The microbiome of cryospheric ecosystems

Abstract

The melting of the cryosphere is among the most conspicuous consequences of climate change, with impacts on microbial life and related biogeochemistry. However, we are missing a systematic understanding of microbiome structure and function across cryospheric ecosystems. Here, we present a global inventory of the microbiome from snow, ice, permafrost soils, and both coastal and freshwater ecosystems under glacier influence. Combining phylogenetic and taxonomic approaches, we find that these cryospheric ecosystems, despite their particularities, share a microbiome with representatives across the bacterial tree of life and apparent signatures of early and constrained radiation. In addition, we use metagenomic analyses to define the genetic repertoire of cryospheric bacteria. Our work provides a reference resource for future studies on climate change microbiology.

Fig. 1. A unique cryospheric microbiome.

A Geographic distribution of the 16 S rRNA gene samples for the two primer pairs (PP) and metagenomes for both cryospheric and non-cryospheric ecosystems, where GPS coordinates were available on NCBI. Symbol size denotes the number of samples per site (see Supplementary Table 7). B Phylogenetic tree based on abundant ASVs (>0.5% relative abundance in at least one sample) in the PP1 dataset. The heatmap (inner rings) shows the presence (at a > 0.5% relative abundance threshold) of ASVs in the four ecosystem types of the cryosphere (ice and snow, terrestrial, coastal ocean and freshwater). The barplot (outer ring) represents the coefficient for the SVM classifier analysis, highlighting discriminating ASVs. C Sorensen’s phylogenetic index of β-diversity (n₁ = n₂ = 84,461 for PP1, and n₁ = n₂ = 99,000 for PP2) and D β-MNTD calculated across pairs of samples in the cryospheric samples (Cryo-Cryo), pairs of cryospheric and non-cryospheric samples (Cryo-Others) and pairs of non-cryospheric (Others-Others) samples (sample sizes are listed in Supplementary Table 2). The top panel (shades of blue) is for PP1, the bottom one (shades of red) for PP2; two-sided Wilcoxon tests were performed to assess significance in panels C and D; the Holm method was used to correct for multiple testing (****: 0–0.0001). Boxplots depict the median and the 25th and 75th quartiles, whiskers extend to values within 1.5 times the interquartile range, and the remaining points are outliers. Effect sizes and exact p-values are available in Supplementary Table 2. Source data are provided as a Source Data file.

Fig. 2. Cryospheric genera and shared genomic properties.

A Taxonomic tree representing the number of cryospheric genera per taxon with colours. Only taxa containing at least two cryospheric genera are shown (down to the class level). B Bar plot showing the bacterial genera significantly enriched in the cryosphere with the highest centered log-ratio (CLR) mean difference (based on ANCOM analysis). The colours represent the phylum level taxonomic classification. Source data are provided as a Source Data file.

Fig. 3. Microbiome structure across various cryospheric ecosystems.

A The probability of presence of members of the core microbiome is shown across cryospheric environments. Colours and facets separate phylum-level taxonomic affiliation. B Unconstrained ordination showing differences (Bray-Curtis dissimilarity) of bacterial communities among different cryospheric ecosystems (k = 2, stress = 0.206). Hulls demark 95% confidence intervals for a multivariate t distribution for the respective ecosystem types. C Mean relative abundance (in percentage) of core/ancillary and cryospheric/others bacterial genera across the four ecosystem types and the two primer pairs. Source data are provided as a Source Data file.

Fig. 4. Functional enrichment analysis and taxonomy of enriched functions.

A Prevalence represented against the number of enriched KOs in cryospheric samples across bacterial genera. The shading represents the cryospheric and others bacterial genera whereas symbol size represents the number of contigs taxonomically assigned to the respective genus within cryospheric metagenomes. The insert represents the distribution of the number of contigs harbouring cryospheric enriched KOs across the cryospheric and others genera. B Heatmap representing the completion of pathways across cryospheric samples based only on the KOs enriched in the cryosphere. Source data are provided as a Source Data file.