"The elevated and dynamic thermal resistance of Salmonella on/in low-moisture foods is an emerging challenge for the food industry. Therefore, the overall goal of this study was to improve the validation process for low-moisture foods by providing new knowledge about the effects that product structure and water activity have on Salmonella thermal resistance in or on low-moisture foods. The specific research objectives were: (1) To quantify the effect of inoculation protocol on the thermal resistance of Salmonella Enteritidis PT 30 in fabricated low-moisture foods (almond, wheat, and date products), (2) To evaluate the effects of long-term storage on the survival and thermal resistance of Salmonella Enteritidis PT 30 on almonds, and (3) To develop Salmonella thermal inactivation models that account for the effects of product structure, temperature, and water activity (for almond, date, and wheat products). For pre- and post-fabrication protocols, samples were inoculated before and after product fabrication. Salmonella exhibited greater thermal resistance on almond and date products (almond meal, almond butter, and date paste) inoculated using the pre-fabrication method as compared to the post-fabrication method. However, the opposite was true for wheat products (meal and flour). Differences in the food product composition may have contributed to these findings. Based on these results, the pre-fabrication method was chosen for all further experiments in this dissertation. In the long-term storage study, Salmonella populations decreased by 3 0303log CFU/g after 103 weeks of storage. However, Salmonella thermal resistance did not significantly change during long-term storage. Primary (log-linear and Weibull) and secondary (Bigelow-type) inactivation models for Salmonella were fit to isothermal inactivation data from eight different products, accounting for product structure (kernels/pieces/meal/flour/butter/paste), temperature (70-90°C), and water activity (0.25-0.65 aw). Overall the log-linear model was the most-likely-correct model, and the Bigelow-type secondary models therefore were incorporated into the log-linear model. Among all products, Salmonella was most heat resistant in 0.25 aw almond meal (D80°C = 75.2 min), and least resistant in 0.65 aw date paste (D80°C = 0.7 min). Decreasing aw increased thermal resistance. Additionally, Salmonella thermal resistance was generally greater on fabricated than whole products. However, these differences were relatively small for wheat products. Salmonella resistance on fabricated wheat products actually was lower than on wheat kernels at 0.45 and 0.65 aw. Variability in some of these effects across products might be attributable to compositional factors (e.g., sugar or moisture content), temperature-induced shifts in sorption isotherms or physical properties, or variable effects of particle sizes and microenvironment within the fabricated products. Overall, the primary-secondary inactivation models fit the various data sets well (RMSE from 0.51 to 1.08 log) and therefore are potential tools to predict Salmonella thermal inactivation for these products. Ultimately, this dissertation shows that low-moisture process validation protocols should account for inoculation methods and specific product structures, both of which can significantly affect process outcomes."--Pages ii-iii.
Read
