Integrating sustainability & food safety: assessing Salmonella serovar fitness in irrigation water & transfer onto crops

Final report for GNE17-150

Project Type: Graduate Student
Funds awarded in 2017: $14,958.00
Projected End Date: 05/31/2019
Grant Recipient: University of Maryland, College Park
Region: Northeast
State: Maryland
Graduate Student:
Faculty Advisor:
Dr. Shirley Micallef
University of Maryland
Expand All

Project Information


The goal of this project was to investigate the relationship between Salmonella enterica hazard in water and observed food safety risk to crops from water use. The survival of various S. enterica in surface and reclaimed water was investigated while evaluating the potential for transition to viable but non-culturable (VBNC) organisms.  Furthermore, surface water used for irrigation was investigated as a priming reservoir for various S. enterica serovars for enhanced transmission onto tomato crops.  Results suggested persistence in water included VBNC subpopulations and was driven by water type. Transfer success onto tomato was driven by serovar, and prolonged incubation in water increased the transfer ability of serovars that initially transferred poorly onto tomato. Finally, attachment to polystyrene and water oxidation-reduction potential were identified as possible indicators of foodborne pathogen transfer success onto tomato. Through collaboration with the Maryland Department of Agriculture (MDA) a seminar was given to enhance food safety education for growers on water quality, testing requirements, and research findings from the present work. Moving forward, a greater understanding of the environmental cues used by S. enterica subspecies enterica responding to the agricultural environment will aid researchers in developing S. enterica targeted on-farm management techniques to ensure safe yet sustainable fresh produce cultivation practices.

Project Objectives:

The goal of this project was to investigate the relationship between Salmonella enterica hazard in water and observed food safety risk to crops if the water is used for overhead irrigation, evaporative cooling, or other plant contact activities. This goal was executed through examining two crucial aspects with multiple serovars and water types; survival in water and transfer ability to crops. Furthermore, in collaboration with the MDA, grower education on good agricultural practices concerning water use was enhanced by incorporating a seminar on irrigation water in Food Safety Trainings. This seminar included sharing research findings and discussing the importance of balancing food safety practices with environmental stewardship.  

The objectives of this study were:
1) SURVIVAL: The persistence of multiple serovars of Salmonella enterica in various types of surface irrigation water sources in Maryland was measured over time. The fate of multiple Salmonella enterica serovars in various water types were used to determine whether viability in water was significantly influenced by water type or serovar. Furthermore, the presence of Salmonella enterica persistence in viable but non-culturable (VBNC) states was quantified by comparing dilution plating and quantitative-PCR (q-PCR) measurements of viable cells.

2) TRANSFERABILITY and ADAPTABILITY: The transferability potential of multiple Salmonella serovars from water to crops was determined. It was attempted to assess the ability of Salmonella to establish on tomato via irrigation water contaminated with varying concentrations of this pathogen.  Additionally, it was evaluated if Salmonella enterica adaptation to water environments increased the risk of water-crop transfer. Assessment of biofilm formation in these water microcosms was conducted to determine Salmonella enterica serovar-dependent risk of biofilm formation in water distribution systems used in irrigation. Targeted gene expression analysis was attempted to evaluate Salmonella pre-adaptation to plant surface colonization. Assessment of relationships were attempted between biofilm formation and surface attachment, and stress response genes which may be important for successful transfer of Salmonella onto plants.



Salmonella enterica has historically been a food safety concern for agriculture on the eastern shore of Maryland, with multiple research teams identifying surface waters as a possible reservoir of this organism (1,2). To combat this hazard, the Food safety Modernization act has compiled stringent guidelines for water quality to be used for growing fresh fruits and vegetables (3). This includes developing a water quality profile based on generic Escherichia coli levels for all irrigation sources used on produce covered by the regulation, as well as employing mitigation steps for water sources that are deemed below quality. Mitigation steps can include a wait period, employing drip tape irrigation, installing treatments in the drip line, or switching to groundwater for irrigation (3). While these interventions may address food safety concerns, all aforementioned actions require added cost, labor or time, and can detract from sustainable farming (4).  If farmers choose instead to switch to other water sources such as groundwater, they can burden a precious natural resource which is already experiencing heavy withdrawals (5).

While microbial safety of fresh produce is a priority, the obligation for growers to comply with current E. coli standards for water is imposed without knowledge of the “real” agricultural risk. Salmonella enterica may or may not co-exist or correlate with E. coli levels, and the risk Salmonella enterica poses if present in these waters is also not well understood. Is there a threshold concentration below which Salmonella enterica is unlikely to transfer onto crops or persist in soil? Can Salmonella enterica take up residency in surface water or is their presence more transient?  Does Salmonella enterica presence in water make it more likely to establish on crops? Are these interactions water type or serotype specific? It has been previously established that culturing media for bacteria can positively influence persistence on plants, such as pre-incubation of E. coli O157:H7 with manure extracts, but impact of environmental media on Salmonella enterica has not been established (6). Furthermore, preferential plant colonization of some Salmonella enterica serovars relative to others has been documented, but whether water environments impact this remains to be investigated (7,8) . The dearth of knowledge provides a strong rationale for further investigation into 1) S. enterica serovar-specific interactions in water and 2) connections between survival in water to persistence on plants. These data are needed to better understand the practical farming impacts of Salmonella enterica presence in surface irrigation waters. With new understanding, better sustainable strategies for risk management can be developed rather than resorting to expensive and energy-consuming remedial actions or use of “higher quality” water such as groundwater that might not promote good environmental stewardship.  

Food Safety trainings conducted by MDA in collaboration with the University of Maryland provide a platform for interactive education on food safety issues and up-to-date research.  Water standards and testing are an important component of this training, but it is often difficult to convey and translate research findings to information growers can use, since so many data gaps on irrigation water-based microbial risks remain. In addition to filling these gaps, this project intends to enhance understanding on water-based food safety risks through grower outreach at food safety trainings. 


  1. Bell RL, Zheng J, Burrows E, Allard S, Wang CY, Keys CE, et al. Ecological prevalence, genetic diversity, and epidemiological aspects of Salmonella isolated from tomato agricultural regions of the Virginia Eastern Shore. Frontiers in Microbiology. 2015;6:415.
  2. Callahan, M. T., Van Kessel, J. A., and Micallef, S. A. (2019). Salmonella enterica recovery from river waters of the Maryland Eastern Shore reveals high serotype diversity and some multidrug resistance. Environ. Res. 168, 7–13.
  3. Standards for the Growing, Harvesting, Packing, and Holding of Produce for Human Consumption, S. 21 CFR Parts 11, 16, and 112, Department of Health and Human Services(2015).
  4. Raudales RE. Characterization of water treatment technologies in irrigation: University of Florida; 2014.
  5. Perlman H. Groundwater depletion: U.S. Department of the Interior    U.S. Geological Survey; 2016 [cited 2017 May 03]. Available from:
  6. Seo S, Matthews KR. Exposure of Escherichia coli O157:H7 to soil, manure, or water influences its survival on plants and initiation of plant defense response. Food Microbiology. 2014;38:87-92.
  7. Zheng J, Allard S, Reynolds S, Millner P, Arce G, Blodgett RJ, et al. Colonization and Internalization of Salmonella enterica in Tomato Plants. Applied and Environmental Microbiology. 2013;79(8):2494-502.
  8. Patel, J., and Sharma, M. (2010). Differences in attachment of Salmonella enterica serovars to cabbage and lettuce leaves. Int. J. Food Microbiol. 139, 41–47.




Materials and methods:

To assess Salmonella population dynamics in various water types, 0.22 µM filter-sterilized non-tidal river, tidal river, pond, and reclaimed water samples were inoculated with various S. enterica strains, both clinical and environmental. Included in this study were S. Heidelberg, Javiana, Typhimurium (multidrug resistant and susceptible), Newport (multidrug resistant and susceptible) and serovar 4,5,12:i:- .

All serovars were singly inoculated into various surface waters to assess for viability over 90 days using both standard plate count and 25µM propidium monoazide (PMA) treated qPCR.

A crystal violet attachment assay was conducted to investigate abiotic surface attachment of all serovars incubating in all waters tested for 48 h in a polystyrene 96 well plate.

To assess transfer potential onto tomatoes, strains were inoculated into non-tidal fresh water and allowed to persist for 30 days. At day 1, 5, 10, and 30 an aliquot of inoculated water was concentrated, washed, then inoculated onto cv. ‘Heinz’ Tomatoes. To accurately assess the effect of incubation time on transfer success, control samples of fresh S. enterica inoculated into re-filter sterilized experimental aliquots were incubated for one day before concentration and inoculation, concurrently with all other samples. Tomatoes were incubated for 14 h before destructive sampling and standard plating. 30 day incubations and transfers were repeated for another non-tidal fresh water source, reclaimed water, and pond water.

To investigate genetic responses of Salmonella negotiating transition from water to tomato, a targeted RT-qPCR assay was employed. S. Newport and S. Heidelberg were incubated for 30 days in non-tidal fresh water, and assay controls were prepared as described above. At the time of inoculation, aliquots of inoculum were fixed with RNAlater. The remaining aliquots were inoculated onto cv. ‘Red Robin’ tomatoes. Bacteria were collected after incuation for RNA extraction. Insufficient amounts of RNA in 30% of the samples delayed further execution of this aim, and augmented experimental designs are being discussed to repeat this objective. 

To explore relationships among data collected, physicochemical parameter data of the water bodies at the time of sampling was paired to each experiment based on water sampling date. Significant correlations between the bacterial parameter of interest and physicochemical data were assessed.

Research results and discussion:

S. enterica serovars declined over time in water, but differentially persisted in water with respect to water type and experimental replicate. Across all serovars, reclaimed water displayed the lowest rate of decay and total log inactivation and was the most consistent between experimental replicates. Low water dissolved oxygen levels at the time of samplings were significantly correlated with high total log inactivation, which was evidenced by decline patterns in brackish water. Serovar specific differences in total log inactivation and rate of decline were evident in non-tidal fresh and tidal brackish water.

Comparing cell viability methods for Salmonella residing in water for 90 days among serovars, non-tidal freshwater samples displayed significantly different total log inactivation between PMA treated qPCR and standard plating quantification methods in both experimental replicates. When comparing each serovar tested, while all serovars in non-tidal fresh water indicated subpopulations entering VBNC states, only select serovars in reclaimed and pond water indicated transition to VBNC. This suggested that specific bacterial serotype interactions in certain water types may drive VBNC transition which presents an exciting opportunity for future research into drivers of alternative bacterial persistence strategies.

Water type and serovar were significant driving factors of attachment capacity. Across all serotypes tested, reclaimed water harbored a significantly lower attachment index than pond and non-tidal freshwater. Across all water types, Salmonella enterica environmental river water strains displayed higher attachment indices compared to non-river water isolates. This being said, attachment for some serotypes varied by water type, suggesting that the interaction of bacterial genotype incubating in different water types could significantly affect attachment success to polystyrene.

With preliminary experiments, it was determined that less than 5 log cfu/mL inoculum level into water rendered the assay highly variable and difficult to obtain data. Therefore it was decided to execute the experiments at one inoculum level and include multiple serovars. Evaluating the effect of increased incubation in water on transfer success, we found that as little as 10 days incubation in non-tidal fresh water significantly increased transferability onto tomatoes compared to incubation in water for 24 hours. Water type did significantly affect transfer success, however this was variable among experimental replicates. Out of three experimental replicates, pond water supported the least among of transfer across all serovars tested.  Serovar type was also a significant factor for transfer success, with serovar specific differences in transferability onto tomatoes evident from day 1 to day 30. Interestingly, while increased incubation in water increased transferability to tomatoes for all serovars, the degree of this increase was serovar specific. For example, some serovars transferred with high success regardless of incubation time, while another serovar tested required increased incubation to achieve high transfer success. Finally another serovar transferred with the least success compared to other serovars despite increased incubation. Taken together this suggests that some serovars may be able to utilize water as a “priming” reservoir for increased success on plants.

Through bivariate analysis of water physiochemical parameters at the time of sampling and experimental data, it was revealed that low water oxidation reduction potential was significantly associated with increased transfer success and increased attachment to polystyrene. As expected, attachment to polystyrene and transfer success after 30 days incubation in water were also significantly correlated and were driven principally by serovar over water type. Taken together, these data provide rationale for further investigation of rapid measurements such as crystal violet assays and water physicochemical characteristics to indicate food safety risk to crops of using surface waters.

Research conclusions:

Much study has been devoted to understanding S. enterica prevalence in surface water body systems, however less is known about S. enterica serovar specific ecology in water, including persistence strategies, associations with water physicochemical parameters, and transfer potential onto crops from water which the present study aimed to address. Our results suggest that surface water is more than a static reservoir for S. enterica and may play an important role in pathogen ecology in the farm to fork continuum.

Persistence in water was found to be significantly influenced by water type. Furthermore, subpopulation transition into VBNC states was found to be a survival strategy in non-tidal fresh and select serovars in reclaimed and pond water. This finding has pertinent public health implications in that there is an increased risk of under-estimating S. enterica prevalence in these water systems, especially in one type of non-tidal fresh water where the discrepancies between culturable populations and VBNC populations were consistently significant. Moving forward, alternative states of persistence should be considered during future environmental risk assessments and exploration into inexpensive, rapid, and reproducible culture independent methods for assessing S. enterica presence in these waters during environmental sampling is warranted to accurately profile water hazards.


Serovar specific differences in persistence in water and transfer to tomato were apparent, clearly relaying that serovars of Salmonella enterica behave in unique ways. More research is needed to evaluate factors which influence long-term success on tomatoes pre- and post-crop contamination. For example, investigation into the genetic basis and environmental triggers of increased transfer (or “priming”) of pathogens to tomato fruit can further illuminate pre-harvest risk factors for more precise mitigation strategies. 

Future work investigating the ability of rapid tests and water physicochemical parameters to predict food safety risk to crops is warranted. For example, the average transfer success onto tomato after prolonged incubation, and not transfer data from 24 h incubation in water, was significantly positively associated with attachment to polystyrene.  This finding provides evidence that biofilm formation may not only be predictive of transfer to tomato fruits but is also a key plant persistence trait which previous history in water affords microbes. Oxidation reduction potential and dissolved oxygen levels of water at the time of sampling significantly correlated with transfer behavior and total log decline from inoculum on agar plates, respectively. At a point where we are still evaluating the effectiveness of sampling for fecal indicators to assess water quality, investigation into additional methods for water quality estimations is very important to improve monitoring approaches.

Participation Summary

Education & Outreach Activities and Participation Summary

2 Webinars / talks / presentations

Participation Summary:

40 Farmers participated
30 Number of agricultural educator or service providers reached through education and outreach activities
Education/outreach description:

Through collaboration with the MDA I was able to give a presentation in February 2019 at an Advanced Good Agricultural Practices training. At this training, my aim was to enhance food safety education for growers on water quality, testing requirements, and research findings from the present work. This seminar provided a platform to translate research into practice and highlight the importance of implementing food safety measures while considering environmental sustainability and conservation.

A technical talk based on this work was also accepted and delivered at the International Association for Food Protection Annual meeting in 2018, IAFP-2018 water Ferelli-SAM-amf.

Currently, I am coordinating with University of Maryland Extension to devise the best strategy for disseminating water use and treatment fact sheets so that efforts are not duplicated.

We expect at least one peer reviewed Journal article from this work.

Project Outcomes

Project outcomes:

Through this project we identified a wealth of study opportunity in Salmonella – water interactions that may have direct profound outputs for mid-Atlantic farmers. In our studies we found that VBNC states may be a strategy for persistence. This physiological state has been under-evaluated in an agricultural context. Moving forward, during routine microbial testing of water sources, it will be important to consider that water samples which test negative for S. enterica may be false negatives and instead may harbor cells in a VBNC state that are not detectable by culture methods. Moreover, investigation of other drivers or environments in the farm to fork continuum used by Salmonella to engage in this persistence strategy remains to be explored.  In terms of translatable outputs, knowledge gained from the present work on S. enterica associated risks in surface irrigation water, as well as updates on current efforts to augment water quality measurements could be disseminated to the grower constituency on a regular basis. 

One year into compliance of the FSMA PSR, especially for smaller more diversified farms (like we have here in Maryland), “one size” adaptation of the Rule will not fit all. This has been recognized on a limited commodity basis; as Mushroom Good Agricultural Practices, Tomato Good Agricultural Practices, and the Leafy Green Marketing Agreement exist to translate food safety principles into sensible on-farm practices.  While these resources are important in helping growers tailor food safety needs to their farm, there is a sense that the next step must be taken in the development of regionally and operational-scale specific guidance whose goals are to enhance food safety minded practices based on individualized hazards and varying technology available for risk mitigation. The present work endeavored to more holistically track one organism, S. enterica, as it moves through the agro-environment, focusing on key negotiation strategies used to combat hurdles from farm to fork.  While we provide some insight on S. enterica dynamics, more research is needed to investigate the genetic responses of this and other foodborne pathogens negotiating the transition from water to crops, both in crop-specific and region-specific realms. Expanding the scope of research to understand 1) new and emerging crop production strategies such as hydroponics, greenhouse cultivation, permaculture, and urban farming and 2) current and future pressures on the farm environment such as climate change and the impact of surrounding anthropogenic activity  is crucial for the future of food safety research. With these perspectives, a better understanding can be gained of current hazards and can be used to anticipate future issues from  S. enterica, other foodborne pathogens, and chemical contaminants which threaten the safety and sustainability of our food and water supply. This variation of a “One health” approach to interdisciplinary public health research can elevate on-farm food safety practices, driving past the previous goal of compliance to regulation towards a farming and research culture which incorporates equitability, innovation, sustainability, and safety.

Knowledge Gained:

This project continued to reinforce our belief that the ‘one size fits all’ approach is not stringent enough to adequately curb food safety risk for fresh fruit and vegetable production. The various factors that influence the variability of Salmonella enterica persistence in the environment do so in a serovar specific manner, and this differential effect extends not only to survival in the primary reservoir (in this cases water) but also to niche transfer (in this case tomato). We must continue to investigate Salmonella ecology and Salmonella enterica movement through the farm-to-fork continuum in a more nuanced fashion, to devise strategies that are effective to the most successful serovars under given environmental conditions in various ecological niches.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.