Predicting habitat to optimize sampling of Pacific sardine (Sardinops sagax)

More than 40 years after the collapse of the fishery for Pacific sardine, a renewed fishery has emerged off the west coasts of the United States and Canada. The daily egg production method (DEPM) and acoustic-trawl surveys are performed annually and, to minimize the uncertainties in the estimates, sampling effort needs to be allocated optimally. Here, based on a 12-year dataset including the presence/absence of sardine eggs and concomitant remotely sensed oceanographic variables, a probabilistic generalized additive model is developed to predict spatio-temporal distributions of habitat for the northern stock of Pacific sardine in the California Current. Significant relationships are identified between eggs and sea surface temperature, chlorophyll a concentration, and the gradient of sea surface altitude. The model accurately predicts the habitat and seasonal migration pattern of sardine, irrespective of spawning condition. The predictions of potential habitat are validated extensively by fishery landings and net-sample data from the northeast Pacific. The predicted habitat can be used to optimize the time and location of the DEPM, acoustic-trawl, and aerial surveys of sardine. The method developed and illustrated may be applicable too to studies of other stocks of sardine and other epipelagic fish in other eastern boundary, upwelling regions.