Abstract
Hydrothermal vents, such as those at Lō'ihi Seamount and the Mariana Arc and back-arc, release iron required to support life from the Earth's crust. In these ecosystems, bacteria and archaea can oxidize the released iron and therefore play an important role in the biogeochemical cycles of essential nutrients. These organisms often form microbial mats, and the primary producers in these communities can support diverse higher trophic levels. One such class of bacteria are the Zetaproteobacteria. This class of bacteria oxidize iron and commonly produce extracellular iron oxyhydroxide matrices that provide architecture to the microbial mats, so they are considered foundational members of the community and ecosystem engineers. Zetaproteobacteria are responsible for the majority of iron-oxidation in circumneutral, marine, low-oxygen environments. To study the composition of these communities, microbial mats were collected using a biomat sampler, which allows for fine-scale collection of microbial mats. DNA was then extracted and amplified for analysis of the SSU rRNA gene. After quality control and filtering, the SSU rRNA genes from Mariana Arc and Lō'ihi Seamount microbial mat communities were compared pairwise to determine which site exhibits a greater microbial diversity and how much community overlap exists between the two sites. In-depth analysis was performed with the rule-based microbial network (RMN) algorithm, which identified a possible competitive relationship across oligotypes of a cosmopolitan Zetaproteobacteria operational taxonomic unit (OTU). This result demonstrated the ecological relevance of oligotypes, or fine-scale OTU variants. The oligotype distributions of the cosmopolitan ZetaOTUs varied greatly across the Pacific Ocean. The competitive relationship between dominant oligotypes at Lō'ihi Seamount and the Mariana Arc and back-arc may be driving their differential distributions across the two regions and may result in species divergence within a cosmopolitan ZetaOTU. This implementation of the RMN algorithm can both predict directional relationships within a community and provide insight to the level at which evolution is occurring across ecosystems.