Model-based clustering for estimating cetaceans site-fidelity and abundance

Estimating the size of animal populations in a given area is of particular interest in ecological studies on wildlife conservation, and this task is commonly handled via capture-recapture methods. A recent work (Pace et al., 2021) adopts a two-step approach for identifying groups of animals with similar site-fidelity patterns - according to specific metrics - and estimating the abundance of bottlenose dolphins between 2017 and 2020 at the Tiber Estuary (Mediterranean Sea, Rome, Italy). In this work, we aim at simultaneously classifying individuals and estimating their abundance in the study area, by introducing finite mixtures within the Open-Population Jolly-Seber framework. In capture-recapture analyses, finite mixture models allow to account for groups heterogeneity and to reduce the bias in the final abundance estimates (Pledger, 2005).