Assessing reliance on vector navigation in the long-distance oceanic migrations of green sea turtles

Vector navigation, i.e., maintaining a constant heading for a given amount of time, is hypothesized to provide a viable basis for the navigational feats of a number of long-distance animal migrants. Since animals following this strategy are subject to drift by wind or by ocean current, performing long migrations relying on vector navigation is particularly challenging. We tested whether vector navigation could be involved in the migrations of green turtles (Chelonia mydas) that migrate between the remote Ascension Island and Brazil. To this aim, a novel approach was followed using individual-based numerical models to simulate migratory trajectories of virtual turtles that were compared to actual routes reconstructed by satellite. Simulated postnesting migrations from Ascension revealed that weak currents enabled modeled turtles to reach the Brazilian coast, but only for a limited range of headings around due West. This conclusion was corroborated by comparing modeled trajectories with the actual routes of previously tracked turtles, with a beeline vector navigation strategy providing the best fit, although a true-navigation strategy directed to the landfall site produced similar results. Finally, we tested if a vector navigational strategy was feasible for the prebreeding migration from Brazil towards Ascension, but modeled routes mostly failed to reach the island or a larger area around it, with individuals drifting away under the influence of currents. We conclude that Ascension turtles can take advantage of vector navigation when migrating towards a wide target like the Brazilian coast, while the demanding prebreeding migration likely requires more complex navigational systems.