Bacteria play a key role in mediating ecosystem level response of coastal wetlands to a rising sea; however, the response of bacterial communities to restoration efforts is rarely explicitly considered, due in part to the 'field of dreams' assumption that bacterial community structure and function will return with plant and hydrologic restoration. In this project, we address two fundamental questions: i) What is the effect of plant species in determining the strength and occurrence of microbial interactions along a flooding and salinity gradient? and ii) How does a restoration burn interact with vegetation and edaphic factors to affect microbial interactions? To address the first question, we compared microbial association networks across 4 vegetation zones (pine forest, fresh marsh, brackish marsh, and salt marsh) of 4 replicate marsh-pine island complexes at the Grand Bay National Estuarine Research Reserve to evaluate the effects of plant species on microbial community structure and microbe-microbe interactions. Two of these islands were then treated with a management burn as part of a plant migration restoration experiment, and the effects of the burn on microbial species interactions and overall microbial association networks were examined.
Results/Conclusions
Prior to experimental treatments, we saw marked taxonomic differences in bacterial community composition, with microbial community composition clearly delineated at each vegetation zone. Microbial association network analysis revealed greater overall co-occurrence of taxa, with stronger correlations, at the lower end of the gradient. While this suggests that increased oxygen stress may lead to greater cooperation among organisms and greater metabolic partitioning of resources across taxa, it may also be a reflection of the strength of habitat filtering along this flooding and salinity gradient. However, these differences were overwhelmed by the effects of a controlled burn, which disrupted existing microbial associations, and led to no detectable differences in community composition across the gradient. Post-burn microbial associations were weak, suggesting a rapid response of pioneer species with little metabolic partitioning. Ongoing work will address how these interactions may influence bacterial community assembly and dynamics, as well as tease apart effects of vegetation from edaphic factors (e.g. flooding and salinity) in shaping wetland soil bacterial communities after a fire disturbance.