Abstract
Modern fisheries management uses complex models to simulate entire ecological systems, which requires a lot of environmental, behavioral, and spatial information. For the multi-species fisheries of Atlantic yellowfin (Thunnus albacares), bigeye (Thunnus obesus) and skipjack (Katsuwonus pelamis) tuna, mixed schooling is a major concern for sustainability. Juvenile bigeye and yellowfin tunas are caught alongside adult skipjack, depleting the reproductive stock of bigeye and yellowfin, leading to population decline of yellowfin and bigeye, and unoptimized catch of skipjack. To reduce juvenile tropical tuna mortality, we developed an agent-based model of tropical tuna movement behavior in the Atlantic Ocean which can highlight areas of significant species overlap. Agent-based models are simulated environments with fish agents that allow us to identify emergent behaviors and test alternative management strategies in a controlled space. To parameterize the model, we completed a systematic review of tropical tuna movement drivers including species-specific speeds, temperature preferences, oxygen preferences, and fish aggregation device interactions. We mapped mark-recapture data from the Atlantic Ocean Tropical tuna Tagging Program (AOTTP) and used remote sensing methods to identify temperature preferences and general spatial trends after 4 months’ time at liberty. A final agent-based model was constructed to simulate a mark-recapture event where all three species were released in one location at the same time and allowed to disperse over the course of a year. The simulated tropical tunas of each species were compared to their real-life counterparts, and we compared movement treatments to identify the closest match. We concluded that tropical tunas follow FADs and prefer different temperature ranges on a species-specific basis. This conclusion could be expanded by increasing the complexity of the agent-based model in future work.