Thermal niche tracking and future distribution of Atlantic mackerel spawning in response to ocean warming

Abstract

North-east Atlantic mackerel spawning distribution has shifted northward in the last three decades probably in response to global sea warming. Yet, uncertainties subsist regarding on the shift rate, causalities, and how this species will respond to future conditions. Using egg surveys, we explored the influence of temperature change on mackerel's spawning distribution (western and southern spawning components of the stock) between 1992 and 2013, and projected how it may change under future climate change scenarios. We developed three generalized additive models (GAMs): (i) a spatiotemporal model to reconstruct the spawning distribution for the north-east Atlantic stock over the period 1992-2013, to estimate the rate of shift; (ii) a thermal habitat model to assess if spawning mackerel have tracked their thermal spawning-niche; and (iii) a niche-based model to project future spawning distribution under two predicted climate change scenarios. Our findings showed that mackerel spawning activity has shifted northward at a rate of 15.9 ± 0.9 km/decade between 1992 and 2013. Similarly, using the thermal habitat model, we detected a northward shift of the thermal spawning-niche. This indicates that mackerel has spawned at higher latitudes to partially tracking their thermal spawning-niche, at a rate of 28.0 ± 9.0 km/°C of sea warming. Under future scenarios (mid and end of the century), the extrapolation of the niche-based model to coupled hydroclimatic and biogeochemical models indicates that center of gravity of mackerel spawning distribution is expected to shift westward (32 to 117 km) and northward (0.5 to 328 km), but with high variability according to scenarios and time frames. The future of the overall egg production in the area is uncertain (change from -9.3 to 12%). With the aim to allow the fishing industry to anticipate the future distribution of mackerel shoals during the spawning period, future research should focus on reducing uncertainty in projections.