Future distribution of early life stages of small pelagic fishes in the northwestern Mediterranean [Erratum: Sept. 2020, v.162(2), p.989]

We studied the effect of climate change on the potential spawning habitats of two marine small pelagic fishes. We examined the projected changes in the potential spawning habitat of the summer-spawning anchovy (Engraulis encrasicolus) and round sardinella (Sardinella aurita) in the northwestern Mediterranean by combining the regionalized projections of RCP scenarios with an existing species distribution model (SDM). The SDM was based on a separate generalized additive model for the eggs and larvae of the two species computed from ichthyoplankton sampling that was conducted with simultaneous readings of surface temperature, salinity and chlorophyll-a values as predictor variables. The SDM was projected for the 2010 decade, which represented the present-day conditions, with these environmental variables obtained from the regionalized POLCOMS-ERSEM biogeochemical model forced by the RCP 4.5 and RCP 8.5 scenarios. The comparison of the present-day projection results with the projections for the middle and final decades of the twenty-first century showed that the suitability of the spawning habitat as defined by the anchovy eggs model was likely to increase over time under RCP4.5 or decrease slightly under RCP8.5, but the habitat for anchovy larvae was likely to decrease in all cases. Loss of habitat was projected to be particularly important in the south of the study area on the Ebre River delta continental shelf. Conversely, the probability of round sardinella occurrence will significantly increase under both scenarios. The potential habitat of this species, which is of subtropical origin, is likely to shift northwards. The limitations of the existing models to extrapolate the current results to future scenarios are discussed regarding (i) the uncertainty in the projections of driving environmental variables (e.g., chlorophyll-a), (ii) the simplified nature of the projection models, which did not capture the dynamics of the early life stages of the fish at a small scale, and (iii) insufficient consideration of important drivers, such as larval transport or retention by mesoscale hydrographic phenomena.