Broad-scale climate influences on cod (Gadus morhua) recruitment on Georges Bank

Climatic influences on Georges Bank cod recruitment were investigated using the North Atlantic Oscillation (NAO) as an index of atmospheric variability and the Atlantic Multidecadal Oscillation (AMO) as an index of sea surface temperature. A quantitative approach based on a simple Cushing-type stock-recruitment model was developed and extended to include climate influences using the technique of generalized transfer functions (ARIMAX modelling). This allowed the autoregressive nature of the interacting exogenous and endogenous processes to be taken into account. Based on two information criteria, the resulting best transfer function contains winter NAO with a lag of 3 years, annual AMO with a lag of 1 year (both as exogenous climate factors), log(e)(spawning-stock biomass) as a structural model component, plus two autoregressive parameters. The model is characterized by the smallest information criteria, 92% of explained recruitment variation (vs. 55% from the simple Cushing-type model), excellent forecasting behaviour, and all model assumptions being fulfilled. It is proposed that the model's recruitment hindcasts (ex post forecasts) and forecasts be incorporated into stock and risk assessments as well as management strategy evaluations, either as a climate-induced recruitment index for projections or as real forecasts to establish sustainable cod fisheries on Georges Bank conditioned by climate as a forcing factor.