The U.S. Food and Drug Administration’s Proposed Produce Rule is a substantively defective product. It’s not a question of the difficulty and costs of compliance and enforcement or of farm size or a commodity’s inherent safety – although these are important issues.

The case I want to present is that FDA made seriously wrong choices on too many issues: irrigation water standards; biological inputs such as compost, radiological and chemical hazards; sprout production and fresh-cut production; epidemiological data and economic analysis. Like other hazardous and defective products, the Proposed Rule should be recalled.

Today’s column is about irrigation water. Here is the relevant section of the Proposed Rule for testing water:

“Article 112.44 (c) When agricultural water is used during growing activities for covered produce (other than sprouts) using a direct water application method you must test the quality of water in accordance with one of the appropriate analytical methods in subpart N. If you find that there is more than 235 colony forming units (CFU) (or most probable number (MPN), as appropriate) generic E. coli per 100 mL for any single sample or a rolling geometric mean (n=5) of more than 126 CFU (or MPN, as appropriate) per 100 mL of water, you must immediately discontinue use of that source of agricultural water and/or its distribution system for the uses described in this paragraph. Before you may use the water source and/or distribution system again for the uses described in this paragraph, you must either re-inspect the entire agricultural water system under your control, identify any conditions that are reasonably likely to introduce known or reasonably foreseeable hazards into or onto covered produce or food-contact surfaces, make necessary changes, and retest the water to determine if your changes were effective; or treat the water in accordance with the requirements of § 112.43.”

I hope that’s clear.

Farmers are supposed to be able to follow the Food Safety Modernization Act without being forced to hire consultants or private certifiers.

Note that there are precise and enforceable – and, as we shall see, useless – metric standards such as “126 CFU per 100 ml” (in a rolling geometric mean) of generic E. coli.

The problem is that the generic E. coli standard is completely uncorrelated with the human pathogens of concern that have caused produce outbreaks and recalls such as E. coli O157:H7 and Salmonella. Worse, it has been shown to be uncorrelated with these human pathogens in actual production of leafy greens and other produce in numerous published studies, including by researchers from USDA and UC Davis, for California production, and by other researchers across the country.

The largest data set, however, is from a California fresh-cut processor whom I interviewed as part of writing a comment on the Rule. For more than six years, they have tested about 2 million pounds of produce a week, as well as the farm inputs used to grow those 2 million pounds. They tested both for specific human pathogens, such as Salmonella and O157:H7 E. coli, and for generic E. coli, which is required for members of the California Leafy Green Handlers Marketing Agreement (CA LGMA). Their summary: the generic E. coli standard can be exceeded without human pathogens being present, and it can be met when human pathogens are actually present in high quantities. It is not a useful standard for produce safety.

Trevor Suslow, in a Pew Issue Brief (1), described the origins of this standard and accurately predicted its increased use:

“A limited, and arguably outdated, set of indicators of fecal contamination has been used by the fresh produce industry to assess the suitability of water used in preharvest crop production up to the point of harvest. Many regional GAP and CSG systems have relatively recently adopted EPA recreational water quality criteria for establishing action thresholds, in the absence of actual risk-based data based on irrigation water…. As internal and external pressure is exerted for national standards, a simple approach has been to adopt these EPA criteria.”

The standard the FDA proposes to enforce on farmers, as federal regulation, is derived from a now-superseded U.S. Environmental Protection Agency (EPA) analysis and recommendation to state and tribal health authorities for recreational water.

The EPA criteria were developed for a different purpose than produce safety and in a different context.

This recommendation, and its successor published in the Congressional Record in November 2012 (2), is designed to prevent an excessive number of gastrointestinal illnesses for swimmers, such as 32 illnesses per thousand persons swimming in an ocean or fresh water body of water. The choice of generic E. coli (one of two standards to be met) is based on the actual correlations between epidemiological data on illnesses after swimming and measured indicator organisms. Generic E. coli  (along with Enterococci) had the most robust correlation with actual GI illnesses for fresh-water recreational use out of all the indicators of (mainly) human fecal contamination that were looked at.

Looking at the update 2012 EPA Recreational Water Quality Criteria (3) and consistent with the previous standard: “Viruses are thought to be the etiologic agent responsible for most of the GI illnesses that are contracted in recreational waters impacted by sources of human fecal contamination….”

Generic E. coli bacteria are used as an indicator of fecal contamination to correlate with largely viral GI illnesses, not for identifying the bacterial pathogens that have caused most produce outbreaks, serious illnesses or deaths (nor with more serious viral illnesses).

The original group to adopt the recreational swimming standard and generic E. coli as the indicator species and metric was the CA LGMA. They had their own reasons and context for this choice. One was that, for all its problems, they found that using other bacterial indicators of  fecal contamination were actually worse standards. Generic E. coli was the best of a bad lot of indicators, for their purposes.

To my surprise, I found on interview that there has never been a study of the effectiveness of the LGMA in preventing outbreaks. I suspect the same is true for the related GAPS programs. FDA itself never reported conducting an assessment of the efficacy of any of these programs before adopting their metrics or standards as models. It never even got to the level of evaluating the context, role and effectiveness of the irrigation water standard.

It gets a little worse. A USDA team reviewed “Irrigation Waters as a Source of Pathogenic Microorganisms in Produce” in 2011 (4). They found: “Despite the general belief that irrigation water poses a potential source of pathogens in food-borne outbreaks, there are relatively few confirmed cases in the USA.”

Most of the surface waters of the United States will fail the generic E. coli standard either persistently, predictably during certain times of the year or intermittently. This is a problem for most farmers who want to irrigate. Due to the peculiarities of the proposed rule, however, farmers using buried drip for irrigation and their own controlled wells as a source of irrigation will have an easier time complying.

Commercial bulb onion producers were among the first to comment on the irrigation standards (5). One production and irrigation region in Oregon and Idaho, which produces about 20 percent of U.S. onions, would have to cease onion production under the metrics of this rule (6). The possibility of generic E. coli levels in irrigation water stopping irrigation and therefore killing a crop would prevent farmers from ever receiving a production loan, long before even considering “treating” the water of a 750,000-acre irrigation system.

On the other side of the country, in Pennsylvania, an interesting two-year study of smaller farms with multiple types of surface-water sources for irrigation found that many farms would fail the generic E. coli tests, although O157:H7 and Salmonella were not detected and while Salmonella could be found occasionally when irrigation water would have passed the tests. It also showed that how samples were chilled and iced for transport to an accepted laboratory could greatly affect results and, therefore, consequences of the testing (7).

The problems with surface-water irrigation, identified in early comments, are one of the two reasons that FDA now has called for a full environmental review (EIS) rather than stating the Proposed Rule is categorically exempt (8). The concern is not with the standards, however. Instead, FDA anticipates that so many farmers will switch to well irrigation, despite the expense, that it may affect the status of fragile aquifers.

When the basic standard itself is problematic, all the rest of the regulatory requirements around that standard look very strange. Farmers will have to follow a sampling program that tells them how often they should test using a meaningless standard. They will have to document that they have sampled according to the frequency protocols for the meaningless standard. They will also have to document the results of their useless sampling so they can demonstrate to an inspecting authority that their irrigation water passed the meaningless standard. There are very exact protocols that must be followed. Overall compliance costs (not just for irrigation) are a high percentage of operating profits, but failing the meaningless standards for irrigation could be economically devastating.

FDA added an extra (and almost sadistic) twist. Farmers can be faced with an entire crop being lost because they have to stop irrigating when irrigation water fails the standard. They have the option of treating their irrigation water with a chemical to kill pathogens. This chemical must, however, be EPA-approved for treatment of irrigation water. There are no such EPA-approved chemicals. There are no EPA standards for testing chemicals to be used to kill pathogens in irrigation water. EPA has no plans, currently, to develop such a standard. A chemical that can kill human pathogens could also wreak havoc with soil microbial ecosystems, farm by farm, in entire irrigation systems and across the nation.

FDA’s solution is to delay implementation of the irrigation-water standard by two years.

My conclusion on irrigation-water regulation is that FDA will be bankrupting farmers, driving them out of business, preventing new farmers from entering, and shipping produce production out of the United States, and without enhancing food safety.

What we will get, instead, is food-safety theater.

* * *

Editor’s Note: Reprinted in full with permission from Food Safety News in Seattle, Washington. To read the original article, visit their site at:

FDA's Proposed Produce Rule is 'Food-Safety Theater'

Daniel Cohen is the author of this article.  He is the owner of Maccabee Seed Company in Sacramento, California and has degrees from Yale University and the University of California, Davis.


(1) Trevor V. Suslow. Standards for Irrigation and Foliar Contact Water. Produce Safety Project Issue Brief. An Initiative of the Pew Charitable Trusts at Georgetown University, 2010.

(2) Environmental Protection Agency. 2012 Recreational Water Quality Criteria. Federal Register, Vol. 77, No. 230; Thursday, November 29, 2012, Notices (pages 71191-71192). Notice of availability and recommendations.

(3) Environmental Protection Agency. Recreational Water Quality Criteria (RWQC). Office of Water 820-F-12-058. 2012.

(4) Yakov Pechepsky, et al. Irrigation Waters as a Source of Pathogenic Microorganisms in Produce: A Review. Advances in Agronomy, Vol. 113 pp 73-138. 2011.

(5) National Onion Association. Comment from the Dry Onion Industry Regarding FDA’s Proposed Rules… (Docket No. FDA-2011-N-0921). Docket for the Proposed Produce Safety Rule, Comments, May 2013.

(6) Sean Ellis. “Water Quality Rules Unleash Flood of Fear: Proposed regulations could bring an end to PNW onion crop.” Capital Press [Agricultural Newspaper]. Updated: Saturday, July 06, 2013, 12:11 AM.

(7) Audrey Draper. Microbial survey of Pennsylvania water used for specialty crop irrigation and development of sampling, handling and shipping procedures for surface water testing. MS Thesis, Food Science, Pennsylvania State University. December, 2012.

(8) Department of Health and Human Services, Federal Drug Administration. Notice of Intent to Prepare an Environmental Impact Statement for the Proposed Rule, Standards for Growing, Harvesting, Packing and Holding of Produce for Human Consumption. Federal Register, Vol. 78, No. 160. Monday, August 19, 2013 (pages 50358-50359). Proposed Rules.