SEARCH

SEARCH BY CITATION

Keywords:

  • Climate change;
  • ensemble forecasting;
  • environmental gradient;
  • homogenization;
  • stream fish assemblages

Abstract

Aim  To investigate the potential impacts of climate change on stream fish assemblages in terms of species and biological trait diversity, composition and similarity.

Location  One-thousand one-hundred and ten stream sections in France.

Methods  We predicted the future potential distribution of 35 common stream fish species facing changes in temperature and precipitation regime. Seven different species distribution models were applied and a consensus forecast was produced to limit uncertainty between single-models. The potential impacts of climate change on fish assemblages were assessed using both species and biological trait approaches. We then addressed the spatial distribution of potential impacts along the upstream–downstream gradient.

Results  Overall, climate change was predicted to result in an increase in species and trait diversity. Species and trait composition of the fish assemblages were also projected to be highly modified. Changes in assemblages’ diversity and composition differed strongly along the upstream–downstream gradient, with upstream and midstream assemblages more modified than downstream assemblages. We also predicted a global increase in species and trait similarity between pairwise assemblages indicating a future species and trait homogenization of fish assemblages. Nevertheless, we found that upstream assemblages would differentiate, whereas midstream and downstream assemblages would homogenize. Our results suggested that colonization could be the main driver of the predicted homogenization, while local extinctions could result in assemblage differentiation.

Main conclusions  This study demonstrated that climate change could lead to contrasted impacts on fish assemblage structure and diversity depending on the position along the upstream–downstream gradient. These results could have important implications in terms of ecosystem monitoring as they could be useful in establishing areas that would need conservation prioritization.