Incorporating Climate Variation into Salmon Forecasts – Do Better Forecasts Improve Management?

Incorporating environmental and climate indicators into stock forecasts has been advocated for improving harvest management. There are examples where use of indicators has improved management-related forecasts, but there has been little evaluation of improvement in terms of achieving management goals. We developed a climate-based predictor for Oregon Coast coho salmon (Oncorhynchus kisutch) that greatly improves management-related forecasts. To investigate whether improved forecasts are likely to improve harvest management, we used a management-strategy evaluation (MSE) approach involving Monte-Carlo simulations of population dynamics linked with harvest strategy implementation. We analyzed the performance of two types of harvest strategies (one precautionary, the other traditional)with respect to two performance metrics: mean annual harvest and frequency of failing a critical spawner density criteria. We found that the precautionary strategy is largely insensitive to forecast skill, while the traditional strategy is sensitive to forecast skill. Our results imply that incorporating environmental variables to improve forecasts is not always useful; the benefits depend on both the general management strategy chosen and the level of precaution adopted in setting harvest targets.