Estimation of Japanese survey biomass index of Pacific saury for 2020 using VAST model

Vector Autoregressive Spatio-temporal (VAST) model was applied to Japanese fishery-independent survey data to predict Pacific saury distribution and estimate biomass index for 2020 under an assumption that age-specific fish distribution is consistent over the research period. Validation test indicated that a VAST model incorporating a quadratic function of logarithm of sea surface temperature as a density covariate performed well to estimate biomass index even when the survey area was reduced in the terminal year. The estimated biomass index from the developed VAST model indicated similar year trends with index from a design-based approach after 2011 and the estimate in 2020 dropped to the historical lowest since 2003.