On the Microgeography of Soil Bacterial Communities

The highly fragmented soil physical environment and the dynamic aqueous phase jointly constrain bacterial life by limiting cell dispersion and modulating diffusion and access to patchy nutrients. A modeling framework that integrates soil hydration conditions with soil organic carbon inputs provides systematic estimates of size distributions and interaction distances among soil bacterial populations. The patterns of bacterial community microgeography provide an important building block for interpreting soil ecological functioning. Experiments and mechanistic modelling show that soil bacterial cluster sizes (measured by counting cell numbers within a community) follow an exponentially truncated power law with parameters (e.g., largest community size) that vary with mean soil water content and carbon inputs across biomes. Theory predicts that similar to human settlement size distributions, tree sizes and other systems in which growth rates are defined by the environment independent of the object size, the resulting bacterial community size distributions is likely to obey the so-called Gibrats law. Our results support a potential for generalization using positively skewed distributions of soil bacterial community sizes (e.g., log normal and Gamma). We show that soil bacteria reside in many small communities (with over 90% of soil bacterial communities having less than 100 cells), supported by theoretical predictions of log-normal distribution for non-interacting soil bacterial community sizes with scaling parameters that vary with biome characteristics. We will use estimates of the fraction of the largest bacterial communities where anoxic conditions may develop under prevailing conditions to constrain the number of anoxic hotspots per soil volume.