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

Research output: Contribution to journalArticlepeer-review

40 Scopus citations

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.

Original languageEnglish
Article numbere02885-17
JournalApplied and Environmental Microbiology
Volume84
Issue number9
DOIs
StatePublished - May 1 2018

Keywords

  • Bacterial community
  • Groundwater
  • High-throughput sequencing
  • Network analysis
  • Nitrate
  • Uranium

Fingerprint

Dive into the research topics of 'Bacterial community shift and coexisting/coexcluding patterns revealed by network analysis in a uranium-contaminated site after bioreduction followed by reoxidation'. Together they form a unique fingerprint.

Cite this