Bacterial Community Shift and Coexisting/Coexcluding Patterns Revealed by Network Analysis in a Uranium-Contaminated Site after Bioreduction Followed by Reoxidation
Bing Li,Wei-Min Wu,David B Watson,Erick Cardenas,Yuanqing Chao,D H Phillips,Tonia Mehlhorn,Kenneth Lowe,Shelly D Kelly,Pengsong Li,Huchun Tao,James M Tiedje,Craig S Criddle,Tong Zhang
DOI: https://doi.org/10.1128/AEM.02885-17
2018-04-16
Abstract:A site in Oak Ridge, TN, USA, has sediments that contain >3% iron oxides and is contaminated with uranium (U). The U(VI) was bioreduced to U(IV) and immobilized in situ through intermittent injections of ethanol. It then was allowed to reoxidize via the invasion of low-pH (3.6 to 4.0), high-nitrate (up to 200 mM) groundwater back into the reduced zone for 1,383 days. To examine the biogeochemical response, high-throughput sequencing and network analysis were applied to characterize bacterial population shifts, as well as cooccurrence and coexclusion patterns among microbial communities. A paired t test indicated no significant changes of α-diversity for the bioactive wells. However, both nonmetric multidimensional scaling and analysis of similarity confirmed a significant distinction in the overall composition of the bacterial communities between the bioreduced and the reoxidized sediments. The top 20 major genera accounted for >70% of the cumulative contribution to the dissimilarity in the bacterial communities before and after the groundwater invasion. Castellaniella had the largest dissimilarity contribution (17.7%). For the bioactive wells, the abundance of the U(VI)-reducing genera Geothrix, Desulfovibrio, Ferribacterium, and Geobacter decreased significantly, whereas the denitrifying Acidovorax abundance increased significantly after groundwater invasion. Additionally, seven genera, i.e., Castellaniella, Ignavibacterium, Simplicispira, Rhizomicrobium, Acidobacteria Gp1, Acidobacteria Gp14, and Acidobacteria Gp23, were significant indicators of bioactive wells in the reoxidation stage. Canonical correspondence analysis indicated that nitrate, manganese, and pH affected mostly the U(VI)-reducing genera and indicator genera. Cooccurrence patterns among microbial taxa suggested the presence of taxa sharing similar ecological niches or mutualism/commensalism/synergism interactions.IMPORTANCE High-throughput sequencing technology in combination with a network analysis approach were used to investigate the stabilization of uranium and the corresponding dynamics of bacterial communities under field conditions with regard to the heterogeneity and complexity of the subsurface over the long term. The study also examined diversity and microbial community composition shift, the common genera, and indicator genera before and after long-term contaminated-groundwater invasion and the relationship between the target functional community structure and environmental factors. Additionally, deciphering cooccurrence and coexclusion patterns among microbial taxa and environmental parameters could help predict potential biotic interactions (cooperation/competition), shared physiologies, or habitat affinities, thus, improving our understanding of ecological niches occupied by certain specific species. These findings offer new insights into compositions of and associations among bacterial communities and serve as a foundation for future bioreduction implementation and monitoring efforts applied to uranium-contaminated sites.