Spatiotemporal Variability in Microbial Quality of Agricultural Water Supplies: Implications for Cooperative Sampling

Introduction: The Produce Safety Rule (PSR) requires growers of covered commodities to inspect, evaluate, and monitor the quality of irrigation water through routine testing for generic Escherichia coli. While each grower is responsible for documenting their own water quality profile (WQP), the law also allows growers to utilize alternative sampling, i.e. developing a joint WQP from common sources, provided there are no potential sources of variability between the sampling site and point of delivery.

Purpose: This study was conducted to (i) evaluate potential drivers of bacterial variability in surface water sources and, (ii) examine the spatial scale at which samples maintained statistical similarity in order to facilitate cooperative sampling strategies.

Methods: Indicator bacteria (E. coli and fecal coliforms) were enumerated in water samples collected from 66 sites across six irrigation districts in central Washington and northern California on a monthly basis through the 2015 irrigation season (n=517). Prevalence of pathogens (Salmonella spp., E. coli O157 and shiga toxin-producing E. coli (STEC)) was determined for all one liter samples and subset (n=149) of high-volume (10 liter) samples. Physiochemical water parameters, in situ meteorological data, and a detailed environmental assessment were also included in the geospatial and statistical analyses.

Results: Escherichia coli results exceeding the PSR standard of 126 CFU/100 ml were rare (55 of 517) and were not associated with pathogen occurrence in 1 or 10 liter samples (P>0.05). Escherichia coli exhibited strong seasonal (P<0.05) and regional trends (P<0.001) with highest concentrations in California occurring during the late Spring. Significant spatial autocorrelation (P<0.01) was evident at the 5-10 km-scale and was not explained by environmental characteristics (P>0.05), including adjacent land uses.

Significance: This preliminary study suggested that adjacent sampling sites share similar water quality characteristics, which could facilitate cooperative sampling amongst growers sharing common irrigation sources. Additional sampling at finer spatiotemporal scales is ongoing and necessary to resolve site-specific trends.