Development And Validation Of Predictive Models For The Growth And Thermal Inactivation Of Salmonella Spp.in Ready-to-eat Tuna

Tuna is usually eaten in raw or just being cooked in lightly fried style.However,the consumption of tuna products,such as sashimi or mildly cooked tuna meat,have a high risk of introducing foodborne disease by Salmonella spp.Predictive microbiology is an important tool for food safety,microbiology it is the basis of the quantitative microbial risk assessment.This aim of this study was to（1）:investigat the growth characteristics of Salmonella spp.in tuna under isothermal conditions,then to construct and validate predictive models through a one-step analysis approach;（2）evaluate the effect of background microflora on the growth of Salmonella spp.in tuna and then to develop competitive growth models based on the interaction between the bacterias;（3）investigate the thermal inactivation of Salmonella spp.in tuna under isothermal and dynamic temperature conditions,and then to develop and compare kinetic models based on static and dynamic method.（1）A two-strain cocktail of Salmonella spp.was inoculated to irradiated tuna.Isothermal studies were performed at different temperatures between 8℃and 35℃to generate growth curves at each temperature.Duplicated experiments were conducted,and the data set from one replicate was used to directly construct both primary model（Huang model and Baranyi model）and secondary model（Huang Square-Root model,HSR）through onestep kinetic analysis.The Fourth order Runge-Kutta method and nonlinear least square method were combined to searching the parameters of the models.The data set from the other replicate under constant temperatures and newly designed dynamic temperature conditions were chosen for model validation.The results showed that one-step approach can be used to analyze the growth curves of Salmonella in tuna sashimi.Though both Huang-HSR models and Baranyi-HSR models had an equal goodness of fit,the former was the recommended model in this study,because of the explicit form of definition for the lag phase in Huang model.The minimum growth temperature of Salmonella was 6.91℃,and the maximum cell density of Salmonella was 9.15 log CFU/g.The root-mean-square errors（RMSE）of validation at isothermal condition and dynamic condition were only 0.37log CFU/g and 0.44 log CFU/g,with the residual errors of predictions following Normal distribution and Laplace distribution,respectively.For each of them,nearly 64.80%and 79.00%of the residual errorswere within the range of-0.5 log CFU/g and 0.5 log CFU/g.（2）A two-strain cocktail of Salmonella spp.and six-strain cocktail of background microflora were both inoculated to irradiated tuna.The inoculated samples were divided into two groups,according to the concentration of the background microflora.Isothermal studies were performed at different temperatures between 8℃and 30℃to generate growth curves The data of duplicated experiments at each temperature were combined and used to directly construct both primary competitive model（Lotka-Volterra model）and secondary model（Huang Square-Root model,HSR）through ones-step kinetic analysis.The competitive growth data of Salmonella spp.and high-and low-concentration of background microflora under 10℃,12℃,and other three sets of dynamic temperature conditions were used for validating the models and parameters.The results showed that there was no significant difference between the growth rates of Salmonella spp.under the competition of high and low concentration of background microflora.The minimum growth temperature of Salmonella spp.estimated by one-step method was6.36℃,and the maximum growth concentration was 9.82 log CFU/g.Both growth of Salmonella spp.and background microflora were accurately predicted.The RMSE of 4 groups of isothermal verified experiments were between 0.31 log CFU/g and 0.87 log CFU/g,and the RMSE of 3 groups of dynamic verified experiments were between 0.47log CFU/g and 0.68 log CFU/g.The residual errors of the isothermal and dynamic validation obeyed Normal distribution and Laplace distribution,respectively.For each of them,nearly 71.3%and 74.3%of the residual errorswere within the range of-0.5 log CFU/g and 0.5 log CFU/g.（3）A two-strain cocktail of Salmonella spp.was inoculated into tuna meat.Isothermal studies were performed by submerging samples under hot water maintained at at 52.5℃,55.0℃,57.5℃,60.0℃,62.5℃.The D values at each temperature were determined and used to calculate the z value by log（D）=log（D₀）-T/z.Dynamic studies were conducted by submerging samples in a water bath with its temperature programmed to increase linearly from 20.0℃to 70.0℃at 1.01℃/min～1.13℃/min.A numerical method was used to calculate the parameters D₀ and z.The results showed that,the thermal inactivation curve of Salmonella spp.in tuna followed the traditional first-order kinetics under isothermal temperature conditions,the D value in the range of 52.5℃～62.5℃was between 0.11 min～16.67 min,and the z value was 4.52℃.The average z value of Salmonella spp.measured under dynamic heating conditions was5.62℃,and the average value of log（D₀）was12.12.However,the kinetic parameters（D₀ and z）determined by static and dynamic methods could be used to predict the survival of Salmonella spp.exposed to linear heating temperature profiles,with statistically equal accuracies.