Development of a Bayesian Mixed-Stock Model for Determining the Origin of Atlantic Bluefin Tuna

Understanding the nature and magnitude of trans-boundary exchange of fish stocks is fundamental to assessment and management of fisheries resources. Recovery of Atlantic bluefin tuna (Thunnus thynnus) populations is confounded by a lack of knowledge regarding trans-Atlantic movement and mixing of individuals from eastern and western stocks. Geochemical tagging shows considerable promise for understanding of connectivity between eastern and western stocks; however, uncertainty remains regarding the actual proportion of expatriates on either side of the management boundary.  While this is due in part to the quality of the baseline (reference) sample, estimates appear to vary considerable depending upon the mixed-stock approach being used. Therefore, a detailed evaluation of mixed-stock procedures used with geochemical data is needed to verify currently accepted ratios and refine the mixed-stock framework.  In this study, we present a Bayesian mixed-stock model for Atlantic bluefin tuna and compare its utility with several methods for classification and mixed-stock analysis.