Evaluation and validation of the efficacy of peroxyacetic acid and chlorine as antimicrobial treatments for agricultural surface water to be used post-harvest

Abstract

The recent increase in food-borne illness cases associated with produce threatens public health, the produce industry, and the national economy. Water used in produce washing is one of the routes of microbial contamination, hence the increasing interest and importance of finding efficient antimicrobial disinfectants to treat wash water to ensure that it is potable and safe for produce washing. Objective one evaluated peroxyacetic acid and chlorine as treatments for simulated surface water inoculated with a non-pathogenic Escherichia coli cocktail (ca. 5 log colony forming units (CFU/ml) to determine their efficacy at achieving ‘no detectable generic E. coli in 100 mL of post-harvest agricultural water'. Simulated surface water was prepared to turbidity levels of two and 100 NTU using PTI Arizona Test Dust. Each of the two turbidities was adjusted to both pH 6.5 and 8.4, and the turbid water was equilibrated to 32 and 12ºC. Each sample was inoculated with ca. 5 log CFU/ml of non-pathogenic E. coli and treated with either free chlorine (C) 25±2 ppm, peroxyacetic acid (PAA) 75±5 ppm, or a sterile water control (W), with a 10 second (s) contact time. Following the 10 s contact time, samples were neutralized in Dey-Engley Broth, and E. coli were enumerated using E. coli/Coliform (EC) Petrifilm® at 0, 5, 10, 60, 1440, and 2880 minutes (min). At each time point, samples were also enriched in 2X brain heart infusion (BHI) broth to be streaked on MacConkey agar plates to confirm the absence of E. coli. All C and PAA treated samples were below the limit of detection on EC (5 CFU/ml) and negative for E. coli on MacConkey plates starting from the 0 min time point. This suggests the achievement of no detectable levels of E. coli by both C and PAA after a 10 s contact time. Objective two validated the efficacy of C and PAA at reducing E. coli in surface water sources to the ‘no detectable generic E. coli’ for agricultural water that will have post-harvest contact with the produce, as required by the Food Safety Modernization Act for post-harvest use. Rain barrel and creek water sources obtained from two different farms were inoculated with ca. 5 log CFU/ml of a non-pathogenic E. coli cocktail and equilibrated at temperatures 32 and 12ºC. Samples were treated as described in objective 1 and enumerated at 0, 5, 10-, 60-, 1440-, and 2880-min post-treatment using the FDA- approved IDEXX Quanti-Tray/2000 Colilert method and EC Petrifilm. Samples were enriched in 2X BHI and streaked on MacConkey agar to test for the absence or presence of E. coli. Using Colilert, E. coli was detected until 60 min contact time in samples treated with C, while E. coli was not detected in samples treated with PAA at the 0 min (10 sec contact time) sampling point. When enumerating with EC Petrifilm, the W sample harbored 5 logs E. coli, which was statistically different than C and PAA (P≤0.05), both of which were negative for E. coli following enrichment in 2X BHI. Data collected during objectives one and two suggest that at the evaluated dosage levels and water conditions, chlorine and peroxyacetic acid effectively achieve no detectable E. coli in surface water sources in less than 60 min post-treatment.