UW Aquatic & Fishery Sciences Quantitative Seminar

Jim Thorson

NOAA, NWFSC

Surplus production: New directions for one of the oldest models in fisheries

Abstract

Surplus production models have been an integral tool in fisheries science for over fifty years,and continue to be used for fisheries management today. These models have changed with the times, progressing from equilibrium to state-space estimation approaches, and continue to shape concepts of fisheries yield and multi-species interactions. In this presentation, I argue that surplus production models still have a vital role in the advancement of fisheries science using two examples from my research. I first explore a semi-parametric approach to estimating the relationship between sustainable yield and population biomass (called the production function). This semi-parametric approach specifies a Bayesian prior on the production function, while allowing data to update the prior when data are informative. By applying the approach to state- space surplus production models, I show that resulting parameter estimates can be used to diagnose and account for model-misspecification. Second, I modify the conventional surplus production model to use a parametric function representing the evolution of fishing effort over time (“effort dynamics”) in place of the conventional index of abundance or fishing effort. I then use theoretical arguments and simulation testing to show that the resulting model can, in some cases, reconstruct biomass using only catch data. This model is applied to catch data for assessed stocks off the U.S. West Coast, and compares favorably with other data-poor methods and stock assessment estimates of biomass. I conclude with other ideas for novel research based on the surplus production model, and encourage future research that uses surplus production models to develop and illustrate advancements in the theory and estimation of fishery models.