Unmixing soil bacterial diversity considering community microgeography

Similar to human settlements, soil bacteria are distributed into numerous cell clusters whose sizes encode their function and environmental context. We show that bacterial cell cluster size distributions follow log-normal patterns, predicted by proportionate growth (Gibrat’s law). This log-normal distribution of cell cluster sizes is robust across biomes spanning a wide range of resource availabilities. Importantly, we show how characteristic cluster sizes vary with soil carrying capacity and transport limitations. In soils with high bulk cell density, cell cluster size distribution gives rise to rare but disproportionately large, ‘mega’ communities, that, in turn, disproportionately drive metabolism leading to anoxic microsites in largely oxic soils. A cursory evaluation of the statistical features of soil bacterial microgeography suggests that standard bulk soil sampling conflates many small, endemic clusters with a few dominant mega clusters. Consequently, accurate assessment of diversity and composition requires “unmixing” of genetic information across the likely original distributions and bacterial cluster size spectrum. We outline an analytical and modeling framework that translates soil carrying capacity into expected community size heterogeneity based on the observed cell cluster size distributions with heavy tails. We combine global soil and microbiome data sets to model putative community size structures across different ecosystems. Our approach reframes soil microbiomes as size-structured meta-communities and provides testable predictions for diversity-function relationships under changing moisture and carbon regimes.