Describing the Die-Off of Generic Escherichia coli on Field-Grown Tomatoes in Virginia Using Non-linear Inactivation Models.

Abstract

Agricultural water has been identified as a source of microbial contamination of fresh produce. When surface water is deemed unsafe or not of adequate sanitary quality under current U.S. regulations, growers can implement mitigation measures such as a time-to-harvest interval. Scientifically relevant data on die-off (i.e., inactivation) rates of generic Escherichia coli under field conditions are highly variable by growing region and season. This study used artificially contaminated water to evaluate generic E. coli die-off on tomatoes under field conditions. Field trials were conducted at two locations in Virginia during the summer and fall of 2015 and 2016. Contaminated water (4 log CFU/ml) was used to spray mature tomatoes, resulting in 2 log CFU/tomato of generic E. coli. Tomatoes were enumerated for 7d post-contamination. The relationship between E. coli, time, location, and season were modeled using Weibull and Biphasic models. Akaike's Information Criterion with small sample size bias adjustment was used to determine model fit. A combined site-season Biphasic model, with two distinct inactivation rates, was the most parsimonious model, with growing season contributing more significantly to model fit than site-specific differences. The Biphasic model estimated the mean daily die-off of E. coli populations before the breakpoint (0.87 - 1.16d), at 1.47 - 2.24 log CFU/d with a mean decrease of 1.54 - 2.40 log CFU/d. After the breakpoint, E. coli populations decreased marginally, not at all, with mean daily die-off ranging from -0.05 - 0.07 log CFU/d. Findings indicate that generic E. coli die-off on tomatoes under field conditions undergoes an initial rapid decline within the first 20 - 28h, followed by a slower, prolonged decrease. Environmental factors relating to seasonal differences were more effective at predicting die-off than site-specific differences. E. coli surviving beyond the breakpoint may drive the risk of foodborne illness.