Integrating data on a mixed stock anadromous fishery to estimate stock-specific escapement and mortalities

ABSTRACT Objective Anadromous fisheries can affect stocks with differing management objectives. Understanding the effects of fisheries on individual stocks by quantifying final dispositions, including escapement and mortalities, is important for sustainable fishery management. Our objectives were to extend an existing management model that is used to evaluate final stock dispositions to estimate confidence intervals and to generalize the model so that it could be applied in other systems. Methods An annual run reconstruction model for estimating final dispositions of Snake River steelhead Oncorhynchus mykiss was previously developed to evaluate mortalities and escapement for each stock. We extended and generalized this model in two ways: (1) to propagate uncertainty using Monte Carlo simulation and (2) to fit the model using maximum likelihood methods. We generated confidence intervals and conducted sensitivity analyses to evaluate sources of uncertainty. Results The Monte Carlo version of the model explicitly propagates uncertainty in inputs through to model predictions, whereas the maximum likelihood version includes additional data sources, such as catch composition and escapement. We estimated disposition metrics for Snake River steelhead using the original formulation, Monte Carlo version, and the maximum likelihood version for the 2018–2019 season, and we found that the results were similar but that the maximum likelihood version provided a better fit to the observed data. Conclusions We demonstrate how to account for uncertainty when synthesizing several kinds of monitoring data. Improved characterizations of uncertainty facilitate better decision making about future data collection and fisheries management.