Modeling the Pacific chub mackerel (Scomber japonicus) environmental niche and future scenarios in the northern Peruvian Current system

Abstract

To understand and characterize the Pacific chub mackerel (Scomber japonicus) spatiotemporal distribution in the northern Peruvian Current System, this study characterized the ecological niche using the minimum volume ellipsoid (MVE) approach considering two meteorological conditions jointly; El Niño (EN) and La Niña (LN). For this purpose, the species presence records collected by the Peruvian onboard observer program from 1997 to 2018 were used. All presence records were matched with the date (day/month/year) and location (longitude, latitude) of the corresponding oceanographic variables (sea surface temperature [SST], sea surface salinity [SSS], chlorophyll, and oxygen). Also, the future projected climate change effects on chub mackerel spatiotemporal distribution under Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios by mid- (2040–2050) and end‐of‐the‐century (2090–2100) were investigated. Ellipsoid evaluations demonstrated that models calibrated with SST and SSS had good performance. Projections for EN conditions predicted areas with high environmental suitability close and along the Peruvian coast. Contrastingly, during LN these areas were extended to the oceanic zone. The projected future scenarios showed that for the RCP 4.5 scenario, environmental suitability maps were similar to those found in LN but with persistent distribution in south-central Peru. For the RCP 8.5 scenario, habitats tended to be coastal. Based on climate refugia –areas where the bulk of the species may aggregate under future environmental change with the potential to re-expand once the stress abates–, the RCP 4.5 scenario did not show changes in the hotspots; conversely, the RCP 8.5 scenario predicted a reduction of around 47% of hotspots from 2040–2050 to 2090–2100 at 10° S-18° S. Finally, the consideration of extreme conditions (EN and LN) in model calibration can potentially generate models with good performance allowing to better characterize species niche when the data is restricted by political boundaries. Nevertheless, further testing must be performed to assert this idea.