Microbial source tracking and land use associations for antibiotic resistance genes in private wells influenced by human and livestock fecal sources

Authors: Burch, Tucker; STOKDYK, JOEL - Us Geological Survey (USGS); FIRNSTAHL, AARON - Us Geological Survey (USGS); KIEKE, BURNEY - Marshfield Clinic Research; Cook, Rachel; Opelt, Sarah; Spencer, Susan; Durso, Lisa; Borchardt, Mark

Submitted to: Journal of Environmental Quality
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/30/2022
Publication Date: 12/8/2022
Citation: Burch, T.R., Stokdyk, J.P., Firnstahl, A.D., Kieke, B.A., Cook, R.M., Opelt, S.A., Spencer, S.K., Durso, L.M., Borchardt, M.A. 2022. Microbial source tracking and land use associations for antibiotic resistance genes in private wells influenced by human and livestock fecal sources. Journal of Environmental Quality. 52(2):270-286. https://doi.org/10.1002/jeq2.20443.
DOI: https://doi.org/10.1002/jeq2.20443

Interpretive Summary: Antimicrobial resistance is a global public health problem, contributing to more than two million cases of infection per year in the United States. It is also often characterized as a “One Health” problem because antimicrobial resistance is widely believed to be driven by antibiotic use in both human medicine and livestock production, and it is mediated by environmental factors as well. However, very few studies on antimicrobial resistance capture all three One Health sectors (i.e., human, animal, and environmental). This study examines the occurrence of antimicrobial resistance genes (ARGs) – the genetic basis of antimicrobial resistance in resistant bacteria – in groundwater from a rural region of northeast Wisconsin where contamination by human and livestock sources occurs. Human sources include discharge from residential septic systems, and livestock sources largely consist of land-applied dairy manure. ARGs were found in groundwater obtained from residential private wells, and their frequency of occurrence was related to land use, rainfall, hydrogeological conditions, and well construction. The most common land use factor identified across all ARGs investigated was the presence of residential septic systems on the landscape. However, ARGs also co-occurred with genetic markers specific to human and bovine sources, and ARG occurrence was similar between the two sources. Thus, this study found that both sources contributed to contamination of groundwater by ARGs in this setting, and neither source dominated. This study could serve as a model for investigating how generalizable this finding is in other agroecosystems with different mixtures of land use, different agricultural practices, and different environmental factors.

Technical Abstract: Antimicrobial resistance is a growing public health problem and requires a One Health approach when formulating policy and management interventions. However, few studies address its animal, human, and environmental components simultaneously. We investigated the occurrence of five antibiotic resistance genes (ARGs) and the integrase gene of class 1 integrons (intI1) in private wells drawing water from a vulnerable aquifer influenced by a combination of residential septic systems and land-applied dairy manure. Samples (n = 138) were collected across all four seasons over approximately 1 year from a randomized sample of private wells in Kewaunee County, Wisconsin. ARG and intI1 occurrence were related to microbial source-tracking (MST) markers specific to human and bovine fecal material; they were also related to 54 risk factors for contamination representing land use, rainfall, hydrogeology, and well construction. ARGs occurred in 5-40% of samples overall depending on target, and septic system density was the most frequently-identified land use risk factor for contamination across all targets. Detection frequencies for ARGs and intI1 were lowest in the absence of source-specific MST markers (1-30%), highest when co-occurring with both human and bovine markers (11-78%), and between these extremes when co-occurring with just one type of marker (4-46%). MST results indicated that both sources contribute to ARG contamination in this setting; neither source dominates. Interventions intended to reduce ARG occurrence in this environment will be most effective when considering both sources, and the approach used here could be expanded to other settings to investigate how ARG contamination in the environment associates with human versus livestock fecal sources.