Résumé
It has been proposed that animals migrating towards a specific destination may rely on vector navigation by maintaining a direction for a certain amount of time. In sea turtles, reliance on this strategy has been proposed especially for migrations directed towards wide targets and carried out in weak sea current flows. In the present study we tested if vector navigation could also be a feasible strategy for turtles facing dynamic oceanographic conditions by considering the case of green turtles nesting in the Comoros archipelago and migrating to their foraging grounds along the African coast. To test this hypothesis, we firstly analysed the turtles’ actual headings estimated considering the currents encountered by migrating turtles along the open sea segment of their routes. We then run individual-based models to simulate the journeys of turtles migrating in the area while relying on a vector or a true navigation strategy, and reconstructed the turtle’s water-related, i.e. motor paths, by removing the drifting effect of sea currents. Tracked turtles did not show any major change in their headings and mostly oriented towards the foraging area. While simulations did not provide homogeneous results, the turtle orientation efficiency estimated from motor paths showed that they possibly relied on a vector rather than a true navigation strategy. The present results suggest that vector navigation is a viable strategy to account for the migratory performances recorded in turtles migrating in dynamic oceanographic conditions, even if the involvement of the more sophisticated true navigation mechanism cannot be completely excluded