Elevational Constraints on the Composition and Genomic Attributes of Microbial Communities in Antarctic Soils

Abstract

The inland soils found on the Antarctic continent represent one of the more challenging environments for microbial life on Earth. Nevertheless, Antarctic soils harbor unique bacterial and archaeal (prokaryotic) communities able to cope with extremely cold and dry conditions. These communities are not homogeneous, and the taxonomic composition and functional capabilities (genomic attributes) of these communities across environmental gradients remain largely undetermined. We analyzed the prokaryotic communities in soil samples collected from across the Shackleton Glacier region of Antarctica by coupling quantitative PCR, marker gene amplicon sequencing, and shotgun metagenomic sequencing. We found that elevation was the dominant factor explaining differences in the structures of the soil prokaryotic communities, with the drier and saltier soils found at higher elevations harboring less diverse communities and unique assemblages of cooccurring taxa. The higher-elevation soil communities also had lower maximum potential growth rates (as inferred from metagenome-based estimates of codon usage bias) and an overrepresentation of genes associated with trace gas metabolism. Together, these results highlight the utility of assessing community shifts across pronounced environmental gradients to improve our understanding of the microbial diversity found in Antarctic soils and the strategies used by soil microbes to persist at the limits of habitability. IMPORTANCE Antarctic soils represent an ideal system to study how environmental properties shape the taxonomic and functional diversity of microbial communities given the relatively low diversity of Antarctic soil microbial communities and the pronounced environmental gradients that occur across soils located in reasonable proximity to one another. Moreover, the challenging environmental conditions typical of most Antarctic soils present an opportunity to investigate the traits that allow soil microbes to persist in some of the most inhospitable habitats on Earth. We used cultivation-independent methods to study the bacterial and archaeal communities found in soil samples collected from across the Shackleton Glacier region of the Transantarctic Mountains. We show that those environmental characteristics associated with elevation have the greatest impact on the structure of these microbial communities, with the colder, drier, and saltier soils found at higher elevations sustaining less diverse communities that were distinct from those in more hospitable soils with respect to their composition, genomic attributes, and overall life-history strategies. Notably, the harsher conditions found in higher-elevation soils likely select for taxa with lower maximum potential growth rates and an increased reliance on trace gas metabolism to support growth.

Keywords: Antarctica; microbial ecology; soil microbiology; soils.

FIG 1

Overview of soil bacterial and fungal community compositions across the Shackleton Glacier region. (A) The Shackleton Glacier region (84°S to 85°S, 174°W to 177°W). Locations of the 10 features where samples were collected are indicated by the colored dots (note that 14 to 26 samples were collected from transects at each of the 10 sampling locations). Map by Mike Cloutier, Polar Geospatial Center (Imagery © 2021 Maxar; reproduced with permission). (B) Relative abundances of the most abundant prokaryotic phyla for each of the 167 samples from which 16S rRNA marker gene sequences were obtained. For panel B, samples are organized from the highest elevation site at the top to the lowest elevation site at the bottom.

FIG 2

Phylogenetic tree of the taxa from the Shackleton Glacier region of Antarctica based on 16S rRNA gene sequencing (n = 167 samples). The ASVs represented in this tree include the top 100 most abundant ASVs that were identified from 16S rRNA gene amplicon sequencing. The inset colors indicate the region of the tree associated with each bacterial phylum, while the gray bars around the ASV labels represent the bacterial order (internal dark-gray ring) and family (external light-gray ring). If no taxonomy is indicated, the ASV is not classified to that taxonomic level (taxonomy, “NA” [not applicable]).

FIG 3

Taxonomic compositions and elevational preferences of 13 prokaryotic modules identified from 167 samples. Modules are groups of ASVs (nearly all bacterial) that were identified as cooccurring based on the results of the network analysis. The modules displayed are modules with distributions best explained by site elevation based on the results of the random forest analysis. (A) Average standardized relative abundances (Z score) plotted against elevation of the 3 high-elevation modules and the 10 low-elevation modules. The numbers of ASVs that are included in each module are listed next to the module number. (B) Phylum- and family-level taxonomic identities of those ASVs associated with each of the 13 modules (3 high elevation and 10 low elevation).

FIG 4

Patterns in estimated doubling times of samples (n = 27) across the Shackleton Glacier region. Fifteen samples make up the above 800-m ASL “high-elevation” group, while 12 samples make up the below 800-m ASL “low-elevation” group. Estimated minimum doubling times, as inferred from gRodon (56), were longer (lower maximum potential growth rates) in higher-elevation samples than in lower-elevation samples (P = 0.042 by a Mann-Whitney U test). We note that as gRodon was not designed for use on communities from these types of environments, the estimates of minimal doubling times are presented for comparative purposes only, and the exact values should be considered with caution.

FIG 5

Abundances of nine genes associated with trace gas metabolism across the Shackleton Glacier region. The 27 samples are grouped from lowest elevation to highest elevation and are grouped into high-elevation and low-elevation categories based on whether they were collected above or below 800 m ASL. Groups of genes associated with specific trace gas oxidation pathways are outlined by colored boxes. Z scores were calculated based on the proportional gene abundances, which are presented in Data Set S1D in the supplemental material. Significant differences in gene abundances between the two groups are starred, and the associated statistical information can be found in Fig. S6.