Escherichia coli is a bacterial species that lives in the gut of all warm-blooded animals, fish, birds as well as reptiles and is commonly used as an indicator of fecal contamination in water. This project assessed currently used culture based media for the detection of E. coli in irrigation waters used in Arizona and California, and will present recommendations towards the most reliable media for the evaluation of irrigation waters used for produce. Currently, no microbial indicator standards exist for irrigation waters used for produce production in the United States. The produce industry suggests that the recreational water standard guideline (126 E. coli/100 ml) established by the United States Environmental Protection Agency (USEPA) be used. There is concern that the false positive rate of E. coli detection may be high in these waters giving false indications of the level of risk from enteric pathogens. This project evaluated three commercially available media for E. coli detection to test irrigation waters from three agricultural areas (Yuma and Maricopa, AZ and Imperial Valley, CA) and then assessed false positive rates by utilizing Polymerase Chain Reaction (PCR) and DNA sequencing of the bacterial isolates. The media that were chosen for evaluation were (1) MI Agar, (2) IDEXX Colilert Quanti-Tray® and (3) m-ColiBlue24® broth, all evaluation media accepted by the USEPA and widely used in the monitoring of irrigation water quality by the produce industry. Four hundred and fifty 1-L irrigation water grab samples were collected between March 2012 and November 2012. The samples were analyzed for both cultural counts and water quality parameters including temperature, salinity and pH. Isolates positive and negative for E. coli were then selected and assessed utilizing PCR and DNA sequencing. The false positive rate of each method was found to be high, with MI Agar, m-ColiBlue24® broth and the IDEXX Colilert Quanti-Tray® at an accuracy of 67%, 72%, and 51% respectively. A false positive result is reported when presumptive E. coli sub cultured from the media is found to be non-E. coli through molecular analysis. Overall the IDEXX Colilert Quanti-Tray® performed at a greater rate of accuracy than the other two media evaluated, however, high false positive rates may lead to inaccurate assessment of water quality.

Surface waters used for irrigation purposes in South Africa have been found to be of poor microbiological quality and to be contaminated with human pathogens. These pathogens can be transferred from contaminated water onto fresh produce and potentially cause human infections. Cryptosporidium and Giardia are waterborne parasitic protozoa that have been found in surface waters worldwide. They can cause morbidity in infected individuals and be lethal when infecting people with compromised immunity. This study was divided into two phases. The first phase was a field survey aimed at determining the incidence of human pathogens Cryptosporidium, Giardia and Salmonella spp. in rivers used for irrigation purposes in South Africa as well as on vegetables irrigated with these rivers. The rivers selected were from three different provinces of South Africa. The relationship between faecal indicators and the presence of Cryptosporidium oocysts and Giardia cysts was also investigated. Out of the 30 water samples analysed, 43% were positive for Cryptosporidium oocysts, 23% tested positive for Giardia cysts and 27% were positive for Salmonella spp. However, no Cryptosporidium oocysts or Giardia cysts were found on the vegetables analysed. No significant differences in the prevalence of Cryptosporidium and Giardia and indicator parameters were observed between the three rivers. Using a logistic regression model, no significant correlations were observed between the incidence of faecal indicators and the presence of Cryptosporidium oocysts and Giardia cysts. In the second phase of the study, the individual and combined effects of chlorine, blanching, blast freezing and microwave heating on Cryptosporidium parvum oocysts inoculated on green peppers were investigated. The viability of the oocysts after treatments was determined using the vital dye propidium iodide. Stained oocysts were counted with a flow cytometer. Chlorine treatments did not significantly affect the viability of the oocysts. Blast freezing affected the viability of 20% of the oocysts. Both microwave heating and blanching affected 93% of oocysts. Combined treatments with chlorine and blast freezing did not affect the viability of the oocysts significantly compared to the control. Combined treatment with chlorine and microwave heating was significantly more effective than microwave heating alone and affected 98.1 % of the oocysts. The data indicate that C. parvum oocysts are sensitive to heat and, to some extent, to freezing temperature but are resistant to chlorine. The results of the survey show the presence of Cryptosporidium and Giardia in irrigation waters and thus a possible health risk associated with the consumption of raw vegetables as those can become contaminated via the irrigation water. The results of the challenge tests indicate that C. parvum oocysts on vegetables are inactivated by blanching and microwave heating but survive blast freezing and exposure to chlorine. Boiling and microwave heating of vegetables should be sufficient to kill C. parvum. On the other hand, ready-to-eat vegetables could be at risk of carrying live C. parvum oocysts as the use of chlorine in washing bath is not expected to inactivate C. parvum oocysts present on vegetables. Copyright / Dissertation (MSc)--University of Pretoria, 2012. / Food Science / unrestricted