Increased denitrification rates associated with shifts in prokaryotic community composition caused by varying hydrologic connectivity

While modern developments in agriculture have allowed for increases in crop yields and rapid human population growth, they have also drastically altered biogeochemical cycles, including the biotransformation of nitrogen. Denitrification is a critical process performed by bacteria and fungi that removes nitrate in surface waters, thereby serving as a potential natural remediation strategy. We previously reported that constant inundation resulted in a coupling of denitrification gene abundances with denitrification rates in sediments, but these relationships were not maintained in periodically-inundated or non-inundated environments. In this study, we utilized Illumina next-generation sequencing to further evaluate how the microbial community responds to these hydrologic regimes and how this community is related to denitrification rates at three sites along a creek in an agricultural watershed over 2 years. The hydrologic connectivity of the sampling location had a significantly greater influence on the denitrification rate (P = 0.010), denitrification gene abundances (P < 0.001), and the prokaryotic community (P < 0.001), than did other spatiotemporal factors (e.g., creek sample site or sample month) within the same year. However, annual variability among denitrification rates was also observed (P < 0.001). Furthermore, the denitrification rate was significantly positively correlated with water nitrate concentration (Spearman's ρ = 0.56, P < 0.0001), denitrification gene abundances (ρ = 0.23-0.47, P = 0.006), and the abundances of members of the families Burkholderiaceae, Anaerolinaceae, Microbacteriaceae, Acidimicrobineae incertae sedis, Cytophagaceae, and Hyphomicrobiaceae (ρ = 0.17-0.25, P = 0.041). Prokaryotic community composition accounted for the least amount of variation in denitrification rates (22%), while the collective influence of spatiotemporal factors and gene abundances accounted for 37%, with 40% of the variation related to interactions among all parameters. Results of this study suggest that the hydrologic connectivity at each location had a greater effect on the prokaryotic community than did spatiotemporal differences, where inundation is associated with shifts favoring increased denitrification potential. We further establish that while complex interactions among the prokaryotic community influence denitrification, the link between hydrologic connectivity, microbial community composition, and genetic potential for biogeochemical cycling is a promising avenue to explore hydrologic remediation strategies such as periodic flooding.