[Applying multi-model inference to estimate growth parameters of greater lizard fish Saurida tumbil in Beibu Gulf, South China Sea].

Age and growth parameters are key parameters in fish stock assessment and management strategies, thus it is crucial to choose an appropriate growth model for a target species. In this study, five growth models were set to fit the length-age data of greater lizard fish Saurida tumbil (n = 2046) collected monthly from December 2006 to July 2009 in the Beibu Gulf, South China Sea. The parameters for each model were estimated using the maximum likelihood method under the assumption of the additive error structure. Adjusted coefficient of determination (R2adj), root mean squared error (RMSE), Akaike's information criterion (AIC), and Bayesian information criterion (BIC) were calculated for each model for fitness selection. The results indicated that the four statistical approaches were consistent in selection of the best growth model. The MMI approach indicated that the generalized VBGF was strongly verified and made up 95.9% of the AIC weight, indicating that this function fitted the length-age data of the greater lizard fish well. The growth function was Lt = 578.49 [1-e -0.05(t-0.14) 0.361.