Reduction of Escherichia coli and Vibrio parahaemolyticus Counts on Freshly Sliced Shad (Konosirus punctatus) by Combined Treatment of Slightly Acidic Electrolyzed Water and Ultrasound Using Response Surface Methodology

Abstract

The aim of this study was to determine the combined effects of slightly acidic electrolyzed water [SAEW (pH range 5.0-6.5, oxidation-reduction potential 650-1000 mV, available chlorine concentration 10-80 mg/L)] containing 0, 15, and 30 ppm chlorine and 0, 50, and 100 min of ultrasound [US (37 kHz, 380 W)] using the central composite design (CCD) on the reductions of Escherichia coli and Vibrio parahaemolyticus (initial value, approximately 6-7 log(10) colony forming unit (CFU) of E. coli or V. parahaemolyticus/g) and the sensory properties on freshly sliced shad (Konosirus punctatus), in comparison with SAEW or US alone. Another aim was to develop the response surface model for E. coli and V. parahaemolyticus in the shad treated with the combination of SAEW and US. Single treatments with SAEW (chlorine 15 ppm), SAEW (chlorine 30 ppm), or US for 50 min caused a much-less-than-1-log(10) reduction in the number of both E. coli and V. parahaemolyticus in the shad. In contrast, the combination of SAEW (15 or 30 ppm chlorine) and US (50 or 100 min) caused > 1-log(10) reduction of E. coli numbers (1.04-1.86 log reduction) and V. parahaemolyticus (1.02-1.42 log reduction) in the shad. In addition, the sensory properties of the shad were not changed under the harshest conditions of the combination (SAEW with chlorine at 30 ppm and US for 100 min). Response surface models were developed for the population of E. coli (Y = 6.15322 -aEuro parts per thousand 0.024732X (1) -aEuro parts per thousand 0.016486X (2) -aEuro parts per thousand 0.00015X (1) X (2) + 0.00024X (1) (2) + 0.00007X (2) (2)) and V. parahaemolyticus (Y = 5.67649 -aEuro parts per thousand 0.042598X (1) -aEuro parts per thousand 0.014013X (2) + 0.00003X (1) X (2) + 0.00006X (1) (2) + 0.00062X (2) (2) ), where Y is the bacterial population (log(10) CFU), X (1) is ppm chlorine in SAEW, and X (2) is the duration of treatment (min) with US. The appropriateness of the models was verified by bias factor (B (f); 1.10 for E. coli, 1.03 for V. parahaemolyticus), accuracy factor (A (f); 1.11 for E. coli, 1.05 for V. parahaemolyticus), mean square error (MSE; 0.0087 for E. coli, 0.0028 for V. parahaemolyticus), and coefficient of determination (R (2); 0.976 for E. coli, 0.982 for V. parahaemolyticus). To produce a 1-log(10) reduction of E. coli and V. parahaemolyticus, US treatment times for E. coli and V. parahaemolyticus were calculated within the maximum of 54 and 67 min, respectively, at chlorine 10 ppm in SAEW. SAEW chlorine concentrations (ppm) for E. coli and V. parahaemolyticus were calculated within the maximum of 38 and 41 ppm, respectively, at 20 min of US. Therefore, the resulting response surface models for E. coli and V. parahaemolyticus should be further validated on slices of other kinds of raw fish. Ultimately, the response surface quadratic polynomial equations may thus be used for predicting the combined treatments of SAEW and against E. coli and V. parahaemolyticus in raw fish production, processing, and distribution.