Forecasts of swordfish (Xiphias gladius) and common sardine (Strangomera bentincki) off Chile under the A2 IPCC climate change scenario

•We developed a method to assess the impacts of climate change on pelagic fisheries.•We projected a slight decline of relative abundance of swordfish and sardine by 2065.•Warmer waters will produce higher abundance of swordfish nearest to the coast.•Change in abundance and distribution may bring ecological and socio-economic effects.

Recent studies have demonstrated the effects of climate change on both oceanographic conditions and the relative abundance and distribution of fisheries resources. In this study, we investigated the impacts of climate change on swordfish (Xiphias gladius) and common sardine (Strangomera bentincki) fisheries using predictions of changes from global models (according to the NCAR model and IPCC emissions scenario A2), bioclimate envelope models and satellite-based sea surface temperature (SST) estimates from high-resolution regional models for the simulation period 2015-2065. Predictions of SST from global climate models were regionalised using the Delta statistical downscaling technique. The results show an SST trend of 0.0196 °C per year in the study area, equivalent to 0.98 °C for the simulation horizon and for a high CO2 emission scenario (A2). The bioclimate envelope models were developed using historical (2001-2011) monthly environmental and fisheries data. These data included the local relative abundance index of fish catch per unit effort (CPUE), corresponding to the total catch (kg) by 1000 hooks in a 1° latitude × 1° longitude fishing grid for swordfish and to the total catch (ton) by hold capacity (100 m3) in a 10′ latitude × 10′ longitude grid for common sardine. The environmental data included temporal (month), spatial (latitude) and thermal conditions (SST). In the first step of the bioclimate modelling performed in this study, generalised additive models (GAMs) were used as an exploratory tool to identify the functional relationships between the environmental variables and CPUE. These relationships were then parameterised using general linear models (GLMs) to provide a robust forecasting tool. With this modelling approach, environmental variables explained 58.7% of the variation in the CPUE of swordfish and 60.6% of the variation in the CPUE of common sardine in the final GLMs.Using IDRISI GIS, these GLMs simulated monthly changes in the relative abundance and distribution of the studied species forced by changes in the regionalised SST projected by the NCAR model under the A2 emission scenario. The simulations predicted a slight decline of 6% (17 kg/1000 hooks) and 7% (3.8 ton/100 m3) for swordfish and common sardine, respectively, in the spatial mean of the potential relative abundance (CPUE) by 2065.