Evaluation of an alternative irrigation water quality indicator

Water used in preharvest production of edible horticultural foods, fruits and vegetables, is universally recognized as a key potential hazard for widespread crop contamination with human foodborne pathogens. Industry guidance, standards, and microbiological and sampling interval metrics uniformly identify nonpathogenic E. coli as the practical and cost-effective indicator of fecal contamination of a water source. Drawn from recreational water quality standards, numbers of E. coli can be associated with the presence of human pathogens. However, years of experience and tens of thousands of irrigation water tests demonstrate that the accuracy of current metrics to predicting risk to consumers following consumption of uncooked produce is very poor. Our project objective is to develop an initial baseline of comparative data for indicator E. coli, Total Bacteroides (an alternative and more robust indicator), in surface water sources to the presence of human pathogenic E. coli and Salmonella in water used for irrigation management, ag-chemical sprays, and other preharvest applications. Studies will be conducted in CA, AZ, and selected regional production districts and diverse types of produce. This project would develop data to support academic, industry, and public health evaluations for the replacement of non-functional quantitative irrigation water standards with a simpler semi-quantitative threshold.

The lack of consistent correlation between chemical or biological indicators of fecal contamination at the local, regional, national, and global level is well documented in the scientific and public health literature. This situation is extensively described in the environmental and public health science literature and encompasses sources of drinking water, regional watersheds, recreational water, wastewater reclamation, and irrigation water. FDA clearly recognizes the absence of clarity and consistency relating to, and the practical limitations of, designing any single or multi-level indicator-based approach to establishing compliant and non-complaint conditions that achieve a measureable advancement in public safety and health-based water quality standards. Recent studies assembled from datasets in different parts of the U.S. and other countries underscore the general lack of correlation between a generic E. coli indicator in irrigation sources and the presence of pathogens in the water or on crops irrigated with the water. In addition, ignoring for the moment the limitations of indicator E. coli as the reasonable practical compromise choice, the selection of quantitative regulatory compliance standards is both unnecessary and costly for many producers. There is no dataset or model we are aware of that could be applied to differentiate the risk associated with use of agricultural water under any combination of conditions between a rolling geometric mean of 125 vs. 126 MPN (CFU) / 100 ml or 234 vs. 235 CFU/100ml as a single sample exceedance point. An alternative qualitative approach is hypothesized. A well-established biological indicator of fecal contamination of water sources is the Gram-negative obligate anaerobic bacterium Bacteroides. Tracking and differentiating animal sources of bacteria in the genus Bacteroides has long been a tool for evaluating fecal contamination of surface water with uses in drinking water sources and for recreational purposes. Recent developments in rapid, quantitative, real-time PCR methods and first generation commercial Total Bacteroides qPCR kits make application in service labs for routine agricultural water testing a plausible reality.

This project would initiate the collection of preliminary data to support collective expert evaluations for the replacement of quantitative irrigation water standards based on generic E. coli with a qualitative presence-absence standard (in reality a semi-quantitative threshold based on a designed riskassumption Limit of Detection) based on Bacteroides as an improved indicator of fecal contamination and more specifically the presence of pathogens. Initially, for the purpose of this project we would use the commercial kit reagents and protocol for Total Bacteroides detection described by GeneSig (http://www.genesig.com/products/9159). 

• Six hundred water samples will be collected from surface water sources in diverse fresh produce production districts predominantly from CA and AZ but including OR and WA

 • Irrigation, foliar contact, and other ag-water samples will be tested for indicators and bacterial pathogens 

• Water samples will be analyzed in a quantitative manner to establish recommended Limit of Detection thresholds for a standard protocol that will define a qualitative assumed or precautionary risk-based metric 

• A qualitative standard is anticipated to decrease the time to reporting (no growth phase needed) and keep cost per test to a comparable level with current expenditures for quantitative E. coli results

 • In parallel, a comparative analysis of published PCR detection systems for Bacteroides gene targets will be conducted to evaluate current kits with non-commercial research detection systems from published literature for fidelity of outcomes (eg. Kitts et al. 2010)

This project resulted in a substantial body of new, detailed data with standardized assays on 730 surface water sources over diverse and numerous locations and times, which greatly increases the knowledge of microbial water quality in key specialty crop regions. The conclusion from these studies is that either FIB (generic E. coli or Total Bacteroides) may provide good indication of a recent and high level of fecal contamination of a water source, but neither FIB is likely to provide a strong indication of bacterial pathogens in that water source. The results showed that the water testing, while required, indicated compliance but did not provide a strong indication of bacterial pathogens in that water source. Regardless, as a precautionary policy, the industry must implement one of the FSMA PSR corrective measures for agricultural water sources if high levels of fecal indicators are present. While the team believes the assay holds much promise as an improved system, the previously known and further verified low correlation with pathogens in typical surface water sources, within the limits of any survey, has not altered this perspective. Further development of a standard method for AllBac testing in accredited labs is worth pursuing for the greater flexibility in overcoming sample-to-submission-compliance time barriers encountered by many farms. Limited detection opportunities to develop a correlation of microbial water quality to pathogen detection on the irrigated crop were possible. It should be noted that the team had fewer than anticipated opportunities to follow the persistence of waterborne pathogens to the irrigated crop in Year 2 of the project because cooperating growers either switched to drip irrigation or to well water or antimicrobial-treated water.