Farm to Faucet? Agricultural Waste and Private Well Contamination in Kewaunee County, Wisconsin.

About 44% of the U.S. population relies on groundwater for drinking water, and more than households use private wells.1 Although a household may own its well, the residents do not control the water that recharges the aquifer they draw upon. This aquifer can become contaminated, in whole or in part, leading to a “tragedy of open access”2 —a subcategory of the classic “tragedy of the commons” defined by ecologist Garrett Hardin.3 A pair of studies in Environmental Health Perspectives4,5 examined such a scenario among private wells in Kewaunee County, Wisconsin.

The highly fractured Silurian dolomite aquifer provides water for 85% of this region’s wells, according to Kenneth Bradbury, a hydrogeologist with the Wisconsin Geological and Natural History Survey. This aquifer is vulnerable to contamination due to its rapid recharge, a function of the region’s shallow bedrock, says Bradbury, who was not involved in the research. Many homes in Kewaunee County are on septic systems, which ultimately discharge to groundwater, he adds. In addition, local dairy farmers spread cattle manure over the landscape as part of nutrient management plans6 that supply agricultural fields with fertilizer and nutrients. To settle a debate7 over whether septic systems or cattle were to blame for private well contamination, the authors of the new papers set out to systematically determine the sources, extent, and resulting health effects of tainted well water.

Proximity to manure lagoons like the ones on this Wisconsin dairy farm was associated with coliform contamination of nearby drinking water wells. The authors say the strength of the association was one of the most surprising findings from the study. Image: © iStock/BanksPhotos.

In the first study,4 the authors quantified the extent and sources of microbial and nitrate contamination in private wells in Kewaunee County. As of 2016, the county was home to 20,600 people, half of whom lived in 4,900 rural homes with wells. The researchers selected private household wells using stratified random sampling. Residents who agreed to participate collected up to two samples from their wells—323 participated during an aquifer recharge period (rainy weather and snowmelt) in November 2015, and 401 participated during a no-recharge period (dry weather) in July 2016. The residents mailed in the samples to be tested for contaminants, including Escherichia coli, total coliform bacteria, and nitrates.

Past work showed that a greater distance from the land surface to the bedrock—and thus to the aquifer—reduced the likelihood of contamination from surface sources.8 Therefore, the authors measured depth-to-bedrock at each well. Higher total coliforms and nitrates both showed strong associations with shallower depth-to-bedrock and a greater amount of nearby agricultural land use, but not with the presence of septic systems.

The team used microbial source tracking to correlate well contaminants to their source, whether septic systems or agricultural fields. Microbial source tracking uses biochemical or genetic markers to determine whether microorganisms in fecal matter originated from human or nonhuman sources, as well as to distinguish among nonhuman sources, such as cattle and pigs. More than 60% of the wells had fecal-associated contamination.

“What struck us is how strongly those risk factors related to agricultural land use were also related to private well contamination,” says Mark Borchardt, lead author on one of the studies and a U.S. Department of Agriculture (USDA) microbiologist. “The one that really threw us is the linkage between coliform contamination in wells and the distance to a manure lagoon [as opposed to a septic system]. We never thought we would see the strength of that relationship.”

Regarding nitrates, Borchardt pointed to nutrient management plans, widely employed to reduce surface runoff from agricultural operations. “Everyone thinks these nutrient management plans help mitigate pollution, and in fact, we found that the more fields around a well with [a plan in place], the greater the likelihood of that well having a nitrate level that exceeds the standard,” he says. “There’s a mindset that if you have a nutrient management plan, you must be managing your nutrients well, and there’s little risk of contamination. That turned out not to be true in our study site.”

The U.S. Environmental Protection Agency set a maximum contaminant level goal for nitrate of for public drinking water supplies.9 However, the standard does not apply to private wells, which are unregulated.

The authors of the second study5 used quantitative microbial risk assessment (QMRA) to calculate the health burden from the water contamination quantified in the first study. First, they predicted how many cases of acute gastrointestinal illness (AGI) due to contaminated private wells could be expected in Kewaunee County based on county- and national-level numbers. Then they associated their predictions with two factors: fecal source (bovine, human, or unknown) and soil depth-to-bedrock ( or less, or greater than ).

Their analysis predicted that well contamination could cause up to 301 cases of AGI per year, 230 of them connected with bovine fecal pathogens, 12 with human sources, and the remainder with unknown sources. Nearly five times as many cases were associated with a depth-to-bedrock greater than , versus shallower wells. This counterintuitive finding reflected the greater number of wells constructed to these depths, which resulted in more people being exposed.

“This is a really high-quality study made possible by the amazing data set generated by the authors in [the first] paper,” says Stanford University civil and environmental engineering professor Alexandria Boehm, who was not involved in the research. “They compared their risk estimates to those produced by others for similar regions and to the actual estimate of illnesses reported to the local health department, and the comparisons were favorable. This lends credence to the results.”

“[The scientists] have taken an exciting leap forward in the use of QMRA for contaminated private groundwater insofar as … they have quite rightly elected to focus in on one specific aquifer and community,” says Paul Hynds, an environmental epidemiologist with the Technological University of Dublin who conducted a comprehensive QMRA well contamination study in Ireland.10 “This approach, which is based on a wonderful 12-month fieldwork campaign—thus allowing them to assess the seasonality of risk—provides a framework for all future QMRAs of this nature.” Despite the lack of local or regional daily consumption values, Hynds—who was not involved in the studies—says “this is the best I’ve seen in the literature to date, and I’ll be trying to emulate it in Europe.”

Another interesting finding from the study was the positive correlation between depth-to-bedrock and total cases of AGI. “Wells drilled in depth-to-bedrock of 20 feet or more are typically thought of as less vulnerable to contamination because all that soil above a well helps to filter out some of the contaminants,” says Tucker Burch, a USDA agricultural engineer and lead author of the second paper. “Our study highlights an important qualification: Wells drilled in deeper depth-to-bedrock aren’t completely invulnerable to contamination.

Although people may propose handling contamination by building better wells, Borchardt says, “We see a very weak relationship with well construction. The biggest bang for the buck in improving the water quality is the land-use risk factors.” This, he says, has important public policy implications given that key factors include septic system density and proximity to manure storage, such as lagoons.