Pelagic resources landings in central-southern Chile under the A2 climate change scenarios

Abstract

Artificial neural networks (ANNs) were used to predict landings of anchovy (Engraulis ringens), common sardine (Strangomera bentincki), and jack mackerel (Trachurus murphyi) in central-southern Chile. Twelve environmental variables were considered along with fishing effort (fe) and landing statistics from 1973 to 2012. During external validation, the best models with all of the selected variables gave r2 values of 90 % for anchovy, 96 % for common sardine, and 88 % for jack mackerel. The models were simplified by considering only fe and sea surface temperature from NCEP/NCAR reanalysis data (SST-NOAA), and very similar fits were achieved (87, 92, and 88 %, respectively). Future SSTs were obtained from the A2 climate change scenario and regionalized using statistical downscaling techniques. The downscaled SSTs were used as input for landings predictions using ANN simplified models. In addition, three scenarios of future fishing efforts (2010–2012 average, average + 50 %, and average − 50 %) were used as the input data for landing simulations. The results of the predictions show a decrease of 9 % in future landings of sardine and an increase of 17 % for jack mackerel when comparing 2015 and 2065 monthly projections. However, no significant differences are shown when comparing the estimated landings for the three fishing effort scenarios. Finally, more integrative and complex conceptual models that consider oceanographic-biophysical, physiological, environmental-resource, and interspecies processes need to be implemented.