On the Variability of the Length¿Weight Relationship for Atlantic Bluefin Tuna, Thunnus thynnus (L.)

Autores UPV
Año
Revista REVIEWS IN FISHERIES SCIENCE

Abstract

Following extensive review, a model of the Atlantic bluefin tuna (ABFT), Thunnus thynnus (L.), length¿weight relationship for the eastern Atlantic and Mediterranean (RW = 0.0000188 *SFL^3.01247; Ec 1) is presented on the basis of samples of ABFT spawners, with an average value of index K = 2.03 +- 0.15SD, collected by the Atlantic traps of Portugal and Spain in the Strait of Gibraltar (1963; 1996¿1998; 2000¿2012), and a set of samples of juvenile fishes from ICCAT¿GBYP (n = 707). The resulting model (Ec 1), together with the model used for the eastern stock assessment (RW = 0.000019607 *SFL^3.0092; Ec 2) and a recently adopted by ICCAT Standing Committee on Research and Statistics (SCRS) (RW=0.0000315551*SFL^2.898454; EAST) are analyzed in using a bi-variant sample [SFL (cm), RW (kg)] of 474 pairs of data with the aim of validating them and establishing which model(s) best fit the reality represented by the sample and, therefore, will have the greatest descriptive and predictive power. The result of the analysis indicates that the model EAST clearly underestimates the weight of spawning ABFT and that model Ec 2 overestimates it slightly, being model Ec 1 that best explains the data of the sample. The result of the classical statistical analysis is confirmed by means of the quantile regression technique, selecting the quantiles 5, 25, 50, 75, and 95%. Other fisheries and biological indicators also conclude that the model EAST gradually underestimates the weight of ABFT spawners (of 2¿3 m) by 9¿12.5 %, and does not meet the criterion that for RW = 725 kg (Wmax), SFL=319.93+-11.3 cm (Lmax).