Purpose: The goal of our study was to quantify microbial levels in potential environmental sources of contamination (soil, irrigation water, farm worker hands) and identify their role in produce contamination.
Methods: From 2011-2012, 237 produce composite (cantaloupe, jalapeño, tomato) and 121 matched irrigation water, 79 soil, and 158 farm worker hand rinse composite samples were collected from 10 farms on the Mexican side of the U.S.-Mexico border. From each sample, generic Escherichia coli, enterococci, fecal coliforms, and coliphages were enumerated. The magnitude of association between produce and environmental sample contamination was assessed by chi-square tests and logistic regression models (prevalence data), Spearman’s correlations (non-normal concentration data), and linear regression models (concentrations). Human ethical research approval (IRB) was secured.
Results: In general, farm worker hand samples had higher concentrations of fecal indicators than soil or water samples. Soil had the lowest prevalence of fecal indicators compared to hands and water samples. Pooling data across the two years of sample collection, the concentrations of E. coli (rho = 0.9), enterococci (rho = 0.5), coliforms (rho = 0.3), and coliphage (rho = 0.7) were significantly and highly correlated between hands and produce, but not between soil or water and produce. The presence of E. coli, enterococci, and coliphage was significantly associated between hands and produce (odds ratios ranged between 2.3 - 8.5 across indicators), but not between soil or water and produce. These trends were also observed when analyses were adjusted for crop type and processing.
Significance: These results suggest that decreasing farm worker hand contamination would be an effective intervention to decrease microbial contamination of produce in the study region.