Investigating the determinants of soil bacterial community structure in a Mongolian global change experiment
Soils contain some of the most diverse bacterial communities, leaving much to be explored about how these communities are structured. We asked how these communities vary within the landscape between years and with climate and land-use change. We undertook a multi-year experiment to test these effects of global change on soil bacterial communities in northern Mongolia, a region where air temperatures have increased by 1.6 °C since 1960 and traditional land-use patterns are shifting with socio-economic changes.
We used open-top chambers to create warmer and drier conditions and year-round fencing to manipulate grazers (grazed versus ungrazed plots). We collected 68 soil samples in 2010 and 2011 from two south-facing slope locations (lower versus upper). Plant-available nitrogen, soil temperature and moisture were measured for each sampled plot. DNA was extracted and 16S rRNA V4 PCR amplicons were sequenced using the Illumina platform.
We used constrained analysis of proximities (CAP) to test the effects of year, slope location, warming and grazing on bacterial communities. Community composition was examined using taxonomic (Bray-Curtis) and phylogenetic (UniFrac) approaches. We created CAP contour plots to assess the relationship between community composition and the environmental factors we measured.
Bacterial community composition (Bray-Curtis) varied with year and slope, but not warming or grazing. With this analysis, three environmental gradients were significant: total available nitrogen, soil temperature and soil moisture. These environmental gradients could be correlated with significant abundance gradients for three phyla.
Weighted UniFrac dissimilarities revealed significant differences for year and warming and identified available nitrate and moisture as significant environmental variables. No significant abundance gradients for phyla were found.
Year and moisture were robust factors structuring the bacterial community in both the taxonomic (Bray-Curtis) and phylogenetic (UniFrac) analysis, while the significance of slope location, warming, temperature and nitrogen type varies between analyses. Such results suggest that different analyses may provide overlapping, yet distinct ecological perspectives. In the taxonomic analysis, Verrucomicrobia, Planctomycetes, and Firmicutes correlate most closely with temperature, moisture and nitrogen, respectively, giving us hints as to the discrete ecological attributes of these phyla. The phylogenetic analyses found no such taxon-specific correlations but suggested the importance of nitrate and moisture in structuring the community as a whole. These results, both at the phylum and community levels, provide fodder for overarching hypotheses that can act as guidelines in future studies of incredibly complex soil systems.