Purpose: This study developed dynamic and probabilistic models to predict the fates of L. monocytogenes on fresh pork.
Methods: A 10-strain mixture of L. monocytogenes was inoculated on fresh pork skin (3×5 cm) at 4 log CFU/cm2. The inoculated samples were then stored aerobically at 4 (10 days), 7 (10 days), 10 (10 days), 15 (4 days), 20 (4 days), 25 (12 h), and 30°C (12 h). The microbiological data were fitted to the Baranyi model to calculate maximum specific growth rate (µmax; log CFU/cm2/h) and lag phase duration (LPD; h). The µmax and LPD were then fitted to a polynomial equation and the Davey model, respectively. Accordingly, L. monocytogenes growth was simulated under constant and dynamic temperatures. Of 49 combinations (temperature (7)×sampling time (7)), the combinations with significant growth (P < 0.05) were designated ‘1’, and the combinations with non-significant growth were given ‘0’. These growth response data were analyzed with the logistic regression. The model performances for dynamic and probabilistic models were evaluated, using observed data.
Results: L. monocytogenes growth was not observed before 12 h of storage at 4-10°C, and the pathogen started to grow (P < 0.05) after 6, 4, 2, and 2 h at 15, 20, 25, and 30°C, respectively. After a primary model was developed, µmax gradually increased, but LPD decreased as storage temperature increased. The developed secondary models were also acceptable (R2=0.974). Growth/no growth interfaces of L. monocytogenes were produced by the probabilistic model at P = 0.1, 0.5, and 0.9. In addition, the model performance was acceptable for kinetic model (RMSE=0.282) and probabilistic model (concordance index=99.2%).
Significance: The results indicate that the developed models should be useful in predicting kinetic behavior and growth probabilities of L. monocytogenes on fresh pork.