Standardized cpue for albacore (Thunnus alalunga) from the Brazilian longline fishery

Standardized CPUE indexes are important components of traditional stock assessment analysis. To be useful the CPUE should be related to the stock biomass. Nevertheless, besides the biomass, several factors (e.g. fishing ground) affect the CPUE. The standardization of the CPUE is necessary to make possible its use as a measure of relative abundance. The generalized linear model (GLM) is an alternative to model CPUE data and to calculate standardized indexes. CPUE data of the albacore (Thunnus alalunga) caught in the west of the South Atlantic by the Brazilian tuna fleet comprised of Chinese Taipei leased longliners were standardized using GLM. An observational log-normal error was assumed for the CPUE. Akaike information criteria (AIC) and analysis of variance (ANOVA) were used to select the relevant factors. The selected model included “year, ” “quarter ” and “area ” as main factors and also second order interactions among these factors. “Area ” was the most important factor to explain variations in CPUE. Residuals of the fitting were normally distributed, confirming the log-normal error assumption was satisfactory. The estimated standardized scaled CPU