Identifying dispersal pathways and critical habitats is essential to evaluate risks and inform effective management strategies of migratory marine species during all life stages. This is especially true for sea turtles that are conservation-dependent and for which management needs usually precedes comprehensive data collection. The aim of this study was to model dispersal pathways (representative of individual behaviour) and compare potential dispersal corridors (representative of population-level behaviour) of hawksbill, loggerhead, leatherback, and green sea turtles from key rookeries in the Western Indian Ocean (WIO) with different dispersal strategies. We used the Sea Turtle Active Movement Model (STAMM) to simulate post-hatchling dispersal under the combined effects of ocean currents and habitat-driven movements. Simulation results confirmed the high connectivity between hatching sites and developmental areas in the WIO; dispersal is mostly driven by ocean currents but differs among species and years with habitat quality also differing among species. Active swimming appeared to have little influence on their dispersal patterns during the first year. We then analysed simulation results using a movement-based kernel density estimation to identify dispersal corridors for each species. There were three distinct dispersal corridors: among equatorial Indian Ocean Islands (hawksbills); along East Africa (green turtles); and around southern Africa (loggerheads and leatherbacks). These results provide a first estimation of the dispersal pathways used by neonate turtles, that are usually lacking in conservation assessments. The results can also assist to develop more targeted management measures like RMU designation or marine spatial planning for the lost years.
«