Neutral models of short-term microbiome dynamics with host subpopulation structure and migration limitation

BACKGROUND: Most empirical studies tend to focus on microbiome dynamics within hosts or microbiome compositional differences between hosts over short periods. However, there is still a dearth of formal models that allow us to investigate the observed short-term dynamics of microbiomes under a unified ecological and evolutionary framework. In our previous study, we developed a computational agent-based neutral framework that simulates microbiome dynamics spanning many host generations with the added dimension of a genealogy of hosts. Although this long-term framework revealed interesting microbial diversity patterns under a simple but plausible evolutionary process and provided a platform for future elaboration of more complex systems, it does not allow us to explore microbiome dynamics within a single host generation. METHODS: In this paper, we developed a computational, agent-based, forward-time framework of microbiome dynamics within a single host generation. As we have done under our neutral long-term models, we incorporate neutral processes of environmental microbiome assembly and microbe acquisition from parents and environment. We also incorporate a Moran genealogical model of hosts, so that the dynamics of microbiome evolution can be studied within a single host generation. Furthermore, we allow host subpopulation structure and host migration to affect microbiome recruitment. RESULTS: We show that microbiome diversity within hosts increases monotonically with increases in environmental contribution, while microbiome diversity between hosts increases with increasing parental inheritance. Host population division and dispersal limitation under high host contribution further shaped the patterns by elevating microbiome differences between hosts and depressing microbial diversity within hosts. Microbiome diversity within the whole population showed strong temporal stability regardless of the modes of microbiome acquisition and subpopulation structures. CONCLUSIONS: We present a computational framework that integrates various processes including host genealogy, microbe recruitment, and host dispersal limitation acting on the short-term dynamics of microbiomes. Our framework demonstrates that the neutral dynamics of microbiomes within a population of hosts is strongly influenced by transmission mode and shared environment. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s40168-018-0464-x) contains supplementary material, which is available to authorized users.