Recent studies have shown that the establishment and maintenance of the Drosophila melanogaster microbiota depend on constant ingestion of microorganisms, and that bacterial populations vary in richness and diversity over the lifespan of a fly. Animals are constantly faced with new potential invaders of their resident gut microbiota, but only some strains successfully establish themselves. We sought to understand to what extent this variability is probabilistic (i.e., due to chance) versus determined by specific host and bacterial traits. We examined these questions in the context of a defined bacterial community of Acetobacter and Lactobacilli in gnotobiotic flies. Using a modified CAFE assay, we delivered a defined dose of bacteria to many hundreds of individual flies and measured their probability of successful colonization and their growth rate in the fruit fly gut. We focused on three different isolates of Lactobacillus plantarum (from human saliva, from Canton-S lab flies, and from wild flies). We quantified fitness and niche differences in vitro and in the fly gut, finding minor differences in both. High resolution microscopy indicates that Lactobacillus is capable of colonizing distinct regions of the fly gut. Notably, the wild fly strain had a greater colonization probability, a greater population size, and lower variation in population size than the other strains.
We are currently conducting experiments to determine the sources of the variation between individual flies (e.g., immune activity) and between the different bacterial strains. In invasions of conventionally reared flies, we find that variance in the resident microbiota population increases with increasing input dose of commensal bacteria, indicating that invasion destabilizes the gut microbiota and suggesting that host-microbiota feedbacks may affect population dynamics.
Overall our data invalidate a simple lottery model and suggest a bistable system where invader-host feedbacks induce transitions between the two stable states (colonized versus uncolonized), with invasion facilitating more frequent state transitions. The results have implications for not only gut microbiome stability but also for macro-ecological invasions where large variation exists in the success of an invasion.