When brown water starts coming out of kitchen faucets and filling bathtubs, it is concerning for all involved. One of the first reactions is to point fingers. Local residents point to area dairy farmers and vice versa.

Lee karen
Managing Editor / Progressive Dairy

Unfortunately, brown water events happen often enough in Kewaunee County, Wisconsin, that Mark Borchardt, with USDA-ARS, and a team of researchers decided to conduct a large study to find the source of contamination in the groundwater.

At the Midwest Manure Summit in Green Bay, Wisconsin, in February, Borchardt explained the Silurian dolomite aquifer in the region is particularly vulnerable. There are many horizontal and vertical fractures that act as pipes and move the water very, very quickly. In addition, there are many parts of the aquifer where the soils are thin. Sometimes there is no soil at all. Other times there may only be a few inches to several feet.

“What this means is: When water hits the surface, like rainfall or snowmelt, things move to the groundwater amazingly fast,” Borchardt said.

Not to mention this is a very productive aquifer that goes beyond northeast Wisconsin around the entire Great Lakes region and stretching as far as Guelph, Ontario, Canada.


Since the flow of water within the aquifer occurs primarily along the bedding plane fractures, there is little to no attenuation of contaminants within the aquifer. Borchardt said the flow rates could vary from 10 to hundreds of feet per day as opposed to a porous media aquifer, like sandstone, where the rates might only be a millimeter per day.

“There’s very little surface runoff,” he said. “Water goes into the groundwater quickly, and recharge is exceedingly rapid. It can carry surface contaminants to the water table.”

Borchardt said he thinks recharge events (rainfall or snowmelt) are the primary drivers for contamination, as brown water events tend to happen during snowmelt.

Disease outbreaks of enterohemorrhagic E. coli and campylobacter have also been associated with the brown water events.

With nearly 100,000 cows, heifers and calves in the county, dairy farms, particularly CAFOs, are being targeted as the source. Despite there being a high nutrient management plan compliance rate and rules that CAFOs can’t spread manure on frozen or snow-covered ground, this contamination continues to happen, Borchardt said.

Equipped with the technology to determine the brown water source, Borchardt decided to collect some samples as part of a training exercise. His team collected samples from 10 households in a biased study because they were households with a known history of water problems.

The test was designed to look for host-specific viruses, meaning viruses specific to cattle or specific to humans.

“For these gastrointestinal viruses, they are host-specific. So when we find one in water, we know where it comes from,” Borchardt said.

In addition to those viruses, they were looking for a number of pathogens that are not host-specific and could come from cattle or humans.

In this pilot study, they found seven out of 10 wells, or 70 percent, were positive for fecal contamination. “That’s a really high contamination rate,” Borchardt said, noting that in a study he published in 2003 from 50 household wells in Wisconsin, there was only an 8 percent hit rate.

Of the seven contaminated wells, the results were 50-50. Three wells had human-specific viruses, three wells had bovine-specific viruses, and one had both.

Six wells were negative for total coliform or E. coli indicators, while four wells were positive for salmonella, and one well was positive for Campylobacter jejuni.

“When you think about drinking salmonella, that’s possibly life-threatening,” Borchardt said.

These results prompted a much larger study designed to gain an understanding of how much contamination there was in private wells in the county, where it is coming from, when it is occurring and the risk factors that lead to contamination.

After categorizing the wells based on depth to bedrock and sending out letters to randomly selected homeowners within each group, they were able to collect samples from 327 participants during a recharge period in November 2015.

Thirty-four percent of the wells sampled were found positive for total coliform, E. coli or high nitrates. With a weighted analysis by depth to bedrock, Borchardt said they estimate 26 percent of the wells in the county were positive.

There was a significant correlation with depth to bedrock, where wells at less than 5 feet had a 50 percent chance of being contaminated, 6 to 20 feet were at 42 percent, and greater than 20 feet depth to bedrock at 23 percent.

Another sample set during a non-recharge period in July 2016 from more than 400 households resulted in 31 percent of the wells contaminated with one of the indicators and a little bit less of an effect on depth to bedrock. Borchardt noted at this time it was only provisional data, and the statistical analysis had not been done yet.

To answer the question on where the contamination is coming from, three rounds of sampling were conducted in April, August and November last year. Out of 82 wells, 18 contained human wastewater, 26 had bovine wastewater, and 29 had micro-organisms that were not host-specific.

“Overall, 52 out of 82 [wells] had something in them they shouldn’t have,” Borchardt said.

In general, he added, the concentrations from bovine are quite a bit higher than human.

As he began to interpret the data, Borchardt noted Kewaunee County is primarily rural with a lot of septic systems, approximately 5,000 of them, and all septic systems are designed to release effluent to the environment.

“It releases it to the subsurface, so we shouldn’t be surprised we’re seeing human wastewater in our groundwater,” he said.

In the researcher’s conceptual model, human septic material is continuously being released at small amounts – but 3 to 4 feet underground so it gets a head start to the groundwater. Whereas, dairy manure is a larger source but only applied periodically at the surface. For it to get underground, it requires a rainfall or snowmelt event.

According to their data, during non-recharge times, there are more samples with human wastewater than bovine. Then, when recharge is occurring, the number of samples with bovine is much higher than human.

To get a better handle on when the pathogens are arriving in wells, they have installed auto-sampling units in two homes. “Ideally, what we’d like to do, if possible, is to find something like chloride or nitrate or conductivity, something that’s easily measured with a sensor in real time that’s associated with this pathogen movement,” he said. That way it could serve as an early warning system.

Once all of the sampling is complete, Borchardt and his team plan to overlay land-use patterns, manure application, well construction and environmental variables to identify the items most closely linked with well contamination.

Thus far, they can conclude that both humans and cows are responsible for well contamination in the county, and there are pathogens of significant concern.

“I think everyone recognizes that both parties bear some responsibility for the groundwater contamination,” Borchardt said. “What I’d like to think – maybe I’m being incredibly naïve – is that both parties can contribute to the solution.  end mark

PHOTO: Brown well water is a concern for residents of Kewaunee County, Wisconsin. Researchers have confirmed that samples contain human waste, bovine waste, or both. Photo by Karen Lee.

Karen Lee