Seasonal ocean forecasts to improve predictions of Dungeness crab catch rates, co-developed with state and tribal fishery managers

Historical (hindcast) and September initialized forecast fields from the J-SCOPE system (more information here: https://www.nanoos.org/products/j-scope/). The historical fields are annual averages while the forecasts are averaged monthly.  The fields here were used to generate the CPUE models described in Norton et al. (2023). 

Abstract from the paper: The commercial Dungeness crab (Metacarcinus magister) fishery in Oregon and Washington (USA) is one of the most valuable fisheries in the region, but it experiences high interannual variability. These fluctuations have been attributed to environmental drivers on seasonal and annual timescales. In this study, researchers and state and tribal fisheries managers develop a statistical model for Dungeness crab catch per unit effort (CPUE) to help inform dynamic management decisions in Oregon and Washington. Fishing observations were matched to seasonally forecast and lagged ocean conditions from J-SCOPE, a regional forecast system (http://www.nanoos.org/products/j-scope/). Inclusion of dynamic and lagged ocean conditions improved model skill compared to simpler models, and the best model captured intraseasonal trends and interannual variability in catch rates, and spatial catch patterns. We also found that model skill relied on fishing behaviour, which varies interannually, highlighting the need for advanced fishing behaviour modelling to reduce uncertainty. The relationships between catch rates and ocean conditions may help elucidate environmental influences of catch variability. Forecast products were co-designed with managers to meet their needs for key decision points. Our results illustrate a seasonal forecasting approach for management of other highly productive, but also dynamic, invertebrates that increasingly contribute to global fisheries yield.