Incorporating spatial and seasonal dimensions in a stock reduction analysis for lower Fraser River white sturgeon (Acipenser transmontanus)

Abstract

We applied a spatially and seasonally structured stock reduction analysis (SRA) model to white sturgeon ( Acipenser transmontanus ) in the lower Fraser River, British Columbia, to estimate trends in abundance since the 1800s and evaluate the current status of the population. We used a sequential Bayesian state–space estimation approach to incorporate prior information from other analyses and evaluate the updating of prior knowledge within the SRA model. The estimated ratio of the abundance of spawning fish in 2004 to relative to unfished conditions was slightly higher than estimates from other studies; on average, 27% of the posterior probability was associated with a 2004 spawning stock abundance of 50% or less of the unfished abundance. Estimates of the current abundance of fish vulnerable to the lower Fraser River recreational fishery were higher than those obtained in other recent SRAs that ignored spatial structure. We also performed the analysis using a spatially aggregated version of the SRA model and obtained lower estimates of unfished biomass and depletion and higher estimates of fishing mortality rates compared with the spatially structured model. We evaluated two structural hypotheses about age-specific vulnerabilities in the historical commercial fishery; assumed vulnerabilities had a marked impact on estimated fishing mortality rates.