COS 9-6 - Structured Community Transitions Explain the Switching Behavior of Microbial Systems

Abstract: Microbial systems within a host appear to exhibit a relatively stable community of species, especially within healthy hosts. However, over short timescales, these systems tend to switch (go back and forth) among few equally dominant communities, increasing the uncertainty of whether a system will change its community or not---similar to the uncertainty derived from flipping a fair coin. While this switching behavior has been mainly attributed to external factors acting on individual species, its association with dysbiosis or inherent structural processes operating at the community level remains unclear. Here, we integrate ecological theory and clinical data to demonstrate that structured community transitions can explain the switching behavior of microbial systems. We propose that unstructured community transitions happen when the transition probability from one community to another is simply proportional to the volume of parameter space of external conditions compatible with such communities. Instead, structured community transitions happen when the transition probability between two communities is, in addition, inversely proportional to their distance in parameter space---which can be treated as a special case of the gravity model. We showcase our theory using generalized Lotka-Volterra dynamics. We corroborate our theory using clinical temporal data of human gut and vaginal microbiota. Our results illustrate how internal constraints operating at the community level are inherently linked to the changing behavior and predictability of microbial systems.