Skillful prediction of tropical Pacific fisheries provided by Atlantic Ninos

Tropical Pacific upwelling-dependent ecosystems are the most productive and variable worldwide, mainly due to the influence of El Nino Southern Oscillation (ENSO). ENSO can be forecasted seasons ahead thanks to assorted climate precursors (local-Pacific processes, pantropical interactions). However, due to observational data scarcity, little is known about the importance of these precursors for marine ecosystem prediction. Previous studies based on Earth System Model simulations forced by observed climate have shown that multiyear predictability of tropical Pacific marine primary productivity is possible. With recently released global marine ecosystem simulations forced by historical climate, full examination of tropical Pacific ecosystem predictability is now feasible. By complementing historical fishing records with marine ecosystem model data, we show herein that equatorial Atlantic sea surface temperatures (SSTs) constitute a valuable predictability source for tropical Pacific fisheries, which can be forecasted over large-scale areas up to three years in advance. A detailed physical-biological mechanism is proposed whereby equatorial Atlantic SSTs influence upwelling of nutrient-rich waters in the tropical Pacific, leading to a bottom-up propagation of the climate-related signal across the marine food web. Our results represent historical and near-future climate conditions and provide a useful springboard for implementing a marine ecosystem prediction system in the tropical Pacific.