John Baillie has often shared with the industry insightful comments regarding how abstract proposals impact real life farmers. We’ve included his comments in pieces such as these:
As floods recently began to effect acreage in Salinas, John sent us this note:
The Salinas Valley is going through a flood event; as we speak, roughly 1000-1500 acres are affected. With ALL the different interpretations of LGMA and all the food safety rules the different processors have, I can see AGAIN who will take the blunt end of the bat again.
The farmers along the Salinas River have been told that ANY CROP that had flood water in the field will be destroyed. This is because of what SOUND SCIENCE-BASED FACT??? As far as I am aware of, no one has been testing the flood water as of this morning – March 29, 2011. A meeting did take place around lunchtime though, and I was told that a group from UC Davis was being requested to come down to start taking samples of the river water.
What will they find??? Will it be any different than the water used to irrigate crops in the deserts of California and Arizona that come out of the Colorado River??? What about the crops currently being harvested in the central San Joaquin Valley that get their water from the Aqueduct??? They pulled 15 cars out of that in the past two months…
My question is why is the water in the Salinas River worse than the Central Valley or the desert??? Do they know something we don’t??? SOUND SCIENCE-BASED FACTS???
— John Baillie
Jack T. Baillie Co, Inc.
We thought we would turn to Trevor Suslow, Ph.D., Extension Research Specialist, Post Harvest Quality and Safety at UC Davis, for more insight on this issue. Dr. Suslow has also contributed much to industry understanding of these issues, including in these pieces:
We asked Dr. Suslow to address some of the issues raised by Jack Baillee:
We are once again activated to assist in questions and responses that have re-surfaced due to the recent flood events along the California Central Coast in and around March 26-28, 2011. I am responding with a bit more formality and lack of brevity as I have received several requests related to this topic over the past ten days, and I felt this is a good opportunity to develop an Extension-style response that has a bit more substance in advance of our formal manuscript regarding the data developed over the past five years.
Tracks In The Flooded Field
Fortunately, the extent of the impacted areas was far less than the widespread flooding in March-April 2006, which followed our earlier grower-supported research effort in a 2004-05 environmental investigation of an outbreak-implicated ranch. The key questions throughout have remained consistent as we attempt to contribute ‘real-world’ data to the development of Best Practices and practical ‘metrics’ for industry and public health regulators to consider in erecting standards and audit criteria:
• What is the appropriate pre-plant interval after flooding?
• What is the appropriate re-plant interval after flooding?
• If the ground is seeded but not emerged?
• If the plants are emerged but at least 30 days from harvest?
• If the plants are head, loose leaf, hearts, or baby-leaf/spring mix?
• What is the appropriate distance from the visible flood boundary?
• How does soil texture (sandy to clay loam) affect sub-surface transport?
• How may this impact well water quality?
• What is the buffer zone from the flood boundary if planted prior to flood?
• How much do I need to disc down to be safe?
It is easy to understand how many possibilities for Actionable Responses and Corrective Action Plans are generated from flooding as a risk factor to food safety management, especially for leafy greens, culinary herbs, and related low-growing horticultural foods. A simplistic, highly conservative standard is a tempting approach to protect both consumer safety, as the overriding goal, and economic integrity and sustainability of our farm community, large and small.
For quick background, the cumulative 2011 rainfall in the Salinas region that preceded the flooding was approximately 5 inches during March with around 2.4 inches falling on already well-saturated ground between the 18th and 23rd. In contrast, rainfall in the same area from March 1 to April 4, 2006, exceeded 10 inches. Since 2006, a significant amount of land susceptible to flooding has been taken out of production, converted to other crops, and some levee and creek/watershed drainage improvements have been made to reduce the likelihood of breaching banks during this type of seasonal weather.
In addition, the metrics and responses to flooding have been adopted as uniformly applied and audited standards for the LGMA signatories. As a backdrop to the standards adopted, the following communication from FDA (November 4, 2005) set the table for a lengthy and impassioned debate during Fall 2006 to Spring 2007 in the development of the COMMODITY SPECIFIC FOOD SAFETY GUIDELINES FOR THE PRODUCTION AND HARVEST OF LETTUCE AND LEAFY GREENS (CSGLLG; April 18, 2007 amended August 4, 2010):
Letter to California Firms that Grow, Pack, Process, or Ship Fresh and Fresh-cut Lettuce/Leafy Greens
FDA considers ready to eat crops (such as lettuce/leafy greens) that have been in contact with flood waters to be adulterated due to potential exposure to sewage, animal waste, heavy metals, pathogenic microorganisms, or other contaminants. FDA is not aware of any method of reconditioning these crops that will provide a reasonable assurance of safety for human food use or otherwise bring them into compliance with the law. Therefore, FDA recommends that such crops be excluded from the human food supply and disposed of in a manner that ensures they do not contaminate unaffected crops during harvesting, storage or distribution.
Adulterated food may be subject to seizure under the Federal Food, Drug, and Cosmetic Act, and those responsible for its introduction or delivery for introduction into interstate commerce may be enjoined from continuing to do so or prosecuted for having done so. Food produced under unsanitary conditions whereby it may be rendered injurious to health is adulterated under § 402(a)(4) of the Federal Food, Drug, and Cosmetic Act (21 U.S.C. 342(a)(4); (US FDA 2004). Areas that have been flooded can be separated into three groups: 1) product that has come into contact with flood water, 2) product that is in proximity to a flooded field but has not been contacted by flood water, and 3) production ground that was partially or completely flooded in the past before a crop was planted.
2006 was certainly not the first time there was regional flooding, and the March-April events preceded the September 2006 E. coli O157:H7 outbreak on spinach that is seen as the tipping-point for consolidating the California and Arizona industry efforts to erect a uniform food safety platform and voluntary marketing agreement audit protocol. However, the FDA letter loomed in the industry’s mind in early 2006.
In short, the outcome was to define operating Best Practices to reduce the potential for cross contamination of crops adjacent to, but not directly impacted by, a flood event; erect audit criteria for re-plant and pre-planting time intervals for impacted areas; and create mitigation steps to reduce the uniform time interval for plant-back restrictions on a case by case basis. The basic tenets (not the specific language, to improve clarity for non-industry readers) include:
1. Identify, buffer, and do not harvest any product within a minimum of 30 feet of the visible flooding leading edge. The basis for the 30-foot distance is to accommodate a generous turn around distance for production equipment to prevent contact with or movement of impacted crop and soil to minimize mechanical cross-contamination of non-flooded ground or produce during seeding, transplanting, or harvest.
2. The need for a non-harvest, non-traffic area greater than 30 feet must be based on a risk analysis conducted by a qualified food safety professional.
3. Flooded ground may be planted or re-planted following a plant-back interval of 60 days, provided that the soil has sufficient time to dry out. Appropriate soil testing can be used to shorten this period to 30 days prior to planting. The 60-day planting or plant-back restriction was the end-point compromise, balancing input from multiple sources, including regulatory opinion and expert solicitation.
In the end, the Actionable Response was held to be consistent with the USDA National Organic Program standards on use of manure for compliance in organic certification. This was weighed against more conservative approaches citing various research publications reporting survival and recovery from soils amended with artificially contaminated manures (E. coli O157:H7 and Salmonella spp.) exceeding 250 days. The rationale applied was that a 60-day restriction prior to planting combined with an average 60-day interval prior to harvest with many, but not all, susceptible crops grown close to the soil level was a practical and effective starting point for the metric the LGMA would use until research could define a revised science basis for increasing or decreasing this timeframe.
The natural E. coli decline after a flood observed (top graph) looks a lot like the decline after a simulated contaminated irrigation event studied in the same field two-years in a row with the generic E. coli and non-pathogenic E. coli O157:H7 cocktails (bottom graph). These slides are derived from studies conducted by a team of individuals led by Suslow UCD and Koike UCCE Monterey.
One of the most frequently cited research papers that had been used to argue for plant-back restrictions of 220 days (some up to one year) and also to support the 60-day interval was the excellent report from Jiang et al. 2002. Fate of Escherichia coli O157:H7 in Manure-Amended Soil. AEM p. 2605–2609. One has to be very careful about reading the details of the study and guard against cherry-picking the data to support a position.
In short, the key discrepancy surrounds focusing on the data derived from non-autoclaved soil (higher microbial competition) and autoclaved soils (low microbial competition) amended with a range of inoculated manure (achieving approximately 1 million E. coli O157:H7/gm of soil) additions to the test soil equivalent to approx. 16 to 160 tons of manure per acre-foot of soil. This level of soil amendment, developed for research purposes, was held to represent a true ‘worst-case-scenario’ and well beyond the expected flood-borne contamination in the absence of a known point-source impact (such as untreated wastewater discharge).
Flood-Impacted Soil Near Head Lettuce
Taking all this into account, 60 days was deemed a responsible waiting period, and provisions were established to lessen this period for faster-cycling crops (spinach, spring mix, etc.) with evidence for multiple ground-working events and adherence to soil testing guidance.
Additional details of the full considerations including risk issues, land use factors, equipment cleaning and sanitation Best Practices, and soil testing recommendations are available in LGMA documents and associated Technical Basis documents (available at http://www.westerngrowers.org/ and http://www.caleafygreens.ca.gov/).
Since that adoption of the CSGLLG, a large body of data has been generated by researchers around the globe that are relevant, in a general or reasonably specific sense, to the original grower- and handler-generated questions that rose to the forefront in 2006. Without going into great detail as to the full range of outcomes from various model and research farm plot studies, I will briefly summarize our findings from the 2006 flood survey in the Salinas Valley and our three years of lettuce and leafy greens on-farm field studies conducted in the same region with Steven Koike and other UC Cooperative Extension associates in Monterey County.
This research was funded by a combination of the California Leafy Greens Research Program, USDA CSREES, and the Center for Produce Safety. Our key findings included the following:
2005-2006 Survey of 10 Pre-plant and Post-plant
Fields and 2 Re-plant Ranch Blocks
Absence of detection of elevated indicator E. coli levels in post-flood soil after water receded or rapid decline of populations following ground work to prepare for planting (undetectable within 30 days).
No differences in coliforms or generic E. coli on flood-impacted lettuce and lettuce in the same block outside the visible leading edge of the flood water, at the time of sampling from older or newly emerged leaves.
Absence of E. coli O157:H7 and non-O157 EHEC in any sample (the same procedures detected culture-confirmed contamination of both types of pathogenic E. coli in subsequent risk-based field assessments not associated with flooding).
Absence of E. coli O157:H7 and non-O157 EHEC in 200 whole lettuce plant tests (flood inundated head lettuce at ‘cupping-stage’ and mid-maturity Romaine lettuce) and 500 loose leaf lettuce plantlets (4 true leaves) seeded as replants to flooded field within 30 days of the flood event. The same procedures have detected culture-confirmed contamination of both types of pathogenic E. coli in subsequent risk-based field assessments not associated with flooding. At the time, the industry was not focused on Salmonella as a primary concern and this was not included in field testing during these events.
Field Research Outcomes Relevant to
Post-flood Plant-back Restriction Intervals
Inoculating soil with waterborne inoculum or a sand-organic carrier matrix of lab-grown generic E. coli (3 isolates) and attenuated (non-toxin forming) E. coli O157:H7 (2 isolates) contaminants resulted in the observation of a rapid decline in recoverable populations from the soil (as short as 15 days and within 30 days overall), limited spread from the point of solid form inoculation (less than 15 inches) and no detectable recovery from plants at maturity, all under overhead irrigation.
Incorporating artificially ‘contaminated’ crop residue into soil extended the period of recovery of the same mixtures of test E. coli to greater than 80 days when no additional ground work was done. These studies are on-going with the addition of various degrees and frequency of sequential soil cultivation. Copies of the most recent reports are available at http://www.calgreens.org/ and Trevor Suslow
Ph.D. Extension Research Specialist
Post Harvest Quality and Safety
We are deeply appreciative both to John Baillie for sharing the poignant position such rules actually place farmers in and to Trevor Suslow for explaining the way the thought on this issue has developed.
Dr. Suslow was kind enough to share with us some of the contemporary work going on right now. Please understand that we had to persuade him to do this as some of the work is not yet final and has not been peer reviewed.
The problem, of course, has to be dealt with now and that means we are often dealing with less than perfect information.
Indeed, to some extent, the problem is that the industry contradicts itself. On the one hand, we insist that food safety regulations be “science based,” yet in many cases we just don’t have very good science to go on.
Inherently most of these restrictions are not purely science in the sense that there is any good research that says a buffer zone of 9’ 11” is inadequate and a buffer zone of 10’ is safe. Equally there isn’t the rigor of science to prove that a replant interval of 59 days puts us all at mortal danger, whereas 61 days assures complete safety.
To a farmer, this is his land and his livelihood, and one can certainly only empathize with a man who loses his living as a result of these rules — when, possibly, the product he has to sell is perfectly safe.
On the other hand, following the spinach crisis of 2006, the industry came to a consensus that we simply can’t wait for perfect science to tell us what to do. That means, inevitably, we were accepting that we would err on the side of caution in establishing industry standards.
For processors, who do not own the land or crop that has been flooded, the risk of a problem almost certainly outweighs any benefit from buying this crop.
What John Baillie is urging is a kind of individualized response. Test the flood waters, seek out pathogens and make a decision based on that.
There are opportunities in the metrics to individualize risk assessments but, for the most part, it is an appealing idea but probably impractical. Even if such a test could be done and produce reliable results for all the land subject to flooding — not at all certain — food safety systems depend on easily followed rules, not individualized responses.
Many things that seem to suggest an injustice, say why river water being used for irrigation in one place is OK but river water flooding someplace else renders the crop unsalable, really must be viewed in the context of risk management perspectives and decisions that are, in fact, highly dependent on local risk factors. The chemical and biological hazards associated with regional storm run-off and flooding in the Salinas Valley are likely different than the situation found in another region. The safety margins for replant restrictions may be longer and the potential for transference of contaminants may vary based on the climate and environment. Not surprisingly, farms in regions that have experienced problems will be held to a higher standard. In fact the founding of the CLGMA was really a consequence of the industry in Salinas saying it wanted higher standards.
Is there any solution for farmers? The obvious two: They either get flood insurance on their crop or, if the land is subject to flooding, look to grow crops that are not eaten raw.
The food safety priority is not a secret, and one impact of the rules of the CLGMA and of the various fresh-cut processors is that it changes the economics of using certain land for certain purposes. That means some farmers will come out winners and others losers. That is not the intent of the food safety rules, but is the effect of all regulatory changes. Farmers are, for better or worse, not exempt from this dynamic.