An unnoticed side effect of the U.S. Food and Drug Administration’s (FDA) proposed Produce Rules will be the impact on the economics of large-scale animal production in the United States because they will be unable to dispose of their manures and litter off-site.
Cattlemen, large chicken producers and other industry groups may have thought that rules on produce could not possibly affect them. FDA, however, in their notice of scoping for an Environmental Impact Statement (EIS) in the Federal Register, states (1a):
“…Comments received caused FDA to reevaluate the proposed requirements for biological soil amendments of animal origin, which propose an increasingly stringent set of application restrictions based on the likelihood of the soil amendment harboring pathogens. These proposed requirements, if finalized, are expected to result in changes in current use of treated and untreated biological soil amendments of animal origin or potentially greater use of synthetic fertilizers. Changes in the type or handling of soil amendments may significantly affect the quality of the human environment.”
This reason for an EIS comes from FDA’s peculiarly schizophrenic views on human pathogens in manures. Understandably concerned about contamination reaching produce – which can be consumed without a kill step – the agency takes the unusual step of preferring only expensive and energy-intensive chemical or physical sterilization for treatment of animal wastes. These require no waiting period after application on-farm. Then they recognize that sterilized manures and wastes are wide open for recontamination by human pathogens. As a consequence, sterilized composts, litters, or manures are potentially even more dangerous. The net result is they anticipate increased use of chemical fertilizer in conventional agriculture and leave large CAFOs (concentrated animal feeding operations) with a waste-disposal problem.
Of course, composting is one of the several pathogen-reduction techniques allowed by the U.S. Environmental Protection Agency (EPA) for human waste, such as sewage sludge (biosolids), and the proposed Produce Rules have no problem with this use.
Disregarding the explicit language of the Food Safety Modernization Act (FSMA), the proposed rules also impose new restrictions on organic farming, including a 45-day non-harvest period after applying compost. This includes composts that have been tested for O157:H7, salmonella, and indicator bacteria and documented before sale. As one organic farmer commented, “Why not just hold the compost an additional 45 days before application if this actually had any value?”
They also increase the waiting period after applying raw animal manures an additional four to five months (depending on crop) beyond the Organic Farming Production Act (OFPA). In a cynical reversal of congressional language and intent, FDA claims this is not a conflict: if farmers meet the new standard, they will also meet the old OFPA standard. FSMA language states the proposed rules should not cause changes in organic standards. Yet current organic farming laws and procedures of the National Organic Program (NOP), in general, have provided a much greater margin of safety for use of manures and composts than conventional agriculture. On the one hand, food safety across the country would be greatly increased if all farms followed the organic law for these soil amendments. On the other hand, the proposed changes would significantly harm organic farmers (see interviews with Tom Willey and Earthbound Farms in my comment on FDA’s proposed Produce Rule).
A good compost is actively suppressive of human pathogens. This may seem surprising because usually the focus is on whether a compost has been made according to guidelines of carbon:nitrogen ratio, aeration, and other methodologies to guarantee high temperature generation for sufficient time to kill pathogens. If only exothermic, low energy, biologically generated heat sterilization were involved, re-introduction of pathogens by contamination after cooling could allow pathogen regrowth. But this is just the first phase in composting.
However, beginning with the cooling phase in compost development, the development of stable microbial diversity in mature composts can be suppressive of both human and plant pathogens when deliberately re-introduced (1b). Conversely, autoclaving mature compost and re-introducing human pathogens allows for pathogen survival and increase. In general, (2):
“Contrary to the declining populations that are often seen in natural habitats, populations of E. coli can increase in such substrates under sterile conditions, that is, without predatory, antagonistic or competing organisms. This indicates that the natural microbiota in such cases has an overriding effect on survival.”
The same can be true for soils (3), and plants, through their roots, can manipulate the soil ecology of the rhizosphere (4).
The whole field is undergoing a major scientific revolution under the impact of genomics and other biotechnologies. I would expect there to be major reviews and books on the molecular ecologies of farms and the interrelationship with human, animal and avian health in the near future as we learn to enhance beneficial microbial interactions at all levels. One of the major advances is that microbial populations can be characterized without having to be culturable, so a broader picture can be seen.
Some examples include the detection of native microbial populations that can suppress O157:H7 in dairy livestock bedding (5) and the detection of a native Salmonella-suppressant microbe that can be grown and used as a treatment in the production of tomato transplants and carry-over protection to the field in the southeastern United States. The latter came from research by Eric Brown of FDA’s CFSAN (Center for Food Safety and Applied Nutrition) division (6). The American Society of Microbiology recently published a free summary book on beneficial microbial interactions in agriculture (7).
I interviewed two quite different people about this for my comment on FDA ‘s proposed Produce Rule (8). Tom Willey, with his wife Denesse, runs a 75-acre certified organic truck farm near Madera in California’s San Joaquin Valley. His farm relies on dairy compost from a well-established service that tests for O157:H7 and Salmonella, as well as for indicator pathogens. They do not produce fresh-cut mixes or bagged produce because of the increased food-safety hazards.
Will Daniels is vice president for food safety and organic integrity for Earthbound Farms, a major producer of bagged fresh-cut salads that handles more than 2 million pounds of produce per week. They run one of the most extensive microbial testing programs of inputs, produce received, and finished products before shipment I have ever heard of. Interestingly, they are in substantial agreement on using research-based ecological approaches to food safety. They also agree on an overall ecological and integrated approach to human, animal and plant health that respects and works with biological diversity, especially microbial diversity.
How good are the composts that are commercially available? Tom Willey referred me to the only published study I have been able to find on microbial testing of commercial composts, noting it would not be pleasant reading. This is “Occurrence and Levels of Fecal Indicators and Pathogenic Bacteria in Market-Ready Recycled Organic Matter Composts” (9), a study submitted in early 2008 and published in 2009. The lead author, W.F. Brinton, Jr., has worked for years to develop the science and controls for composting, including the response to newly emerging pathogens such as O157:H7 (10 – 13).
This study looked at point-of-sale commercial greenwaste composts produced in the three western states of Washington, Oregon and California. The authors were generally favorably impressed by the results because human pathogens were generally rare, although Salmonella was found in one compost. But they found a very wide range of fecal coliforms, and a slight majority of samples exceeded the EPA 503 standards. Furthermore, 6 percent of the samples had detectable O157:H7.
A few years after the spinach crisis:
“One facility produced compost with a very high fecal coliform level, and this facility was in a noted vegetable production area …. We detected measurable E. coli O157:H7 in samples from three facilities. These facilities were in the large facility group and were situated within important vegetable growing regions.”
Really bad composts were associated with exceeding the indicator bacteria levels even for Class B sewage sludge. Safer composts were correlated with windrowing compost methods, smaller facility size, and compost maturity as measured by the California Compost Maturity Index.
“These data indicate that compost that is hygienic by common standards can be produced, but more effort is required to improve hygiene consistency in relation to management practices.”
What is striking about this paper is that it was based on studying greenwaste, which is generally treated as safe for composting and use, including in FDA’s Proposed Rule. It did not take animal manure as a known input to create an unsafe compost. Some states allow a percentage of manure in greenwaste, but contamination can come from a range of practices in collecting urban greenwaste in particular.
There are strong economic pressures to dispose of a range of waste products on agricultural lands. Urban sanitation systems, ash from biomass-based power plants, greenwaste to avoid landfill limitations, port and river dredging and CAFOs all are examples. All of them have an impact on food safety if they end up in farmland used for produce. But they are not all accurately managed for food safety.
What about U.S. commercial manure-based composts? The best survey I have seen on this is a master’s thesis from Clemson University from 2011. Cortney M. Miller looked at 103 samples of organic fertilizers of various types, including manure compost, from nine states (collected from 2007 to 2010). The initial survey showed (14):
“All of the organic fertilizer samples analyzed in this study were found to be free of Salmonella, E. coli O157:H7 and L. monocytogenes.”
This may be classic risk-adjusted behavior. Knowing there is a potential hazard, the composters using manure (and other biological fertilizers) may be more careful than those using “only” greenwaste, which is assumed to be safe. If there were routine annual surveys of composts made from animal manures, state by state, method by method, and company by company, we would have useful data to work with and compare with similar surveys for greenwaste. But these data do not seem to be reported, if they are ever collected.
Instead, there have been about 20 years of research papers on inoculating various soils, composts, and other inputs, as well as plants directly with known pathogens or their disabled surrogates, following their fate in the particular system, and this became the basis of various regulatory proposals or policy statements, including by FDA and the U.S. Centers for Disease Control (CDC).
As few as 10 E. coli O157:H7 cells can be sufficient to cause a human infection (15). That is well worth being paranoid about or extremely careful. How many O157:H7 cells does it take to contaminate vegetables when the soil or growing media are contaminated by water or manure or compost?
A typical set of papers by Islam et al (cited in the prologue to FDA’s proposed Produce Rule) looked at deliberately contaminated manure compost added to soil (or water) followed by growing leaf lettuce, parsley, carrots and green onions.
“Pathogen concentrations were 10 to the 7th CFU/g [ten million cells/gram] of compost” (16, 17).
For interest’s sake, what is the concentration of O157:H7 in the manure from a supershedder cattle? Supershedders are relatively rare individuals responsible for the majority of O157:H7 shedding from a herd of cattle. Sampling each animal’s own contamination levels in Angus, Brahman and Angus-Brahman breeds, Jeon et al. used the definition of more than 100,000 cells per gram – 10 to the 5th (18) for “supershedder.”
Brahman cattle were highly resistant to carriage of O157:H7 compared to Angus and mixed Angus-Brahman breeds in this study.
The normal range defining supershedders is above 1,000 to above 10,000 O157:H7 CFU per gram – 10 to the third to 10 to the fourth. Some reports have been as high as one million cells per gram – 10 to the sixth (19).
Looking at the highest of these concentration numbers, it would be at least 10 times safer to grow vegetables directly in the manure of supershedder cattle than under the research experimental conditions.
Johannesson et al. (20) gave an example of lower-level deliberate inoculation of O157:H7, not in compost but in dairy cattle manure and urine slurry. This was used as fertilizer one week before transplantation of crisphead lettuce (cv “Coquette”). This was at a level of “only” 10,000 (10 to the fourth) CFU/gram final soil concentration.
“As the bacterium was not detected in the edible parts of the lettuce, the outer leaves of the lettuce, or the lettuce roots at harvest it was concluded that transmission of E. coli O157:H7 from contaminated soil to lettuce did not occur. The pathogen persisted in the soil for at least 8 weeks after fertilizing but was not detected after 12 weeks.”
Interestingly, “Pseudomonas fluorescens, which inhibited growth of E. coli O157:H7 in vitro, was isolated from the rhizosphere.” A “good” soil can be suppressive of human (bacterial) pathogens. This seems to hold for Salmonella as well as O157:H7.
There have been some great recent experiments on the possible contamination of tomato fruits from plants grown in pathogenic Salmonella(s)-contaminated drip irrigation water and for models of floral infection (splashing, overhead irrigation). Jie Zheng and co-authors were trying to prove the possibility of Salmonella surviving and infecting tomato fruits after a persistent outbreak of Salmonella (Salmonella enterica serovar Newport) on tomatoes on Virginia’s eastern shore (21). In this case, the outbreak strain was also found in a pond used for irrigation water, including when drip irrigation was used. Using a 100 million cell/ml root zone treatment, they were able to find one infected fruit, as well as persistence in stems. In contrast, blossom infection, presumably modeling sprinkler irrigation or splash dispersal, was a much more facile route for fruit infection.
So there is a possible pathway that suggests confirmation of the original outbreak investigation, and that even drip irrigation might be a possible pathway. However, as the authors point out, it’s also true that young transplants (with associated root wounding) were at a more susceptible stage – for any transport of Salmonella within the root or stem – and, in routine commercial field production, transplants were going into a routinely sterilized soil (under plastic).
“… interior root colonization might occur passively through wounds in roots that are damaged during transplantation. Moreover, methyl bromide has had a long history of use in tomato cultivation as a soil fumigant in the eastern United States, and recent metagenomic studies have shown that such practices have diminished overall soil microbial diversity, perhaps increasing the potential for Salmonella colonization and persistence in the soil.”
It takes introducing multiple risk factors to get Salmonella-contaminated water to infect a tomato fruit through contamination of roots. Risk could be mitigated in a number of ways to stop this, including a bacterial antagonist root drench at the end of transplant production in the greenhouse, possibly not sterilizing soil, and (of course) not using Salmonella-contaminated water for irrigation, particularly during the time window just before or after transplanting. Soil microbial ecology, or its absence, can be critical.
Here is a trick question: How many cells of Salmonella does it take to contaminate a tomato fruit? No more than 100 cells, if one does it correctly, injecting directly into the peduncle of a young fruit. Unfortunately, Salmonella was able to thrive in this environment and grow to a density of 10 to the seventh per gram fresh weight of pulp.
This was part of a University of Florida team’s experiment looking at what they called worst-case conditions for the contamination of tomato plants through above-ground parts (the phyllosphere, as opposed to the root-influenced zone in the soil, or rhizosphere). They also used a surfactant, Silwet L-77, to aid the contamination of leaves (leaflets) dipped into a solution of Salmonella enterica Typhimurium at a concentration of 10 to the ninth CFU per ml (one billion cells per milliliter) two to four times. They actually found internalization of Salmonella and movement to fruits, although rarely under these conditions.
The authors argue that very low probability events, multiplied by the number of tomato plants commercially grown, show the possibility of this being an actual hazard. This is a slightly foolish justification for the research. Another way to look at it is: don’t sprinkler irrigate or surface irrigate with water contaminated with Salmonella at a concentration of one billion CFU per milliliter after spraying the field with surfactant. This would remove even the very low probability of occurrence through this route of contamination.
The authors state that the practical implication of their work may be “… that application of surfactants, especially Silwet L-77, could enhance the entrance of bacterial pathogens into leaf tissues.”
The more worrisome aspect is when a human pathogen and a plant pathogen synergistically aid each other (23): “The importance of phytobacteria in the persistence of human enteric pathogens on plants first came to light from supermarket produce surveys that demonstrated that 60 percent of produce showing symptoms of soft rot also harbored presumptive Salmonella …. Biotrophic plant pathogens, like P. syringae and Xanthomonas campestris, were also shown to promote growth or survival of Salmonella and enterohaemorrhagic E. coli on plants.”
One of the soft rot pathogens, Dickeya dadantii (Erwinia chysanthemi 3937) is also a pathogen of the pea aphid (24). It’s a complex system.
Internalization of a pathogen is the produce industry’s nightmare because no surface sanitization could even make a difference. Surface contamination seems to do the job of creating outbreaks perfectly well, however.
If you put the studies together, you get what is becoming a common conclusion: while rhizosphere and soil microbial ecology systems can be well-buffered against pathogens, the phyllosphere or above-ground plant parts or edible portions are much more problematic, with much greater variability and much less stability than soil and root ecological systems. And, of course, a research protocol using a sufficient concentration of a human pathogen, far beyond what one could expect on farms, can overwhelm the rhizosphere microbial ecology, often not measured or described in these studies.
The potential ability of human pathogens to infect plants through flower parts, on the other hand, seems particularly troublesome.
FDA did get one part of one aspect of this quite right, in my view. Any application to the above-ground portions of crop plants consumed raw should be evaluated for human pathogen issues of the water used. This includes pesticide applications, where human pathogens can use the pesticides themselves as energy sources (25), dust-control water, where the contaminated dust will then blow back onto growing fields after drying, and any other aerial application – as well as, of course, irrigation. Except for irrigation, these are inputs and sometimes professions (for example, pilots) where being part of the food-safety chain may not have been apparent. Modest awareness could bring large safety improvements, particularly when applications are close to time of harvest (26).
FDA consistently ignored microbial ecology in drafting the proposed Produce Rules. In the case of above-ground plant parts, the phyllosphere, this happens to have some justification – as a first approximation. In the case of soils, rhizosphere ecology, and soil amendments such as composts and manures, it leads them into a disastrous set of proposals.
In the absence of comments from the animal industries, we have the curious situation that CAFOs may only be protected by the objections to the proposed Produce Rule from organic farmers and the sustainable agriculture community (27). FDA’s proposals, remarkably, would require the consumption of massive amounts of energy and cash to make manures and composts – and probably soils – more dangerous.
(1a) 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.
(1b) N. Paniel et al. 2010. Assessment of survival of Listeria monocytogenes, Salmonella Infantis and Enterococcus faecalis artificially inoculated into experimental waste or compost. Journal of Applied Microbiology, 108, 1797 – 1809.
(2) van Elsas et al. 2011. Survival of Escherichia coli in the environment: fundamental and public health aspects. ISME Journal, 5, 173 – 183.
(3) van Elsas et al. 2012. Microbial diversity determines the invasion of soil by a bacterial pathogen. [E. coli O157:H7]. PNAS January 24, 2012. 109 (4) 1159 – 1164.
(4) Berendson et al. 2012. The rhizosphere microbiome and plant health. Trends in Plant Science. August 20, 2012. 17 (8) 1360 – 1385).
(5) Westphal et al. 2011. General suppression of Eschericia coli O157:H7 in sand-based dairy livestock bedding. Applied and Environmental Microbiology, March 2011, p 2113 – 2121.
(6) Described in: Richard Coniff, 2013. Enlisting bacteria and fungi from the soil to support crop plants is a promising alternative to the heavy use of fertilizer and pesticides. Scientific AMerican, September 2013, p 76 – 79.
(7) Ann Reid and Hannon E. Gates. How Microbes Can Feed the World. Report on an American Academy of Microbiology Colloquium Washington DC // December 2012. American Society of Microbiology. August, 2013. See:
HYPERLINK http://academy.asm.org/index.php/food-microbiology/5111-how-microbes-can-help-feed-the-world” http://academy.asm.org/index.php/food-microbiology/5111-how-microbes-can-help-feed-the-world
(8) Daniel B. Cohen. Comment on: Standards for the Growing, Harvesting, Packing and Holding of Produce for Human Consumption. Proposed Rule Document issued by the Food and Drug Administration (FDA). Docket No. FDA-2011-N-0921 Regulatory Information Number RIN 0910-AG35. http://www.regulations.gov/#!documentDetail;D=FDA-2011-N-0921-0196
(9) Brinton et al. 2009. Occurrence and levels of fecal indicators and pathogenic bacteria in market-ready recycled organic matter composts. Journal of Food Protection, Vol. 72, No. 2, pp 332 – 339.
(10) David Stip, 1991. “At Cafe Brinton, Today’s Special is Chicken a la Sawdust: the Julia Child of garbage cooks up tasty compost for the microbial palate”. The Wall Street Journal, Wednesday July 31, 1991.
(11) Mary L. Droffner and William F. Brinton, 1995. Survival of E. coli and Salmonella populations in aerobic thermophile composts as measured with DNA gene probes. Zbl. Hyg. 197, pp 387 – 397.
(12) William F. Brinton, 2000. Compost quality standards and guidelines. Final Report by WIlliam F. Brinton, Woods End Research Laboratory, December 2000. Prepared for: New York Association of Recyclers.
(13) Brinton et al. 2005. Herbicide residues in composts: pH and salinity affect the growth of bioassay plants. Bull Environ Toxicol, November 2005. 75 (5) 929 – 936.
(14). Cortney M. Miller, 2011. Microbiological safety of organic fertilizers used for produce. Thesis in partial fulfillment of the degree Master of Science, Microbiology. (See page 44).
(15) Y. Hara-Kudo and K. Takatori, 2010. Contamination level and ingestion dose of foodborne pathogens associated with infections. Epidemiology and Infection. Volume 139; Special Issue 10; October 2011, pp 1505 – 1510.
(16) M. Islam et al., 2004. Persistence of Enterohemorrhagic Eschrischia coli O157:H7 in soil and on leaf lettuce and parsley grown in fields treated with contaminated manure composts or irrigation water. Journal of Food Protection; 67 (7) pp 1365 – 1370.
(17) M. Islam et al., 2005. Survival of Escherischia coli O157:H7 in soil and on carrots and onions grown in fields treated with contaminated manure composts or irrigation water. Food Microbiology; 22(1) pp 63 – 70. January 2005.
(18) Soo Jin Jeon et al., 2013. Evaluation of animal genetic and physiological factors that affect the prevalence of Escherichia coli O157:H7 in cattle. PLoS ONE 8 (2) e55728 (9 pages).
(19) Elaine D. Berry and James E. Wells, 2010. Escherischia coli O157:H7: Recent advances in research on occurrence, transmission and control in cattle and the production environment. In Steve L. Taylor, editor: Advances in Food Nutrition Research, Vol. 60. Burlington Academic Press, 2010, pp 67 – 118. Elsevier.
(20) Gro D. Johannessen et al., 2005. Potential uptake of Escherischia coli O157:H7 from organic manure and compost into crisphead lettuce. Applied and Environmental Microbiology; 71(5), pp 2221 – 2225. May, 2005.
(21) Jie Zheng et al., 2013. Colonization and internalization of Salmonella enterica in tomato plants. Applied and Environmental Microbiology, 79 (8): pp 2494 – 2502.
(22) Ganyu Gu et al., 2011. Internal colonization of Salmonella enterica serovar Typimurium in tomato plants. PLoS ONE 6(11), e27340. (11 pages).
(23) Maria T. Brandl et al., 2013. Salmonella interactions with plants and their associated microbiota. Phytopathology 103: 316 – 325.
(24) Anne-Marie Grenier et al., 2006. The Phytopathogen Dickeya dadantii (Erwinia chrysanthemi 3937) Is a Pathogen of the Pea Aphid. Applied and Environmental Microbiology, 72 (3); pp 1956 – 1965. March, 2006.
(25) Lopez-Velazco et al., 2013. Growth of Salmonella enterica in foliar pesticide solutions and its survival during field production and postharvest handling of fresh market tomato. Journal of Applied Microbiology; 114 (5) pp 1547 – 1558. May, 2013.
(26). See the interview with Scott Horsfall, CA LGMA in Cohen 2013 (above).
(27) An alternative approach to FDA’s would be much more modest and limited to try and implement known steps that reduce hazards in produce. The current organic rules on composted manures, manure and sludge application are more protective than those for conventional agriculture. They could be adopted for all produce. This includes banning sewage sludge (“biosolids”) from produce production for multiple reasons, including non-pathogenic contaminants.
FDA could regulate the commercial and other inputs to produce, rather than the farmers (or just the farmers).
Manures and greenwaste could be tested for a range of human pathogens before they are even processed for farm application. The value of human pathogen-free manure or greenwaste should be higher in a free market because it will take less work to make them into a safe compost. The value of a pathogen-contaminated raw material should be lower. FDA could mandate testing of raw materials and let the market proceed.
There could be improved consistent grading standards for composts. A good compost has no human (bacterial) pathogens. A great compost would be known to be human-pathogen suppressive. A perfect compost would be plant-pathogen suppressive and have killed any viable weed seeds to boot, but that is beyond the food-safety issues discussed here.
FDA could regulate the government agencies that deliver water. The simplest regulation would be to notify farmers of dredging work that disturbs channel banks or bottoms and potentially releases pathogens. The more complex would be to mandate reporting of water quality to farmers by water agencies, not from farmers to FDA.
A layered approach to produce safety in on-farm production has to include microbial ecology in soil amendments, composts, and, in fact, soils. The complexity of these ecological systems, and of farms and farmers, suggests that centralized management from a single rule and metrics applied everywhere – in the U.S. and for foreign produce – is guaranteed to fail. Working with farmers for improvements in safety, not an impossible guarantee of safety, is more likely to succeed.© Food Safety News