Abstract: The enterotoxin of Staphylococcus aureus in pasteurized dairy would cause harm to human health. The growth kinetic of Staphylococcus aureus in pasteurized milk was monitored under 4, 10, 16, 20, 24, 28 ℃, and fitted by the Modified Gompertz model, Logistic model, Huang model and Baranyi model, respectively. The square root model and Arrhenius model were fitted to describe the maximum growth rate (μ_max) with temperature. It was shown that the Modified Gompertz model performed the best, as the determination coefficient (R²) was greater than 0.98, and root mean square error (RMSE) was smaller than 4.6. So, the Modified Gompertz model was chosen as the primary model. Compared with the square root model, the Arrhenius model was a better choice, which was indicated as R²=0.99 and RMSE=0.60. The external verification results showed that the accuracy factor (A_f) and the bias factor (B_f) of the Arrhenius model were 1.39 and 0.87, respectively, which well indicated its prediction potential. The results of this study offered references to predict the shelf-life of dairy products, and provided parts of the input data for microbial risk assessment of Staphylococcus aureus in pasteurized milk.

Key words: Staphylococcus aureus, pasteurized milk, predictive model