Assembly of skin microbiomes is more neutral than gut microbiomes in multiple animal species

Abstract

The gut and external tissues of most animals are colonized by communities of microorganisms that can influence the health, development, and fitness of the host. The composition of these communities can vary greatly between individuals within a host species, and both selective factors (e.g. host immune response) and neutral processes (e.g. random loss of microbial cells) have been shown to contribute to this variation. While it is known that microbiome composition differs between tissues within an individual host, less is known about the ecological processes that underline these differences. To address this, we investigated whether the contribution of neutral ecological processes to microbiome assembly differs between external (skin and scale) and internal (gut) host tissues for a diverse panel of animal hosts. To do this, we fit a neutral ecological model to microbial communities from external and internal tissues across a variety of animal hosts. Strikingly, we discovered that the neutral model was equally or better fit to skin or scale microbial communities across all hosts, suggesting that neutral processes play a larger role in the assembly of skin or scale microbiomes compared to gut microbiomes. Furthermore, we observed that this trend is robust to different definitions of the metacommunity (i.e. the microbial taxa available to colonize a host). Finally, we leveraged a simulation framework to compare the model fits of empirical vs simulated microbial communities. We found that neutral model fits to empirical communities can differ from simulated communities, emphasizing the importance of temporal sampling in profiling animal microbiomes.