Abstract: The inactivation characteristics of nisin, ?-polylysine (?-PL) and pH on Bacillus subtilis isolated from vacuum packed cooked Penaeus vannamei were studied. Growth/no-growth interface models were fitted using simple Logistic and polynomial Logistic equations and validated, and the synergistic or antagonistic effects between the antibacterial factors were comparatively analyzed. The results showed that the R2-Nagelkerke values of the simple Logistic and the polynomial Logistic models were 0.79 and 0.91, respectively; the Hosmer-Lemeshow test showed χ2 of 2.00 and 0.76, and P of 0.981 and 1 for the simple and polynomial Logistic models, respectively, indicating that both models can well describe the growth/no-growth of B. subtilis under the influence of nisin, ?-PL, pH and their interactions, and that the polynomial Logistic model can better predict the probability of growth of the spores of B. subtilis. In addition, nisin, ?-polylysine, pH and their interactions had significant effects on the growth/no-growth interface of B. subtilis isolated from shrimp (P < 0.05). With increasing the concentration of nisin and ?-PL or decreasing the pH, the growth area gradually decreased, while the no-growth area gradually expanded, and the growth of B. subtilis was inhibited. As the nisin concentration increased, the growth/no-growth limit of B. subtilis shifted in the direction of higher ?-PL concentration and lower pH. Regardless of the concentration of ?-PL, when the nisin concentration was 60, 100, 140 or 180 μg/mL, B. subtilis showed no growth at pH values of 4.0, 4.25, 4.5, and 5.0. When the ?-PL concentration was 60, 100, 140 or 180 μg/mL, the growth probability of B. subtilis was zero at pH values of 4.0, 4.5, 4.75, and 5.25. At pH < 5.0, the growth of B. subtilis was strongly inhibited, and pH below 5.0 was a major factor affecting the growth/no-growth of B. subtilis. The growth/no-growth model predicting growth probability of B. subtilis can provide an effective means for quantitatively describing microbial inhibition on aquatic products and for microbial risk assessment.

Key words: Bacillus subtilis; growth/no-growth model; nisin; ?-polylysine; bacteriostatic effect; pH