Environmental and Spatial Influences on Biogeography and Community Structure of Benthic Diatoms

Several theoretical and practical reasons suggest that benthic microalgae could be useful bioindicators. For instance, an ideal indicator species or community would be associated with a given habitat due to local physical conditions or biotic interactions (i.e., ‘environmental filtering’), not due to dispersal limitation. Due to their small size, immense abundances, and reliance on passive dispersal, the popular notion about micro-organisms is that ‘Everything is everywhere, but, the environment selects’ (Baas-Becking 1934). Although much recent research concerning planktonic bacteria and dispersal limitation has been conducted, very little in this regard is known about microeukaryotes, especially benthic microbes. The purpose of our study was to identify and compare spatial and environmental influences on benthic diatom community structure and biogeography. In summer 2015, sediment was sampled at various spatial scales from four barrier island beaches in South Carolina, USA, and high-throughput (Ion Torrent) DNA sequencing was used to characterize diatom assemblages. ANOSIM and principal coordinates analysis revealed that communities were statistically distinct on the four islands. Community dissimilarity was compared to both spatial distance and environmental differences to determine potential influences of these variables on community structure. We found that geographic distance had the strongest correlation with community similarity, with and without one anomalous location, while differences in temperature (air, water, and sediment), nutrients, organic matter, and turbidity also had significant but weaker relationships with community structure. Surprisingly, air temperature, which changes on very short time scales, appeared to be the environmental factor most strongly related to diatom species composition, potentially implicating some unmeasured variable (e.g., cloud cover). However, we also found that temperature and geographic distance were strongly correlated. Future research will expand the spatial scope of this preliminary study and employ techniques (partial Mantel tests) to control for co-variation among variables.