The hunt for the cause of salmonella contamination

Diane Ducharme collects samples in a field of North Carolina tomatoes. She’s collecting the samples to do a salmonella analysis on tomatoes.
Photo courtesy of Dave Caldwell, NCSU.

Most people avoid salmonella on vegetables and fruit, but not the scientists at North Carolina State University (NCSU) and the Plants for Human Health Institute (PHHI), with campuses in Raleigh and Kannapolis, N.C. Their goal is to determine how and when salmonella contaminates produce, especially tomatoes. If they can answer these questions, they hope to help farmers grow tomatoes that are less likely to be associated with foodborne illness.

Behind this research is a team of experts in food safety, environmental microbiology and postharvest physiology: Otto “Chip” Simmons III, NCSU research assistant professor of biological and agricultural engineering; Diane Ducharme, NCSU graduate student, extension associate in horticulture and food safety, and coordinator of a Good Agricultural Practices (GAPs) program at the PHHI on NCSU’s research campus in Kannapolis; Chris Gunter, NCSU assistant professor of horticultural science; Lee-Ann Jaykus, NCSU William Neal Reynolds distinguished professor of food science; and Penelope Perkins-Veazie, professor of horticultural science at the PHHI.

Study objective

In collaboration with the U.S. Food and Drug Administration (FDA), NCSU began a three-year study to look at how salmonella could contaminate tomatoes in a production environment. The FDA had previously conducted a similar project with Virginia Tech in Blacksburg, Va.

Simmons notes that the FDA’s concern focuses on why outbreaks of salmonella associated with red round tomatoes have continued to occur periodically in Virginia, Florida and several other states.

More specific rules are likely forthcoming under the Food Safety Modernization Act (FSMA), which was signed into law by President Barack Obama on January 4, 2011, to ensure a safer food supply. The goal is for federal regulators to take a preventive approach, rather than a reactive approach. Part of that prevention process is finding out where and how foodborne illnesses occur.

According to Simmons, the FDA first started talking to NCSU in late 2011, and researchers began collecting samples from North Carolina farms during the 2012 growing season.

NCSU’s role is to conduct an environmental survey on the production of red round tomatoes grown on three selected North Carolina farms. Red, round, fresh-market tomatoes are of particular concern because these are the tomatoes that go to consumers as fresh produce. The three participating farms are located in western and central North Carolina, the state’s major produce-growing region.

Researchers hope to learn how and when salmonella contaminates produce, especially tomatoes. This may help farmers grow tomatoes that are less likely to be associated with foodborne illness.
Photo by pippalou/

The farmers receive FDA and NCSU funding to offset some of the costs they may incur, notes Simmons. They provide the land and tend the crops, and then allow the NCSU scientists, specifically Ducharme, to collect samples.

“We’re trying to look at every aspect of production, so we’re looking at the soil; we’re looking at the irrigation water that’s used; we’re looking at the pre-fruit blossoms; we’re looking at the fruit; we’re looking on leaves and a host of other samples that are collected by Diane during the growing season,” Simmons explains.

Ducharme’s objective is clear: “To identify specific environmental niches for salmonella and correlate salmonella presence with generic E. coli concentrations in environmental waters within tomato field production systems,” according to information from a presentation given at the 2013 International Association for Food Protection (IAFP) annual meeting (

In 2012, Ducharme collected 49 water samples between August and November. She used the IDEXX Colilert and Quanti-Tray/2000 system, with concentrations expressed as most probable number (MPN) per 100 milliliters.

She analyzed 469 samples of tomato (including fruit, blossom and leaf), weed, soil, water, sediment and fecal material. She looked for salmonella by using a modified Bacteriological Analytical Manual method as well as a real-time polymerase chain reaction.

Salmonella was isolated from the July, August and September sampling periods. According to the IAFP presentation, “Forty-nine percent (27/55) of these isolates were from water, 42 percent (23/55) from stream sediment, 5 percent (3/55) were isolated from inside tomatoes, and 4 percent (2/55) were isolated from the surface of tomatoes. Using parameters for generic E. coli from the currently proposed FSMA rules for agricultural waters, 16 percent (8/49) of the water samples met the single-sample action level of 235 colony-forming units per 100 milliliters, indicating fecally contaminated water sources that may require mitigation.”

Otto “Chip” Simmons III says a team of scientists at North Carolina State University is working to better understand why and how salmonella contamination occurs in produce.
Photo by Rocky Womack.

Ducharme transports the samples to the FDA facilities in Maryland and analyzes them using similar conditions and protocols as the previous study conducted in Virginia.

In 2013, NCSU entered into the second year of the three-year study. “We really don’t have any conclusive findings at this point, other than the fact that we found very, very few positive salmonella samples here in North Carolina – and those are primarily associated with the water,” Simmons says.

From this information, the NCSU scientists hope to better understand the routes salmonella might take from one place to another in the environment, and how it can be controlled. They also want to learn if certain farming practices might help control salmonella and other bacterial pathogens. Possible answers in these areas might allow them to “parlay that into our extension program so we can actually make recommendations through our Good Agricultural Practices training to farmers that would help them mitigate any risks that they might have from these pathogenic organisms in the environment,” Simmons says.

This Salmonella typhimurium bacterium is about half a micrometer (millionth of a meter) wide.
Photo by Peter Cooke/Lenier Tucker, courtesy of USDA-ARS.

“One thing I have to say is: We haven’t had an outbreak of salmonella here in North Carolina,” Simmons says. This means something is different in North Carolina as compared to states where there have been periodic outbreaks. The difference could be in the environment or in the handling of produce.

Ducharme says her results indicate a strong correlation between salmonella found in water and sediment. “A clear line was established with this first year of research that shows limited isolation of salmonella in and on tomatoes,” she says. “This is surprising, with the increasing numbers of foodborne salmonella outbreaks associated with fresh-market tomatoes.”

Ultimately, Ducharme hopes her research will help in the management of on-farm tomato production systems. She believes better management may lead to fewer human illnesses.

Generic E. coli

In addition to collecting field samples of tomatoes and tomato plants, NCSU scientists are conducting water testing “based on a commonly used microbial indicator bacteria, generic E. coli. So one of the other aspects of the project is to do a survey of generic E. coli and compare those numbers to any positives we might find in salmonella for the water samples,” explains Simmons. “We don’t look for generic E. coli in all the samples, just in the water, and that will provide a lot of really good information as to how this indicator organism – that the FDA is proposing – might match up with any positives that we might find in water samples.”

Even though they’re testing the water for these indicator organisms, he says they’re not actually collecting fecal samples from animals on or around the farm.

They also know that certain pathogenic bacteria infect both animals and humans. “So we’ve tried to do an environmental survey around our farms to see if there is animal agriculture upstream,” he notes. “All of these can impact surface waters used for produce irrigation; none of the irrigation water used on these farms is from groundwater sources.”

Simmons says these environmental surface waters are considered risky, at least from an irrigation standpoint, because they may contain runoff from other locations. His thinking is that perhaps animal agriculture (farm livestock or wild animals) may be upstream from the produce farms, and runoff could negatively impact watersheds. “We’re trying to take a lot of that information into account when interpreting these data and trying to look at the generic E. coli and salmonella numbers that we find,” Simmons adds.

He stresses that generic E. coli are not pathogenic, so they’re only used as an indicator organism. In addition, for the tests so far, the few salmonella found are not of human origin, so the NCSU team hopes to do a genetic analysis to better understand those organisms. They plan to do this using a type of source tracking to see where those organisms came from and whether they are of wildlife or livestock origin, or even from animals such as turtles in the environment, which he says generally do not pose significant health risks to humans.

Through this environmental survey, the scientists are trying to gain a better understanding of how all this fits into the research puzzle, and what the impact is on North Carolina farms. They also want to know which key components are different in North Carolina than they are in other states that have recurring outbreaks of salmonella associated with tomatoes. By the end of the study this year, they may be able to answer that question and others.

Maybe they will also have some answers or clarifications in case the FSMA rules are adjusted, which could happen. Simmons says the rules are based on scientific evidence and scientific studies, so perhaps the FDA will look at the findings from the NCSU and Virginia Tech studies to refine the rules going forward.

More questions

While it’s good that the NCSU team has not found many instances of salmonella in the state, it makes it hard to come up with enough data to support a comprehensive analysis. This may leave them with more questions than answers. With that in mind, Simmons says they have identified possible steps they can take beyond the three-year study.

“This has led to a whole cascade of other questions,” he says. “Some of the questions we’re really interested in are: Is there an interaction between plant pathogens and human pathogens? What does the community structure look like on the surface of plants? There are a lot of other bacteria and things there, many of which are helpful for the plants. What does that community structure look like? And how do human pathogens fit into that if they’re introduced?”

Simmons explains that by community structure he means bacteria and other types of organisms, such as fungi. “Bacteria are everywhere, and most of them are harmless. Many of them are actually helpful. If it wasn’t for yeast, we wouldn’t have bread, or we wouldn’t have alcoholic beverages. A lot of times we use them for beneficial purposes. It’s those few bad apples in the bunch, the few pathogens that come along that cause us major problems,” he says.

“How do human pathogens interact with the naturally occurring flora associated with the plants, leaves, tomatoes or even the soil?” he asks. “How do the soil microbes interact with this? I would say this is just the tip of the iceberg, and it’s led us to a lot of other questions to think about with this.”

Simmons concludes: “The bottom line is trying to understand the environmental transmission for human pathogens. This public health aspect is really what this project is grounded in – understanding how microbial contaminants move in the environment, how they make it into our food supply. The biggest thing is how we might be able to control them, and that’s a whole other series of research questions. We have to first understand where they’re coming from and how they get there before we can think about developing control plans and mitigation plans.”

Rocky Womack has written about agriculture and business for more than 25 years.