Community-level microbial responses to recharge and fecal contamination in a karst aquifer

Authors: Heffron, Joseph; Cook, Rachel; FIRNSTAHL, AARON - Us Geological Survey (USGS); STOKDYK, JOEL - Us Geological Survey (USGS); Burch, Tucker

Submitted to: American Society for Microbiology Annual Meeting
Publication Type: Abstract Only
Publication Acceptance Date: 2/27/2024
Publication Date: 6/15/2024
Citation: Heffron, J.A., Cook, R.M., Firnstahl, A.D., Stokdyk, J.P., Burch, T.R. 2024. Community-level microbial responses to recharge and fecal contamination in a karst aquifer. American Society for Microbiology Annual Meeting. June 13-17,2024.

Interpretive Summary:

Technical Abstract: Despite the importance of subsurface microbial communities to groundwater quality, little is known about how land use and other surface influences impact groundwater microbes. Our study aimed to relate groundwater microbial communities to fecal contamination from domestic wastewater and livestock manure. Water samples (n=138) were collected from private wells in a rural northeast Wisconsin county. Samples were analyzed using 16S (V4) Illumina sequencing. Amplicon sequence variants (ASVs) were generated from the demultiplexed reads and decontaminated using pre-amplified sample dsDNA concentrations. Groundwater communities from across the county were highly homogenous. Core taxa, defined as microbial orders appearing in over 95% of samples, included representatives of Patescibacteria, Proteobacteria, Firmicutes, Verrucomicrobiota, Crenarchaeota, Dependentiae, and Nanoarchaeota. The median sample relative abundance of these core taxa was 94%, though some samples included as little as 19% core taxa. To relate microbial community composition to environmental and land use factors, sample communities were clustered by Bray-Curtis distances (k-means, n = 2). Cluster assignment was evaluated relative to land use, geology, groundwater recharge, pathogen and antibiotic resistance gene (ARG) occurrence, and microbial source tracking (MST) markers for human and bovine fecal contamination. Random forest classification was used to select the most important attributes for predicting microbial clusters. This subset of attributes was then tested for significance using logistic regression. Multiple attributes related to water infiltration from the surface (e.g., snowmelt, estimated recharge, precipitation, and presence of sinkholes) were found to be successful predictors of clustering. No attributes related to land use, pathogens, ARGs, or MST markers were significantly associated with microbial clusters. Well water represents the planktonic subsurface community, which may be more susceptible than groundwater biofilms to dramatic turnover during recharge and infiltration. In addition, the wells sampled in this study had all previously exhibited nitrate and/or bacterial (total coliform) contamination. Groundwater communities established under chronic fecal inputs may be less sensitive to acute fecal events. The long-term impacts of fecal contamination on both attached and planktonic communities require further study.