The year 2018 was a hard time for romaine lettuce. Two massive, multistate E. coli outbreaks associated with leafy greens occurred (even extending into Canada), sickening hundreds of people and killing at least five. But as the outbreaks wound down, the culprit appeared even more difficult to control than leafy greens: It was a matter of contaminated agricultural water. And while the puzzle of making agricultural water safe has challenged food safety experts for a long time, the 2018 outbreaks introduced the question to the general public, many of whom now joined in asking, “How do we ensure the water used to irrigate fresh fruit and vegetables is safe?”
“Both the spring and fall outbreaks were devastating to the leafy greens industries,” says Jennifer McEntire, PhD, vice president of food safety and technology for United Fresh Produce Association in Washington, D.C. “In both cases, it appears that there is an environmental source of the pathogen. This is a new type of challenge in that it’s not about preventing an occurrence at a single point in time—it requires a revaluation of the system. Agricultural water has always been recognized as a risk and has always been managed.”
Luke LaBorde, PhD, professor of food science at Penn State’s College of Agricultural Sciences, concurs, underlining that the safety of agricultural water has always been notoriously hard to control.
“Especially,” he tells Food Quality & Safety, “if you’re using surface water from streams and irrigation ditches and things like that—it can change very suddenly. It can take some contamination upstream, what they call ‘point contamination,’ and spread it over wide distances, sometimes unexpectedly. In the case of the romaine outbreaks in Yuma, Arizona, it’s thought that there were irrigation ditches used for their produce farms that ran adjacent to animal-holding facilities. Now, we’ve known for a long time that that’s not a good idea. But still it happens.”
This contradiction is at the crux of agricultural water safety, Dr. LaBorde says. Though many farms know what approach they should be taking, they can be limited in their ability to implement them by financial concerns and other challenges like a lack of control over how neighboring farms handle waste.
“It’s tough, because people are slow to change their whole operation,” he says. “There are all sorts of things people can do, like keeping the animal facilities farther away and watching for drainage and run-off, but if it’s not your property, it’s kind of hard to control what other people do.”
Yet Rebecca Ozeran and Brooke Latack, livestock advisors at University of California Cooperative Extension, note that industry has hardly been inactive on the issue of agricultural water safety—particularly not the leafy greens industry.
“The food safety community has long been aware of the risks of water contamination, and water quality has not been ignored in food safety practices,” Ozeran and Latack jointly explained in an email to Food Quality & Safety. “Most leafy greens growers in the U.S. (producing about 95 percent of our lettuce) are already certified by the Leafy Greens Marketing Agreement, which […] regularly audits food safety practices at all certified operations. But there are inherent risks of contamination from eating any food produced in an open system. It is not feasible to grow all our crops in sterile, enclosed environments. Although operations can and do follow many practices to reduce the risks of food contamination and foodborne illness, the risks will never completely disappear. Even the best possible regulations and management will never achieve permanent zero percent contamination.”
Sampling Falls Short
The overarching problem, says Dr. LaBorde, is that regardless of regulation, some producers will inevitably end up facing situations they couldn’t have predicted. Existing regulations demand sampling and provide metrics for microorganisms, particularly (the largely harmless) generic E. coli, a good indicator of fecal contamination that can signal pathogenic E. coli or Salmonella. Even when sampling water according to regulations, a facility may miss pathogens.
“Taking some samples periodically is just not going to catch unexpected situations,” Dr. LaBorde says. “You can’t ask people to do more unless there are some assurances that what they are actually doing is going to have a public benefit. This E. coli sampling, for instance, sometimes it works, sometimes it predicts, sometimes it doesn’t. But all the studies that have come out say it’s a very poor predictor of human pathogens in water.”
Phyllis Posy, vice president, strategic services and regulatory affairs, for UV-water-treatment firm Atlantium Technologies, agrees. “Taking a sample is just that—an indication of that time and place,” she says. “How many folks are going to sample right after the rain stops, or when the temperature is coldest or warmest? We don’t catch the extremes and that may be when the contamination is measurable.”
Compounding the problem, says Dr. LaBorde, is the likelihood that sampling will miss contaminated silt in a reservoir or ditch. In the case of the Yuma E. coli outbreak, he recalls, “Some people say it might have been silt at the bottom of a ditch that was stirred up. Bacteria can attach to silt, so you may not have picked them up in the water itself. But if that silt or sediment is upset by mixing and flowing, then it can get into the irrigation system and onto the produce.”
Ozeran and Latack also note the difference between surface-water quality and the presence of pathogens in water sediment. While changes to weather can change the constitution of surface water, “sediment that collects at the bottom of a reservoir, canal, or stream tends to have higher concentrations of bacteria than the water itself. With a rainstorm causing hillside runoff or flooding, sediment and the bacteria living there can become suspended in the water. Turbid water after a storm therefore may have a higher concentration of bacteria than it had before the storm.”
While Ozeran and Latack encourage sampling as frequently as possible, and recording general site conditions while sampling (such as whether a flock of birds has recently been in the reservoir) in order to be informed about conditions affecting water quality, they acknowledge that available technology doesn’t allow constant monitoring, making constant measures for the moment impossible.
United Fresh’s Dr. McEntire puts it even more bluntly: “The bottom line is that there is no good testing protocol.”
Drinking Water Is Not Irrigation Water
Dr. LaBorde says a foundational problem with discussions about water safety is that standards applied to drinking water are not applied to irrigation water, with good reason. Initial approaches to E. coli were based on incidences of sudden contamination such as a baby’s diaper in a small lake, which is difficult to apply to irrigation.
“Transferring [an approach grounded in public health epidemiology] over to water that gets on a plant, that’s different to your exposure when you’re swimming and your mouth is open and you’re gulping water,” Dr. LaBorde adds. He notes that research on this subject is ongoing, with researchers trying to improve the means of quickly identifying fecal contamination. “But it’s very frustrating for everybody because the systems just aren’t that good right now, and while there is a lot of research to improve that, it takes time. I don’t think there is a good answer right now. The thinking has been, ‘Well this is good enough for now.’ We’ve got to do something.”
Meanwhile, says Posy, it’s unwise to attempt to apply EPA drinking water standards to irrigation due to volume alone. She stresses that EPA has zero responsibility to oversee agricultural water, since its domain is limited to drinking water.
“If you use [drinking water] for something else—say a company uses it as an ingredient or a processing aid—you can’t rely on EPA drinking water rules since the risk assessment on which those rules are based is for a household use of drinking water at less than one illness for 10,000 exposures,” says Posy. “But a household won’t have 10,000 exposures for about two to three years; a food plant will have 10,000 exposures the first day of the month!”
Is Treatment an Option?
Approaches to treating water differ, but for most the principal sticking point for irrigation water treatment is the price. Daniel Snow, PhD, lab director of the University of Nebraska’s Water Sciences Laboratory, argues, “I think monitoring and understanding impacts to irrigation water quality is clearly important, but treatment will be extremely difficult and prohibitively expensive in comparison to monitoring and treating drinking water.”
The difference, says Dr. LaBorde, is the massive volume of water used for irrigation—much of it turbid and full of organic matter.
“It’s not like you can put some chlorine in a bucket of water,” Dr. LaBorde says. “This is thousands and thousands of gallons. But some people have done that: They’ve had systems that run irrigation water through large pellets of chlorine, and it comes out treated. But chlorine isn’t a very good sanitizer for high turbidity water.”
The other question, he notes, is whether treatment is actually practical. Trying to imagine how much sanitizing agent one would need to treat the water for a major leafy-greens farm, Dr. LaBorde comments it simply costs too much, though such an approach might be workable in smaller farms that would not have to invest as much as a major operator to treat the volume of their irrigation water.
“Some people are actually shifting away from surface water and going to well water,” Dr. LaBorde says. “In theory it’s quite a bit less risky because if it’s a good well and it’s maintained, it’s not exposed to the surface and unexpected events and flooding.” Yet wells—which demand access to a potable aquifer—are not always available, Dr. LaBorde notes, while “out in California, they have had a drought. Those wells are getting low, and there’s only so much water.”
Posy, however, believes there is a role for treatment to play in agricultural water safety. “Since chemicals can have disinfection by-products and require serious control to make sure they do not have impact on the produce itself, I’m a fan of UV disinfection,” she says. “Even if you overdose the water, there’s no negative impact downstream. Watering cycles are not 24/7—we need to get smarter about how to employ treatments that are well controlled and portable as well as remotely controlled. There are systems available that treat water by measuring the real-time quality of the water and using only the dose needed, when needed. Integrated reporting means that the machine tracks key parameters and tells you when you need to pay attention. The key is to use validated systems, not just ‘check-the-box’ systems.”
Best Pathogen Practices
There are simply no easy solutions to the questions surrounding irrigation water safety, says the University of Nebraska’s Dr. Snow.
“At the very least, we might consider developing programs to evaluate irrigation water safety in areas with intensive food crop production,” he says, adding, “Regulation is seldom an efficient means for solving complex problems. If regulations governing irrigation water quality are being considered, there should be sound science supporting the type of monitoring and controls to be implemented.”
This is what concerns Dr. LaBorde as well. Above all, he says, we simply don’t know very much about irrigation safety, and many of the practices we’ve attempted to put into place for reducing pathogens are very expensive but may not be proven to reduce contamination risks. For smaller operators, shelling out for testing procedures they’re not even sure will keep their produce safe is a source of resentment.
Dr. LaBorde says, “It always comes back to: ‘Are you really sure that it’s the best way to be certain the water is safe for its purpose?’”
We may not have very many answers right now, but the shortest route to better practices is discussion, says Posy. “There is a real disconnect between the food safety community and the water safety community, both on the government and food processor/supplier/grower sides,” she says. “There’s even less dialogue on the discharge/recharge side. We need more dialogue and forums.”
For Ozeran, getting irrigation safety right comes down to developing high-level policy that balances scientific evidence, industry capacity, and pre-existing practices in the industries in question.
“For instance, if an industry is already implementing the best available practices but still experiencing food contamination,” she says, “then we need to support additional research to identify better ways to reduce contamination risks. If new policies cannot be implemented, or lack scientific support, then those policies will not improve food safety.”
The Food Safety Modernization Act carries with it a set of regulations pertaining to produce and agricultural water, but for Dr. McEntire it’s important to remember those regulations reflect minimal standards for a marketplace that has internal requirements generally exceeding those demanded by regulation, and often more frequently inspected.
“Most growers are subject to one or more annual food safety audits that include requirements around agricultural water,” says Dr. McEntire. “In California and Arizona, most producers are part of the Leafy Greens Marketing Agreements. Those water testing requirements go above and beyond the requirements in the Produce Safety Rule.”
Dr. LaBorde considers various factors that come into play, such as the increasing public appetite for raw produce, as well as the practice of irrigating with overhead sprayers, which douse produce in water.
“There are lower risk types of application methods such as drip or burrow irrigation,” he says, “which can be designed such that it wouldn’t contact the crop, reducing a lot of risk. That’s possible in very large farms. A lot of people who use drip irrigation here in Pennsylvania are small- or medium-size farms.”
Dr. LaBorde is also paying attention to studies on time-before-harvest, which aim to determine whether leaving produce in the sun will kill off enough pathogens to get produce down to safe levels. “That’s not always practical either because sometimes you have to water the crop just before harvest.”
He’s open to any new ideas about irrigation safety, but that openness is tempered by a clear understanding of how many factors come into play in trying to keep agricultural water safe. Dr. LaBorde—and many others—are as quick to dismiss silver-bullet thinking with real-world challenges.
The problem with irrigation safety, Dr. LaBorde concludes, is, “There aren’t a lot of easy answers.”