Spatiotemporal Variability in Microbial Quality of Western US Agricultural Water Supplies: A Multistate Study

Abstract

In 2011, the US Congress passed the Food Safety Modernization Act, which tasks the US Food and Drug Administration to establish microbiological standards for agricultural water. However, little data are available for the microbiological quality of surface water irrigation supplies. During the 2015 irrigation season, we conducted a baseline study on the microbial water quality of large irrigation districts in California (n = 2) and Washington (n = 4). Monthly samples (n = 517) were analyzed for bacterial indicators (fecal coliforms, enterococci, and Escherichia coli) and pathogens (Salmonella spp., E. coli O157, and non‐O157 Shiga toxin‐producing E. coli [STEC]). Although there was a high degree of variability (μ ± SD = 59.13 ± 106.0), only 11% of samples (56/517) exceeded 126 colony‐forming units (CFU) 100 mL−1, and only six samples exceeded 410 CFU 100 mL−1. Two volumes of water were collected for pathogen analysis (1 L and 10 L); prevalence of Salmonella in 10‐L samples (68/149) was nearly double of that found in 1‐L samples (132/517). We found STEC during ∼9% of sampling events (58/517); serotypes O26 and O45 were the most common at 31 and 26%, respectively. Pathogens were not associated with exceedance of the regulatory threshold, yet the odds of detecting Salmonella increased approximately threefold (odds ration [O.R.] = 3.14, p < 0.0001) for every log increase in turbidity. Microbiological outcomes were highly district‐specific, suggesting drivers of water quality vary across spatiotemporal scales. The true risk of contamination of produce from irrigation water supplies remains unknown, along with the optimal monitoring strategy to improve food safety. Core Ideas Western US irrigation water supplies rarely exceed FDA agricultural water standards. Indicator bacteria are not predictive of pathogen prevalence or concentration. Prevalence of pathogens in irrigation water is higher than previously reported. Odds of detecting Salmonella improved >270% with an increase in sample volume. More research is needed to provide science‐based guidance to the produce industry.

Microbial Water Quality: Monitoring and Modeling

“…Change in spatial scale, such as catchment size, was shown to affect average concentrations of indicator organisms in agricultural landscapes (Harmel et al, 2010). A regional study toward this end was done by Rafi et al (2018), who analyzed E. coli data from 743 monitoring stations in the Central Great Plains, Cross Timbers, and South Central Plains ecoregions of Texas and Oklahoma. The analysis of data from “all sites,” including least‐impacted watersheds and those impacted by wastewater outfalls and urban areas, shows a weak negative relationship between E. coli concentrations and stream order/watershed size, with E. coli generally decreasing with increasing stream order/watershed size.…”

“…High spatial variability of microbial water quality metrics is commonly encountered. An example of this is presented in the large baseline study that focuses on fecal indicator organisms and Shiga toxin-related genes across large irrigation districts in California and Washington (Partyka et al, 2018). Median concentrations of generic Escherichia coli decreased from May to October in both states.…”

“…Both sampling location and sample size depend on the monitoring purpose. A change in sample size from 1 to 2 L doubled the prevalence of Salmonella in the multisite survey in the western United States (Partyka et al, 2018). One could expect that samples of recreation and irrigation water should be taken at different distances from the banks.…”

Meeting Report: Key Outcomes from a Collaborative Summit on Agricultural Water Standards for Fresh Produce

“…In some instances, datasets from monitoring programs such as those completed by EPA are already in existence but may not be specific to bodies of water utilized for agricultural purposes. Fortunately, research is beginning to take place to fill this need; Partyka, Bond, Chase, and Atwill () completed one of the early studies to regularly monitor irrigation water in several surface water irrigation districts throughout an entire irrigation season. Reitz () also has been working toward a regional approach to sampling water from irrigation canals in the Treasure Valley region of Eastern Oregon and Western Idaho to reduce growers’ cost and time investment on a broader geographic level.…”

“…Partyka et al. () surveyed agricultural water quality in California and Washington in multiple surface water irrigation districts during an entire irrigation season and found a high degree of spatial and temporal variability in microbial indicators of quality. McEgan, Mootian, Goodridge, Schaffner, and Danyluk () characterized water quality for sources in Florida and found a lack of correlation between Salmonella levels and microbial indicators, as well as a lack of correlation between Salmonella levels and physicochemical indicators.…”