With climate change challenging marine biodiversity and resource management, it is crucial to anticipate future latitudinal and depth shifts under contrasting global change scenarios to support policy-relevant biodiversity impact assessments [e.g., Intergovernmental Panel on Climate Change (IPCC)]. We aim to demonstrate the benefits of complying with the Paris Agreement (United Nations Framework Convention on Climate Change) and limiting environmental changes, by assessing future distributional shifts of 10 commercially important demersal fish species.
Time period
Analyses of distributional shifts compared near present-day conditions (2000–2017) with two Representative Concentration Pathway (RCP) scenarios of future climate changes (2090–2100): one following the Paris Agreement climate forcing (RCP2.6) and another without stringent mitigation measures (RCP8.5).
Methods
We use machine learning distribution models coupled with biologically meaningful predictors to project future latitudinal and depth shifts. Structuring projections with information beyond temperature-based predictors allowed us to encompass the physiological limitations of species better.
Results
Our models highlighted the additional roles of temperature, primary productivity and dissolved oxygen in shaping fish distributions (average relative contribution to the models of 32.12 ± 10.24, 15.6 ± 7.5 and 12.1 ± 6.1%, respectively). We anticipated a generalized trend of poleward shifts in both future scenarios, with aggravated changes in suitable area with RCP8.5 (average area loss with RCP2.6 = 13.3 ± 4.1%; RCP8.5 = 40.9 ± 13.3%). Shifts to deeper waters were also predicted to be of greater magnitude with RCP8.5 (average depth gain = 25.4 ± 21.5 m) than with RCP2.6 (average depth gain = 10.4 ± 7.9 m). Habitat losses were projected mostly in the Mediterranean, Celtic and Irish Seas, the southern areas of the North Sea and along the NE coast of North America.
Main conclusions
Inclusion of biologically meaningful predictors beyond temperature in species distribution modelling can improve predictive performances. Limiting future climate changes by complying with the Paris Agreement can translate into reduced distributional shifts, supporting biodiversity conservation and resource management.
