Debris from fisheries pose significant threats to coastal marine ecosystems worldwide. Tropical tuna purse seine fisheries contribute to this problem via the construction and deployment of thousands of humanmade drifting fish aggregating devices (dFADs) annually, many of which end up beaching in coastal areas. Here, we analyzed approximately 40,000 dFAD trajectories in the Indian Ocean and 12,000 dFAD trajectories in the Atlantic Ocean deployed over the decade 2008-2017 to identify where and when beachings occur. We find that there is tremendous promise for reducing beaching events by prohibiting deployments in areas most likely to lead to a beaching. For example, our results indicate that 21% to 40% (depending on effort redistribution after closure) of beachings can be prevented if deployments are prohibited in areas in the south of 8°S latitude, the Somali zone in winter, and the western Maldives in summer for the Indian Ocean, and in an elongated strip of areas adjacent to the western African coast for the Atlantic Ocean. In both oceans, the riskiest areas for beaching are not coincident with areas of high dFAD deployment activity, suggesting that these closures could be implemented with relatively minimal impact to fisheries. Furthermore, the existence of clear hotspots for beaching likelihood and the high rates of putative recovery of dFAD buoys by small-scale fishers in some areas suggests that early warning systems and dFAD recovery programs may be effective in areas that cannot be protected via closures if appropriate incentives can be provided to local partners for participating in these programs.
Highlights► dFAD beaching rate increased dramatically over 2008-2013 and then stabilized at ~18%. ► dFAD beaching locations clearly identify coastal beaching hotspots. ► Beaching risk depends on the upper ocean circulation and its seasonal variability. ► Prohibiting dFAD deployments in the highest risk areas significantly reduces beaching. ► Spatial closure optimization can account for seasonal variability in beaching risk.