Environmental modulation of microbial communities

Abstract

Microbial communities are crucial to the health of all ecosystems, and their vast diversity constitutes the majority of species on Earth. A single sample of such a community immediately reveals its complexity, but simultaneously the challenges to understanding its form and function. First, a sample provides a snapshot of a community's current state, but fails to explain how hundreds to thousands of species might emerge and remain together. Second, microbial communities are in constant flux, with species abundances changing in response to biotic interactions with each other as well as abiotic environmental conditions, making it difficult to surmise which forces drive community dynamics. In this thesis, I describe the "bottom-up" approach my colleagues and I use to build microbial communities in the laboratory. By using tractable experimental microcosms and tuning particular abiotic parameters while holding others constant, we discover the effects of environmental changes on community structure. Furthermore, we find that we can verify simple theoretical predictions about how the environment changes interactions and, by extension, community composition. We employ and modify a simple phenomenological model, the Lotka-Volterra interspecific competition model, to make predictions about the effects of increasing mortality, increasing temperature, and environmental fluctuations. We verify these predictions with a diverse set of bacterial species engaged in pairwise competition, and use these pairwise results to successfully predict the outcomes of communities of three or more species. Despite the fact that we knew little about the species other than simple attributes such as their growth rates, the model was generally successful, indicating universal behaviors in response to environmental changes. Our results can provide intuition for experimentalists in tuning laboratory conditions, as well as to scientists studying the effect of the environment on field communities