Soil microbes produce and consume large quantities of greenhouse gases (GHGs) while cycling carbon (C) and nitrogen (N)—particularly carbon dioxide, methane, and nitrous oxide. Understanding the factors that influence the activity of soil microbial communities is therefore valuable for accurate climate modeling. Previous studies have found that with increasing depth within a soil profile, there is strong turnover of microbial communities and processes, with little variation across geospatial locations. Many of these studies, however, used PLFA analysis and compared sites across tens of miles, while the present work analyzes a collection of soil profiles separated by thousands of miles.
We collected soils from 20 unique sites across the United States, encompassing a diversity of biomes, pedology, and land use. Soils were sampled to 100cm at each site, when possible, and high-throughput 16S sequencing was conducted for all ~200 samples. Additionally, about half of the samples were shotgun sequenced for metagenomic analysis, using established KEGG orthology pathways. In combination with a wide variety of edaphic analyses and site-contextual data, we can examine the relative contributions of both depth and location on microbial communities and their functional capabilities.
Results/Conclusions
We found that location is a greater predictor of soil microbial community structure and functional capabilities than depth. This does not necessarily contradict the previous literature, however, since geographically close pairs of sites showed a high degree of community similarity. Furthermore, measurements of community diversity showed that similar ecosystems tended to have more similar trends than did dissimilar environments. For example, soils from grasslands showed a much sharper decrease in diversity with increasing depth, as compared to forests and croplands.
We then sought to examine the trends of greenhouse-gas cycling potential between locations and vertically with depth. We compared the relative abundance of shotgun reads matching genes associated with processes relevant to greenhouse-gas cycling, such as nitrification, denitrification, anammox, methanotrophy, and methanogenesis. None of the processes analyzed showed trends that were significant across depth (0.5 < p < 1), but many sites showed significant differences in relative abundances of genes associated with processes such as nitrification, denitrification, and methanotrophy. In other words, the GHG cycling potential varied more between geospatial locations than across depth within the soil. These findings challenge previous ideas about the importance of location and depth on soil microbial activities, especially as related to climate change.