E. coli (Escherichia coli) belong to a group of bacteria sourced to the intestinal tract of vertebrates, including human beings. Much of the mass contained within a stool is the living and dead bodies of these bacteria, and billions upon billions of them are necessary to orchestrate one movement.
There are many strains of E. coli and most do not cause disease. At least six strains, however, are known to be pathogenic (i.e. disease-causing) with the most common of these producing a toxin called “Shiga” which can damage your intestinal lining and cause diarrhea. This strain is commonly known as E. coli 0157:H7, and has also caused kidney failure and death in people that ingest it.
Chlorine easily kills E. coli and thus illness resulting from its presence in treated drinking water is rare. Public (municipal) water supplies regularly monitor for it, precisely because it is so abundant in fecal matter. It’s a good indicator that a drinking water system may have been compromised by an intrusion of untreated sewage, usually in the piping that distributes water throughout a community.
Because the number of species of microorganisms present in a lake or river is vast, it’s impossible to test for them all and so we also use E. coli as an indicator for safe water recreation. While E. coli themselves may not be pathogenic (although indeed they may be), their presence indicates the possible co-presence of other fecal organisms that are likely to be pathogens for most people. These include things like Camphylobacter, Cryptosporidium, and Salmonella, all of which can be found in animal and human waste. These organisms cause more than 3 million annual illnesses in the U.S.
Determining a “safe” or “protective” level of E. coli bacteria in a lake or river is complicated and not without controversy. EPA has settled on a value of 235 colony forming units (CFU)* of E. coli per 100 ml of water. E. coli levels can vary dramatically over short periods of time, and so regulatory agencies can evaluate the geometric mean of multiple samples taken over 30 days; when the geo mean is used, the protective threshold for recreation is 126 CFU/100 ml.
What does EPA mean by “protective of human health” when considering E. coli and contact recreation in a lake or stream, where immersion and ingestion of the water is likely? In this case, a threshold of 235 CFU/100ml (or a geometric mean of 126) would be expected to produce illness in no more than 36 people in 1000 (i.e. 3.6%). So meeting the standard does not equate to zero risk.
Many Iowa lakes and streams harbor unacceptable levels of E. coli. Our state has 813 impairments of water quality on 622 lakes, stream stretches, and wetlands (some water bodies have more than one impairment). Of the 813 impairments, about half are caused by elevated E. coli levels, implying those waters are not suitable for body contact recreation sometime during the year. Other types of impairments include fish kills, low dissolved oxygen, and toxic chemicals.
Recently, I took a look at some E. coli data for the two major streams that meet in downtown Des Moines, the Raccoon and Des Moines Rivers. Water skiing is popular on the Des Moines River above the Center Street dam; the Scott Street bridge and dam just below the confluence of the two rivers is the most fished place in Iowa; and the Raccoon River flows through Des Moines Water Works Park (largest municipal park in the U.S.) as well as Gray’s Lake Park, Raccoon River Park, Walnut Woods State Park, and Browns Woods County Park, the last three situated in West Des Moines. You would be hard-pressed to find any city in the U.S. with a greater potential for river recreation than Des Moines and surrounding area.
But the water.
Using samples collected at the Des Moines Water Works intakes from 2010 up until the present day, I determined that the Raccoon River consistently meets the 235 CFU/100 ml threshold only during the cold weather months of November through March, when water recreation (at least for me) is limited to pulling fish up through holes in the ice. In the recreational season of April-October, more than half the days (50.2%) see E. coli levels exceed the protective threshold for safe recreation in the Raccoon River.
The situation is modestly better for the Des Moines River, with 22% of the days in the recreational season (Apr-Oct) exceeding the protective threshold. It’s not a mystery why E. coli levels are lower in the Des Moines River—detention of water 6 miles upstream in Saylorville Reservoir exposes the bacteria to sunlight and other stressors, causing them to perish before they exit the reservoir.
It’s no secret that we have a lot of livestock in Iowa and clearly some of these bacteria come from animal manure and the soils it is applied to. But E. coli do live in the gut of humans, pets and wildlife, so we can’t exclude the possibility these other sources are impairing the metro rivers. To try to get a handle on this, I compared the Raccoon and Des Moines River data with what I thought might be a similar situation—the Mississippi River in downtown Minneapolis. It’s bigger than both the Raccoon and Des Moines, but not by a lot; Minneapolis Miss is about 1.4 times larger than the combined flow of the Raccoon and Des Moines Rivers below Scott Street. It’s reasonable to assume that the human, pet, and wildlife contributions in Minneapolis are no greater than they are in Des Moines.
Using data from the Minnesota Pollution Control Agency, I found that the Mississippi is far cleaner, at least when it comes to fecal matter, than the Raccoon and Des Moines Rivers. Looking at the April-October recreational period (no winter data was available), the Mississippi exceeds the recreational threshold only 9.2% of the time in downtown Minneapolis, compared to 22% of the time for the Des Moines River and 50.2% for the Raccoon River (2010-2019). What is especially striking are the “worst” or highest samples: 130,000 CFU/100 ml (Raccoon), 17,200 (Des Moines), but only 2420 for the Mississippi.
There is some controversy in Iowa about whether or not E. coli is the best indicator for recreational suitability, and some would like to change the standard or the indicator or both. The impaired waters list is a lingering fart in the room that to some smells worse than liquid hog manure on a November night. My guess is that if we picked something other than E. coli to measure, these folks would stop grunting, at least for a while, because they think water quality would improve magically overnight—on paper. But the motives behind this idea are as transparent as gas station toilet tissue, in my estimation. Besides, if you start looking at other indicators, chances are you’re going to find them unless they’re unique to camels or elephants or some other exotic creature. This is definitely a “be careful what you wish for” scenario.
It’s been more than a year ago now that I wrote “Iowa’s Real Population”, which featured the map below. Considering all our livestock animals and the waste they excrete, we effectively generate the fecal waste equivalent of Chicago, Tokyo, Rome, Paris, Oklahoma, and Alabama upstream of the city of Des Moines. Frankly, it would be a surprise if E. coli weren’t lousy in the water flowing through Des Moines. And believe me when I tell you this: I’m not the first person doing science or policy to have thought this through.