Global patterns in gene content of soil microbiomes emerge from microbial interactions

Abstract

Microbial metabolism sustains life on Earth. Global sequencing surveys reveal ubiquitous patterns linking the taxa and genes responsible for metabolism with environmental conditions. One explanation for these patterns is environmental filtering: local conditions select for strains with particular traits. However, filtering assumes ecological interactions do not influence patterns. Here, we demonstrate a clear example of interactions driving global patterns in topsoil microbiomes. We find a global trade-off between two nitrate reduction genes and pH:nargene abundances increase whilenapdecreases with declining pH. Experiments contradict filtering by revealing strains possessing Nar thrive at low pH due to an interaction with Nap strains that mitigates nitrite toxicity. Thus, associations between environmental variables and gene abundances arise from environmentally modulated interactions driven by conserved metabolic traits.

Dynamic coexistence driven by physiological transitions in microbial communities

“…Our work argues that their simple mechanism of dynamic coexistence -also explored in recent works [41,[48][49][50][51][52] -is more relevant, not less, given the observed complex physiology and nonlinear growth dependencies in real microbial ecosystems. If physiological state changes turn out to be dominant drivers of dynamical niches, as seen in [12,16,53], dynamics cannot be "averaged" over but become the essential link between physiology and ecology.…”

“…This contrasts starkly with dynamical analysis based on commonly used bottom-up models which invariably involve a large number of unconstrained interaction parameters (e.g., the species interaction matrix in generalized Lotka-Volterra models, or the nutrient consumption matrix in Consumer-Resource models). Additionally, it emphasizes intra-cycle dynamics which has been largely neglected except for a few recent studies [16,50,51,53], and gives concrete predictions, e.g., on the growth rate and duration of early vs late species, that can be tested directly by data. In this sense, the Community State model is a phenomenological model that can be updated directly from data.…”

Predictive microbial community changes across a temperature gradient

“…Forecasting changes in biological communities is a central goal in ecology, yet the complexity of these communities makes this a challenging task. Changes in the environment affect the physiology of individual organisms and interactions between them, leading to changes in the composition and function of biological communities (Dell, Pawar, and Savage 2014;Burnside et al 2014;Crocker et al 2023; D. J. Smith and Amarasekare 2018;Amarasekare 2015).…”