Towards Environmental Control of Microbiomes
Microbial communities have consequential effects on health and the environment yet remain uncontrollable due to their complex dynamics. Ecological modeling offers a platform to overcome their nonlinear and interconnected nature but traditionally does not account for context-dependence. Here, we extend the generalized Lotka-Volterra (gLV) model to accommodate a varying environment by identifying how environmental changes alter species growth rates and interactions in a manner that predicts full community trajectories across environmental gradients. We identify key environment-varying interactions within a synthetic community derived from the Oryzae sativa rhizosphere, and demonstrate how variations in the environment change fixed point compositions and rates of convergence. With our model, we simulate how precise perturbations of the environment can offer improvements in an optimal control problem of driving a community to a target composition. We show that environmental perturbation can minimize the total species input (direct species perturbation) and greatly expand the set of initial states from which a desired target can be reached despite stochasticity. This work demonstrates that a formal perspective on environmental influence of community dynamics is valuable for not only understanding seasonal changes or anthropogenic manipulations, but is critical for improving control of the microbiome.