A Bayesian Model to Estimate Movement Rates from Historical Tagging Data and Catch per Unit Effort Data for Atlantic Tropical Tunas

Movement rates of Atlantic tropical tunas, Thunnus obesus, T. albacares, and Katsuwonus pelamis, were estimated using a spatially structured Bayesian tagging model informed by conventional tagging data for from the International Commission for the Conservation of Atlantic Tuna (ICCAT) tagging database and relative abundance indices obtained by standardizing CPUE data. Bayesian posterior movement parameters were estimated using a three region model of the entire Atlantic for yellowfin, an eight region model of the Eastern Atlantic for bigeye, and a six-region model of the Eastern Atlantic for skipjack.  Catch per unit effort data were standardized in each of the eight regions for each of the three species using a Delta lognormal model. Migration parameters for yellowfin showed less than 10% of the population within a region moving to a different region annually.  Individual bigeye move frequently between regions, with net seasonal northward movement between the Gulf of Guinea and northeast Atlantic. Skipjack tuna net movement tends to be towards the northwest Atlantic. Both bigeye and skipjack exhibit non-directional movements within the Gulf of Guinea.  This study represents the first attempt to quantitatively estimate movement rates and the timing of the movement of tropical tunas.