Validation of the Growth Equation Applicable to the Eastern Atlantic Bluefin Tuna, Thunnus thynnus (L.), Using L-max, Tag-Recapture, and First Dorsal Spine Analysis

Abstract

The growth equation currently used for Atlantic bluefin tuna, Thunnus thynnus (L.), eastern stock (L-t = 318.85 [1-e(-0.093(t + 0.97))]) is validated using several approaches. The first method involved a comparison of studies with von Bertalanffy parameter estimates in which, different methods for the age estimation are utilized, taking as references the maximum size of this species (L-max = 319.93 +/- 11.3 cm) and the growth equation of the western Atlantic stock (L-t = 314.90 [ 1-e(-0.089(t + 1.13))]). The result of this analysis showed that the growth equation used by ICCAT's Standing Committee on Research and Statistics Atlantic bluefin tuna assessment group for the eastern stock perfectly fits L-max. Second, an analysis was realized from first dorsal spine rings, 578 samples (age groups 0 to 3) of ABFT collected from the Bay of Biscay and the Mediterranean Sea, enabled the interpretation of the wide opaque bands (fast growth), formed during the boreal late spring and completed by autumn (June to November), and the translucent rings (hyaline rings, slow growth), formed during boreal autumn to late spring (November to May-June). In addition, first dorsal spine sections bands of two recovered fish that had carried conventional and electronic archival tags are also consistent. The chronological analysis of the opaque bands and hyaline rings of one fish tagged with an archival tag and recovered in the Bay of Biscay (the first time such a spine had been available for such analysis) revealed that transatlantic migrations may lead to double hyaline ring formation in the spine. Finally, the validation of the ABFT growth equation is made by superimposing tag-recovery data from tagging surveys in the Bay of Biscay, western Mediterranean and western Atlantic (N = 131) and spine readings (N = 299) to the eastern stock ABFT growth equation and analysing residuals. The coefficient of determination (R-2 = 97.98) and the residual's distribution indicated good performance of the model. Although no important differences between the growth model of the eastern stock and that of the western stock are found, in all cases studied, the predictive accuracy indicators are better for the eastern model.