Contrasting adaptive genetic consequences of stream insects under changing climate

Abstract

Freshwater biodiversity undergoes degradation due to climate change. Researchers have inferred the effects of climate change on neutral genetic diversity, assuming the fixed spatial distributions of alleles. However, the adaptive genetic evolution of populations that may change the spatial distribution of allele frequencies along environmental gradients (i.e., evolutionary rescue) have largely been overlooked. We developed a modeling approach that projects the comparatively adaptive and neutral genetic diversities of four stream insects, using empirical neutral/ putative adaptive loci, ecological niche models (ENMs), and a distributed hydrological-thermal simulation at a temperate catchment under climate change. The hydrothermal model was used to generate hydraulic and thermal variables (e.g., annual current velocity and water temperature) at the present and the climatic change conditions, projected based on the eight general circulation models and the three representative concentration pathways scenarios for the two future periods (2031-2050, near future; 2081-2100, far future). The hydraulic and thermal variables were used for predictor variables of the ENMs and adaptive genetic modeling based on machine learning approaches. The increases in annual water temperature in the near- (+0.3-0.7 °C) and far-future (+0.4-3.2 °C) were projected. Of the studied species, with different ecologies and habitat ranges, Ephemera japonica (Ephemeroptera) was projected to lose rear-edge habitats (i.e., downstream) but retain the adaptive genetic diversity owing to evolutionary rescue. In contrast, the habitat range of the upstream-dwelling Hydropsyche albicephala (Trichoptera) was found to remarkably decline, resulting in decreases in the watershed genetic diversity. While the other two Trichoptera species expanded their habitat ranges, the genetic structures were homogenized over the watershed and experienced moderate decreases in gamma diversity. The findings emphasize the evolutionary rescue potential, depending on the extent of species-specific local adaptation.

Keywords: Ecological niche modeling; Evolutionary rescue; Gradient forests; Hydrological simulation; Invertebrate; Machine learning; Species distribution modeling; Water temperature.