Summer Dynamics of Microbial Diversity on a Mountain Glacier

Abstract

Glaciers are rapidly receding under climate change. A melting cryosphere will dramatically alter global sea levels, carbon cycling, and water resource availability. Glaciers host rich biotic communities that are dominated by microbial diversity, and this biodiversity can impact surface albedo, thereby driving a feedback loop between biodiversity and cryosphere melt. However, the microbial diversity of glacier ecosystems remains largely unknown outside of major ice sheets, particularly from a temporal perspective. Here, we characterized temporal dynamics of bacteria, eukaryotes, and algae on the Paradise Glacier, Mount Rainier, USA, over nine time points spanning the summer melt season. During our study, the glacier surface steadily darkened as seasonal snow melted and darkening agents accumulated until new snow fell in late September. From a community-wide perspective, the bacterial community remained generally constant while eukaryotes and algae exhibited temporal progression and community turnover. Patterns of individual taxonomic groups, however, were highly stochastic. We found little support for our a priori prediction that autotroph abundance would peak before heterotrophs. Notably, two different trends in snow algae emerged-an abundant early- and late-season operational taxonomic unit (OTU) with a different midsummer OTU that peaked in August. Overall, our results highlight the need for temporal sampling to clarify microbial diversity on glaciers and that caution should be exercised when interpreting results from single or few time points. IMPORTANCE Microbial diversity on mountain glaciers is an underexplored component of global biodiversity. Microbial presence and activity can also reduce the surface albedo or reflectiveness of glaciers, causing them to absorb more solar radiation and melt faster, which in turn drives more microbial activity. To date, most explorations of microbial diversity in the mountain cryosphere have only included single time points or focused on one microbial community (e.g., bacteria). Here, we performed temporal sampling over a summer melt season for the full microbial community, including bacteria, eukaryotes, and fungi, on the Paradise Glacier, Washington, USA. Over the summer, the bacterial community remained generally constant, whereas eukaryote and algal communities temporally changed through the melt season. Individual taxonomic groups, however, exhibited considerable stochasticity. Overall, our results highlight the need for temporal sampling on glaciers and that caution should be exercised when interpreting results from single or few time points.

Keywords: alpine; alpine glacier; biological albedo reduction; cryosphere; glacier biology; snow algae.

FIG 1

(a) Location of our study site on the Paradise Glacier of Mount Rainier, Washington, USA. Imagery from Google Earth. (b) Sentinel-2 satellite imagery of the study site from late May to September 2019. A fresh snowfall occurred between the final two sampling time points in September. (c) A conceptual image of primary production and heterotroph activity on a temperature mountain glacier over the course of 1 year. Overlaid on this conceptual framework are monthly averages of minimum temperature, average precipitation, and average snowfall for the nearby Paradise Ranger Station (1.655 m) from 1916 to 2016 (data from the Western Regional Climate Center). (d to f) Principal coordinate analysis of community composition based on Bray-Curtis dissimilarity for (d) bacteria, (e) eukaryotes, (f) and fungi (g to i). Shannon diversity through time for the same sampling points and communities: (g) bacteria, (h) eukaryotes, and (i) fungi. Data were binned by sample collection month for visualization (see Table S1 in the supplemental material), and each circle in d to i represents one replicate. After processing, our final sample sizes were the following: 16S rRNA, May: n = 2, June: n = 3, July: n = 5, August: n = 4, September: n = 2, late September: n = 2. For 18S rRNA, May: n = 2, June: n = 3, July: n = 3, August: n = 3, September: n = 3, late September: n = 3. For ITS, May: n = 2, June: n = 3, July: n = 4, August: n = 4, September: n = 2, late September: n = 2.

FIG 2

Temporal abundance of common taxonomic groups for each data set overall and broken down for select taxa: (a and b) 16S rRNA, (c and d) 18S rRNA, and (e and f) fungal ITS. Each circle represents one replicate. Taxonomic groups comprising the largest percent relative abundance in each library are shown in a, c, and e. The most abundant operational taxonomic units (OTUs) in each data set are shown in b, d, and f, where taxonomy has been assigned to each OTU at the highest resolution possible (see detailed methods in the supplemental material). Box plots show mean percent relative abundance of the group (a, c, and e) or OTU (b, d, and f). Data sets are binned by month of sample collection except for early and late September, where the first seasonal snow occurred between sampling efforts.