Science-based evaluation of regional risks for Salmonella contamination of irrigation water at mixed produce farms in the Suwannee River watershed.

Outbreaks of human illness associated with produce have resulted in questions about the safety of the water used for irrigating these products. We have assembled an experienced team from the University of Georgia at Tifton and the University of Florida to address the water quality of vegetable irrigation ponds in the Suwannee River watershed. Irrigation water quality standards are not currently regulated or determined by scientifically based metrics. Coliform bacteria are widely used as indicators of fecal contamination, but their validity as indicators of bacterial pathogens is questionable. We propose to investigate the relationship between the occurrence and distribution of these indicator bacteria with that of a specific pathogen, namely Salmonella enterica. Growers in this principal produce production area of the United States have agreed to allow periodic collection of water samples from irrigation ponds to provide preliminary data and validate methodologies. Proposed research will systematically examine bacteria in and around 10 irrigation pond sites for a two-year period under a variety of environmental conditions. This research will identify management practices, environmental parameters, and locale characteristics associated with increased risk of pathogen contamination by irrigation water and will provide a research-based comparison of indicator organisms and Salmonella in a major fruit and vegetable growing area.

Irrigation water is a recognized risk factor for contamination of produce; however, associated risks have not been systematically evaluated. Science-based data are needed to establish and validate Good Agricultural Practices (GAPs) metrics to prevent or mitigate contamination of produce by human pathogens at the production level. In order to evaluate, modify, and implement GAPs, studies must establish appropriate assessment protocols for risk evaluation that are specific for the agricultural practices of that region. Current methods may not be appropriate for detecting specific pathogens and/or lack the sensitivity needed to detect low pathogen levels. The proposed research will evaluate methodologies and examine the hypothesis that specific agricultural practices, and/or environmental parameters contribute significantly to contamination of irrigation water by Salmonella enterica. Most probable numbers (MPN) of Salmonella and fecal coliform bacteria will be determined from irrigation ponds (n=10) at large and small farms with various vegetable and fruit crops in the upper Suwannee River watershed. This region was selected as the survey site because it is a "hot spot" for environmental sources of salmonellosis. Case rates within this region were 50/100,000 in 2007, which is 1.5 times higher than the national average (CDC, 2009), and river water collected locally over a 12-month period was positive for Salmonella in 79.2% of samples (Haley et al., 2009). Salmonella densities directly correlated with water temperature (r=0.49 p<0.05) and precipitation levels (r=0.68 p<0.05) and also increased 62% in summer compared to spring months. These results support the notion that increased environmental distribution of Salmonella may play a role in disease transmission in this area. Parameters to be investigated include the source, type and microbial diversity of irrigation water, physical/biochemical conditions, presence of grassed or buffer areas surrounding irrigation ponds, evidence of wildlife, soil characteristics, fertilization practices and current mitigation measures. Laboratory studies on pathogen survival under specific pond conditions will complement the field observations for the development of a mathematical risk model, which will be used to compare strategies to minimize contamination and improve GAPs. Research goals will be achieved by fulfilling the following objectives: 1) Determine the spatial and temporal incidence of Salmonella in irrigation water. 2) Evaluate the utility of generic microbial, chemical, and environmental indicators as predictors of Salmonella contamination of irrigation water. 3) Experimentally investigate the associated risk factors of Salmonella in irrigation water in order to identify improvements leading to safer irrigation practices 4) Provide education and training to stakeholders on improved agricultural practices to reduce the risk of Salmonella contamination in irrigation water. Results of this research will identify local characteristics that correlate with increased risk of pathogen contamination in irrigation water and determine mechanisms to reduce this risk. The PIs bring combined expertise in Salmonella biology and ecology, pathogen detection and modeling systems, and agricultural extension. The ultimate goal will be to develop predictive models that can be used to implement improved agricultural practices and education programs that will reduce produce contamination.

The original research objectives were modified during revisions of the proposal, and the current proposal includes the following research objectives: 

1. Determine the spatial and temporal incidence of Salmonella in irrigation water and the sediments associated with the different irrigation sources. On-site sampling of 10 farms in the Suwannee River watershed will determine Salmonella distribution by using methods that will maximize the recovery. 

2. Evaluate the utility of generic microbial, chemical, and environmental indicators as predictors of Salmonella contamination of irrigation water and the sediments associated with the different irrigation sources. Salmonella will be monitored and related to fecal indicator assessments, microbial diversity of irrigation sources and types (pond vs. river or other, open vs. closed systems, reclaimed water, drip, pivot, sprinkler, etc.), environmental parameters (temperature, rainfall, nitrogen, phosphate, dissolved oxygen, carbon, conductivity, and flood events), and the extent of buffering capacity and encroachment by wildlife and/or domestic animals.

The following describes the summary of our findings: 

1. Salmonella was distributed throughout all irrigation ponds in the Upper Suwannee River watershed. Although there were some ponds with consistently higher levels of pathogen detection, there were no clear pond characteristics or agricultural practices that distinguished these ponds from those with lower levels. 

2. Several environmental factors showed significant associations with pathogen occurrence and levels, but the correlations were weak (r<0.3)

3. Although we were able to identify significant associations with some environmental parameters and the prevalence of Salmonella, we are still struggling with one of the inherent problems to the investigation of pathogens in the pre-harvest environment, namely the fact that relatively low density of these organisms persist. We have detected Salmonella at all sites, but frequently the MPN values are right at the limit of detection, which begs the question of the accuracy of these values. Attempts to increase sensitivity by milipore filtration, modified Morse filter and IMS and the latest efforts with tangential flow filtration were not successful. 

4. The practice of “cross-plating” (isolation from enrichment onto XLT4 followed by confirmation of CHROMagar) showed 100% agreement with more standard methods using PCR confirmation. This method could be a cost effective alternative to standard methods that would increase the capacity and sensitivity of Salmonella evaluation. 

5. Operational taxonomic units (OTUs) from analysis 16S community bacterial DNA by DGGE also did show strong association with Salmonella occurrence and levels and could be exploited as molecular markers that could serve as alternatives to fecal indicators. 

6. Molecular typing (DiversiLab rep-PCR) indicated an increased incidence (50%) of strains with similarity to clinical isolates as compared to a prior study on the Suwannee River where only 12% showed similarity to clinical strains. Planned genomic studies in collaboration with FDA should help determine the significance of the findings and better define the associated public health risks