Temperature and microbial interactions drive the deterministic assembly processes in sediments of hot springs.

Abstract Terrestrial geothermal ecosystems, as a representative of extreme environments, exhibit a variety of geochemical gradients, and their microbes are thought to be under high stress through environmental selection. However, it is still unclear how stochasticity and biotic interactions contribute to the microbial community assembly in hot springs. Here, we investigated the assembly processes and co-occurrence patterns of microbiota (i.e. bacteria and archaea) in both water and sediments sampled from fifteen hot springs in the Tengchong area, Southwestern of China, using 16S rRNA gene sequencing combined with multivariate ecological and statistical methods. These hot springs harbored more specialists than non-geothermal ecosystems, which are well-adapted to the extreme conditions, as shown by extremely high nearest-taxon index (NTI) and narrower niche width. Habitat differentiation led to the differences in microbial diversity, species-interactions, and community assembly between water and sediment communities. The sediment community showed stronger phylogenetic clustering and was primarily governed by heterogeneous selection, while undominated stochastic processes and dispersal limitation were the major assembly processes in the water community. Temperature and ferrous iron were the major factors mediating the balance of stochastic and deterministic assembly processes in sediment communities, as evidenced by how divergences in temperature and ferrous iron increased the proportion of determinism. Microbial interactions in sediments contributed to deterministic community assembly, as indicated by more complex associations and greater responsiveness to environmental change than water community. These findings uncover the ecological processes underlying microbial communities in hot springs, and provide potential insight into understanding the mechanism to maintain microbial diversity in extreme biospheres.