Catch Efficiency and Biomass Estimates of Walleye Pollock ( Gadus chalcogrammus ) in Relation to Lagrangian Indicators and Surface Temperature of Water in the Okhotsk Sea

A generalized additive model (GAM), incorporating sea surface temperature (SST) as a predictor of catch per unit effort (CPUE; tons per hour) from commercial fishing, has been used for the assessment of the major stock of walleye pollock in the Okhotsk Sea. We evaluated 47 GAM configurations and 5 generalized linear model (GLM) configurations for commercial CPUE standardization, including SST and Lagrangian indicators (LIs) as explanatory variables. Incorporating LIs as a regression spline in the GAM improved model performance. Moreover, the vector autoregressive spatiotemporal model (VAST) and generalized additive mixed model (GAMM) including SST and/or LIs were tested to standardize total stock biomass (TSB), based on area‐swept data (t/km 2 ) from scientific surveys. The stock assessment model using standardized TSB outperformed models without the standardized TSB with respect to retrospective analysis and model convergence. These findings highlight the importance of SST and LIs in standardized indices for biomass estimation using cohort models, thereby improving the stock assessments of walleye pollock in the Okhotsk Sea.