In-season assessment and management of salmon stocks using a Bayesian time–density model

Abstract

We document a time–density model for in-season assessment of salmon stocks that integrates both relative and absolute indicators of abundance and incorporates preseason information on run size and migration timing using a Bayesian framework. We evaluate different data collection programs for Fraser River sockeye salmon (Oncorhynchus nerka) by examining the precision, bias, and timeliness of resulting run-size estimates with a retrospective analysis of 20 years of data. We quantify the run-size bias if migration was early versus late and evaluate the impact of run-size uncertainty on the ability to reach management objectives. In-season assessments greatly improve the accuracy and precision of run-size estimates compared with preseason forecasts. For the in-season assessment of Fraser River sockeye, catch-per-unit-effort (CPUE) data from seaward marine test fisheries, although less precise, were more informative at the peak marine migration than more precise terminal, in-river hydroacoustic data obtained on the same date due to the time lag associated with migration from the marine test fishing locations to the lower river. Throughout the season, the best fisheries management results were obtained by relying on in-season assessments using both marine CPUE data as well as marine reconstructed abundance estimates derived from in-river hydroacoustic estimates, thereby taking advantage of the benefits of both sources of information.