Specialized meltwater biodiversity persists despite widespread deglaciation

Abstract

Glaciers are important drivers of environmental heterogeneity and biological diversity across mountain landscapes. Worldwide, glaciers are receding rapidly due to climate change, with important consequences for biodiversity in mountain ecosystems. However, the effects of glacier loss on biodiversity have never been quantified across a mountainous region, primarily due to a lack of adequate data at large spatial and temporal scales. Here, we combine high-resolution biological and glacier change (ca. 1850-2015) datasets for Glacier National Park, USA, to test the prediction that glacier retreat reduces biodiversity in mountain ecosystems through the loss of uniquely adapted meltwater stream species. We identified a specialized cold-water invertebrate community restricted to the highest elevation streams primarily below glaciers, but also snowfields and groundwater springs. We show that this community and endemic species have unexpectedly persisted in cold, high-elevation sites, even in catchments that have not been glaciated in ∼170 y. Future projections suggest substantial declines in suitable habitat, but not necessarily loss of this community with the complete disappearance of glaciers. Our findings demonstrate that high-elevation streams fed by snow and other cold-water sources continue to serve as critical climate refugia for mountain biodiversity even after glaciers disappear.

Keywords: biodiversity; climate change; glacier loss; invertebrate communities; mountain streams.

Fig. 1.

Glaciers have rapidly receded in GNP since the LIA. (A) Map and histogram of glacial loss for all 129 sampling sites (colored circles) since the end of the LIA (ca. 1850). White to dark shading represents topographic relief. Sampling sites (red points), LIA glacier extent (yellow), and historical photos of glacier loss in Sperry (B), Boulder (C), Grinnell (D) basins are shown. Sites where glaciers were absent in LIA are not included in the histogram.

Fig. 2.

Glacier cover influences the diversity of mountain stream communities. Taxonomic richness and community diversity of stream macroinvertebrates as a function of glacier cover (A and C) and water temperature (B and D). Taxon turnover (β-diversity) between sites and the regional species pool is influenced by glacier cover (C) and water temperature (D). Solid lines are predictions from the best fit linear mixed effects regressions and dotted lines indicate 95% CIs (SI Appendix, Table S1).

Fig. 3.

A specialized cold-water community persists in both glacial and nonglacial streams. (A) Relative abundances of the most represented taxa in each of two distinct communities of alpine invertebrates identified by LDA (SI Appendix, Table S3). (B) The strength of association (magnitude of log-ratio) between taxa and a specific community. (C) The relative abundance of the cold-water community across sites with varying glacier cover (blue), including sites without glaciers (orange). (D) The relative abundance of the cold-water community as a function of distance to source in glacial (blue) and nonglacial streams (orange). (E) A histogram of the relative abundances of the cold-water community across sites that had no glaciers (green), lost their glaciers (yellow), or experienced glacier decline (blue) since the end of the LIA. (F) The relative abundance of the cold-water community as a function of stream temperature in sites where glaciers have been absent or that have lost glaciers since the end of the LIA; Lines (blue and orange) in all panels are model averaged predictions from linear mixed effects models (SI Appendix, Tables S4 and S5).

Fig. 4.

Climate change will likely shrink habitats supporting abundant cold-water communities. Model predicted cumulative distributions of sites supporting the cold-water community under current (black) and future climate (blue) conditions (corresponding percent reductions are dashed lines). Predictions were made using model averaging across linear mixed effects models relating elevation, distance to source, temperature, and glacial cover to the current distribution of the cold-water community (SI Appendix, Table S6).