Habitat filtering at multiple scales : a case study with crab-associated microbial communities

Local diversity is the result of colonization from a species pool and local habitat filters. The relative contribution of these processes delineates our ability to predict changes in biodiversity, key in a changing world. Although theory suggests that these interactions are critical for the maintenance of biodiversity, empirical work has been restricted by limitations of studies across biogeographical regions and laboratory microcosms. Alternatively, host-associated microbial communities represent discrete, tractable and replicable assemblages. Hosts provide distinct habitat patches with different habitat filters and are surrounded by environmental bacteria that determine the regional species pools. My dissertation focuses on colonization and habitat filtering processes driving species composition of microbial communities associated with coastal crabs. First, I investigated colonization from the water column and filtering associated with the striped shore crab (Pachygrapsus transversus). Using a field experiment I asked whether species pool size interacts with habitat filters to structure microbial communities. Results indicated that the carapace was a stronger filters than the surrounding surfaces and are thus less influenced by changes in the regional species pool. Next, I examined whether the temporal scale of convergence and divergence in community composition was habitat dependent. In a mesocosm experiment, gut communities converged in community composition while carapace communities converged at day 8 and diverged thereafter. These findings indicate that the gut represents a strong filter when compared to carapace habitats, whose communities were driven by species interactions during biofilm formation. Second, I investigated whether multiple colonist pools can influence microbial communities. In a field study, I found that surface and burrow sediment colonize fiddler crab gut and carapace communities To better understand the interaction between multiple colonist pools and habitat filters, I used a mesocosm experiment where I manipulated sediment bacteria. While carapace communities were influenced by burrow sediment, removing bacteria from the surface also impacted carapace microbial composition. In contrast, gut communities responded weakly to colonist pool manipulations suggesting strong filtering. These findings highlight complex interactions between local communities and colonist pools. Importantly, identifying when multiple colonist pools or habitat filters drives community composition should lead to a more predictive community ecology.Ecology, Evolution and Behavio