An integrative framework to combine migratory connectivity and demographic data

Migratory species experience various conditions and events throughout their annual cycle that influence their spatial and demographic dynamics. To understand these dynamics, it is essential to describe the origin and destination of individuals. Migratory connectivity, which is defined as the geographic linkage between populations across the annual cycle, is increasingly incorporated in population models to relate population trends to environmental variables at different stages of the cycle. However, such information on migratory movements is obtained independently from the study of population dynamics despite the interaction between both processes. Expanding on the growing use of integrated modelling approaches, we developed an integrated framework that allows the sharing of information between migratory connectivity and population data.We first assembled an integrated migratory connectivity model and an integrated population model to join the analysis of GPS, live-reencounter, dead-recovery, capture–mark–recapture, and population count data within a unified framework. Based on simulated data, we assessed the ability of the resulting integrated connectivity and population model to produce unbiased and precise connectivity and demographic estimates. We then applied the same assessment to real data using the Eurasian Curlew (Numenius arquata) as a case study.On simulated data, the integrated connectivity and population model estimated connectivity and survival parameters with no bias and similar precision to the connectivity model alone. However, it outperformed the population model in estimating fecundity in the absence of explicit productivity data. When applied to the Eurasian Curlew, the integrated connectivity and population model produced overall similar migratory connectivity and more accurate demographic estimates than the connectivity model alone, consistent with previous studies. Additionally, the model was able to estimate fecundity, whereas the data were too sparse for the population model alone to disentangle juvenile survival and fecundity.The sharing of information between migratory connectivity and population data improved the estimation of demographic parameters by the population model and improved connectivity parameter estimates when data were scarce. This flexible framework can be generalised to include diverse data on migration movements, population structure, individual heterogeneity or environmental variables, allowing further investigation of the interaction between migration patterns and population dynamics.