EFFECTS OF PRODUCT STRUCTURE, TEMPERATURE, WATER ACTIVITY, AND STORAGE ON THE THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 IN LOW - MOISTURE FOODS By Pichamon Limcharoenchat A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Biosystems Engineering - Doctor of Philosophy 2018 ABSTRACT EFFECTS OF PRODUCT STRUCTURE, TEMPERATURE, WATER ACTIVITY, AND STORAGE ON THE THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 IN LOW - MOISTURE FOODS By Pichamon Limcharoenchat 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 a ctivity 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 o f 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 an d 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 composi tion 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 log 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 di fferent products, accounting for product structure (kernels/pieces/meal/flour/butter/paste), temperature (70 - 90°C) , and water activity (0.25 - 0.65 a w ). Overall the log - linear model was the most - likely - correct model, and the Bigelow - type secondary models the refore were incorporated into the log - linear model. Among all products, Salmonella was most heat resistant in 0.25 a w almond meal (D 80°C = 75.2 min), and least resistant in 0.65 a w date paste (D 80°C = 0.7 min). Decreasing a w 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 ke rnels at 0.45 and 0.65 a w . 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 o r physical properties, or variable effects o f 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 inactivatio n 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. iv ACKNOWLEDGEMENT S When I start my journey to pursue my degree, I never imagine d that I would meet many people who inspired, supported, and encouraged me through this long journey. Dr. Bradley Marks, my best advisor . I am so honored to have you as my advisor and mentor. I greatly appreciate how much you taught me, not only in academic knowledge , but how to live in a good life. Without your support and guidance, this dissertation would not have been completed. I woul d like to give my appreciation to my committee members: Dr. Kirk Dolan for helping me with all modeling and parameter estimation , Dr. Sanghyup Jeong for his valuable comments relevant to my work, and Dr. Elliot Ryser for microbiology knowledge. I learned a lot from your suggestions and advice. Special thanks to the undergrads , Sarah Buchholz , Renee Schwartz , and Justine Williams . I really appreciate your hard work for collecting the experiment data. I would also like to thank our lab manager s , Mike James and Nicole Hall, for their assistance on laboratory work during all experiments , and their comments/edits on my academic writing. As an international student, my language ability improved a lot with your help. The undergrads team who work ed hard to prepar e all materials in our lab and helped me plating bacteria . Thank you so much. The graduate students in our lab, Francisco, Ian, Quincy, Beatriz, Nur u l, Dani , Phil. Thanks for your help in the lab, your comments on my work, and your encouraging words. v To my friends, in US and Thailand, thanks for always standing by my side. Thanks for traveling across the world to meet me here (with all foods from Thailand ! ) . your support. Lastly, my beloved family, thanks for believing in me and gi ving me strength from the beginning to the end. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ .......................... x LIST OF FIGURES ................................ ................................ ................................ ..................... xiv 1 INTRODUCTION ................................ ................................ ................................ ................... 1 1.1 Problem Statement ................................ ................................ ................................ ........... 1 1.2 Research Goal, Objectives, and Hypotheses ................................ ................................ .... 4 2 LITERATURE REVIEW ................................ ................................ ................................ ........ 6 2.1 Low - Moisture Foods of Interest ................................ ................................ ....................... 6 2.1.1 Almonds ................................ ................................ ................................ .................... 6 2.1.2 Dates ................................ ................................ ................................ ......................... 7 2.1.3 Wheat ................................ ................................ ................................ ........................ 7 2.2 Salmonella ................................ ................................ ................................ ........................ 8 2.3 Salmonella Survival in a Low - Moisture System ................................ .............................. 8 2.4 Factors that impact Salmonella Thermal Resistance in Low - Moisture Foods ............... 10 2.4.1 Water a ctivity ................................ ................................ ................................ .......... 10 2.4.2 Temperature ................................ ................................ ................................ ............ 12 2.4.3 Inoculation method ................................ ................................ ................................ . 13 2.5 Changes in Physical and Chemical Properties during Fabrication and Heating ............ 15 2.5.1 Microenvironment ................................ ................................ ................................ ... 15 2.5.2 Thermal treatment ................................ ................................ ................................ ... 16 2.5.2.1 Calorimetry for heat transfer ................................ ................................ .................. 16 2.5.2.2 Water activity at high temperatures ................................ ................................ ....... 16 2.5.3 Primary model ................................ ................................ ................................ ......... 17 2.5.4 Se condary models ................................ ................................ ................................ ... 17 2.5.5 Model selection ................................ ................................ ................................ ....... 18 2.5.6 Modeling Salmonella inactivation in low - moisture foods ................................ ...... 19 2.6 Conclusion ................................ ................................ ................................ ...................... 21 vii 3 INOCULATION PROTOCOLS INFLUENCE THE THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 IN FABRICATED ALMOND, WHEAT, AND DATE PRODUCTS ................................ ................................ ................................ ................................ .. 22 3.1 Materials and Methods ................................ ................................ ................................ ... 22 3.1.1 Almond meal and almond butter ................................ ................................ ............. 23 3.1.2 Wheat meal and wheat flour ................................ ................................ ................... 23 3.1.3 Date paste ................................ ................................ ................................ ................ 23 3.1.4 Inocul ation and equilibration ................................ ................................ .................. 24 3.1.5 Equilibration ................................ ................................ ................................ ........... 26 3.1.6 Water activity measurement ................................ ................................ ................... 26 3.1.7 Thermal treatment ................................ ................................ ................................ ... 27 3.1. 8 Recovery and enumeration ................................ ................................ ..................... 27 3.1.9 Statistical analyses ................................ ................................ ................................ .. 29 3.2 Results and Discussion ................................ ................................ ................................ ... 31 3.2.1 Sample preparation and water activity control ................................ ....................... 31 3.2.2 Model selection ................................ ................................ ................................ ....... 32 3.2.3 Replication error ................................ ................................ ................................ ..... 36 3.2.4 Product effects ................................ ................................ ................................ ........ 36 3.2.5 Structure effects ................................ ................................ ................................ ...... 36 3.2. 6 Almond products ................................ ................................ ................................ ..... 37 3.2.7 Date products ................................ ................................ ................................ .......... 38 3.2.8 Wheat products ................................ ................................ ................................ ....... 39 3.3 Conclusion ................................ ................................ ................................ ...................... 40 4 SURVIVAL AN D THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 ON ALMONDS AFTER LONG - TERM STORAGE ................................ ................................ .. 41 4.1 Materials and Methods ................................ ................................ ................................ ... 41 4.1.1 Almond kernels ................................ ................................ ................................ ....... 41 4.1.2 Inoculation preparation ................................ ................................ ........................... 41 4.1.3 Almond inocualtion ................................ ................................ ................................ 42 4.1.4 Water activity equilibration ................................ ................................ .................... 42 4.1.5 Long - term storage ................................ ................................ ................................ ... 43 4.1.6 Thermal treatment ................................ ................................ ................................ ... 43 4.1.7 Enumeration ................................ ................................ ................................ ............ 44 4.1.8 Statistical analyses ................................ ................................ ................................ .. 44 viii 4. 2 Results and Discussion ................................ ................................ ................................ ... 46 4.2.1 Water activity and moisture content of stored almonds ................................ .......... 46 4.2.2 Survival of Salmonella Enteritidis PT 30 after storage at room temperature ......... 48 4.2.3 Reduction of Salmonella Enteritidis PT 30 during thermal come - up ..................... 49 4.2.4 Thermal resistance of Salmonella Enteritidis PT 30 heated at 80°C ...................... 50 4.3 Conclusion ................................ ................................ ................................ ...................... 53 5 EFFECTS OF PRODUCT STRUCTURE, TEMPERATURE, AND WATER ACTIVITY ON THE THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 ...................... 54 5.1 Materials and Methods ................................ ................................ ................................ ... 54 5.1.1 Wheat products ................................ ................................ ................................ ....... 54 5.1.2 Almond products ................................ ................................ ................................ ..... 55 5.1.3 Date products ................................ ................................ ................................ .......... 55 5.1.4 Inoculation ................................ ................................ ................................ .............. 56 5.1.5 Equilibration ................................ ................................ ................................ ........... 57 5.1.6 Wa ter activity measurement ................................ ................................ ................... 59 5.1.7 Water activity measurement at 80°C ................................ ................................ ...... 59 5.1.8 Differential scanning calorimetry ................................ ................................ ........... 59 5.1.9 Thermal treatment ................................ ................................ ................................ ... 60 5.1. 10 Recovery and enumeration ................................ ................................ ..................... 61 5.1.11 Statistical analyses of properties ................................ ................................ ............. 62 5.1.12 Generalized linear model for testing factors affecting Salmonella inactivation ..... 62 5.1.13 Primary models ................................ ................................ ................................ ....... 63 5.1.14 Secondary model ................................ ................................ ................................ ..... 63 5.1.15 Model performance and selection ................................ ................................ ........... 64 5.2 Results and discussion ................................ ................................ ................................ .... 67 5.2.1 Initial inoculation ................................ ................................ ................................ .... 67 5. 2.2 Generalized linear model (GLM) ................................ ................................ ............ 67 5.2.3 Primary models ................................ ................................ ................................ ....... 72 5.2.4 Secondary model ................................ ................................ ................................ ..... 83 5.2.4. 1 Reference conditions ................................ ................................ .............................. 83 5.2.4.2 Model evaluation ................................ ................................ ................................ ... 84 5.2.5 Water activity effects ................................ ................................ .............................. 95 5.2.6 Product type and structure effects ................................ ................................ ........... 98 ix 5.2.6.1 Wheat products ................................ ................................ ................................ .... 101 5.2.6.2 Almond products ................................ ................................ ................................ .. 102 5.2.6.3 Date products ................................ ................................ ................................ ....... 103 5.2.6.4 Comparison between similar product structures. ................................ ................. 104 5.3 Conclusion ................................ ................................ ................................ .................... 107 6 OVERALL CONCLUSIONS AND RECCOMMENDATIONS ................................ ........ 109 6.1 Other methodological/preliminary work ................................ ................................ ...... 109 6.2 Overall Conclusions ................................ ................................ ................................ ..... 109 6.3 Future Work ................................ ................................ ................................ ................. 111 APPENDICES ................................ ................................ ................................ ............................ 113 Survivor Data for the Inoculation Protocol Experiment (Chapter 3) ................ 114 Survivor Data for the Long - Term Storage Experiments (Chapter 4) ................ 120 Product Properties ................................ ................................ ............................. 122 Photographs of Experimental Work ................................ ................................ .. 123 Come - Up Time for Thermal Inactivation ................................ ......................... 125 Survivor Data for Water Activity, Product Structure, and Temperature Experiments (Chapter 5) ................................ ................................ ................................ .......... 126 Matlab Codes for the GLM Regression ................................ ............................ 178 Matlab Codes for Model Fitting ................................ ................................ ........ 179 Salmonella Population Reductions during Thermal Come - Up Time (Chapter 5) ................................ ................................ ................................ ........................ 183 Shape Factor in Weibull Model ................................ ................................ ........ 185 Scaled Sensitivity Coefficient for the log - linear/Bigelow - Type Model ........... 188 Differential Scanning Calorimetry ................................ ................................ .... 192 Effects of the Fabrication Process on the Water Properties in Almond Products not Subjected to Complete Equilibration. ................................ ................................ ................ 193 Effect of Almond Skin Integrity on Salmonella Thermal Resistance ............... 195 Effect s of Equilibration Protocol, Water Properties, and Product Structure on Salmonella Thermal Resistance on/in Almond Kernels, Almond Meal, and Almond Butter . 197 Effect of the Type of Inactivation Container Used on Salmonella Thermal Resistance ................................ ................................ ................................ ....................... 200 REFERENCES ................................ ................................ ................................ ........................... 202 x LIST OF TABLES Table 3.1 Salmonella population (± standard deviation) and water activity (± standard deviation) of almond meal, almond butter, date paste, wheat meal, and wheat flour subjected to pre - fabrication and post - fabrication inoculation protocols before heating ................................ ......... 32 Table 3.2 Standard errors of replications, D values (± standard error) determined by non - linear regression of the Salmonella Weibull parameters for the almond meal, almond butter, date paste, wheat meal, and wheat flour (~0.40 0.45 a w ) subjected to pre - fabrication and post - fabrication inoculation protocols. ......... 34 Table 4.1 The a w (± standard deviation), and Salmonella Enteritidis PT 30 survival (± standard deviation) for whole almonds aft er 0 (groups I and II), 7 (I), 15 (I), 2 7 ( I), 68 (I), 70 (II) and 103 (II) weeks of storage at room temperature (and prior to re - equilibration). ................................ ... 47 Table 4.2 The D Salmonella Enteritidis PT 30 survivor curves for whole almonds (~0.45 a w ) after 0 (groups I and II), 7 (I), 15 (I), 2 7 ( I), 68 (I), 70 (II ) and 103 (II) weeks storage at room temperature. ................................ .. 52 Table 5.1 Equilibration time for wheat, almond, and date products. ................................ ........... 58 Table 5.2 Experimental design for the thermal inactivation of Salmonella Enteritidis PT30 on almond, wheat, and date products at 0.25, 0.45, and 0.65 a w between 70 - 90 ° C. ......................... 61 Table 5.3 GLM regression fo r the effect of treatment on Salmonella inactivation (log CFU/g) in ................................ ................................ ................................ ............. 71 Table 5.4 GLM regression for the effect of treatment on Salmonella inactivation (log CFU/g) in ................................ ................................ ................................ ........... 71 Tab le 5.5 GLM regression for the effect of treatment on Salmonella inactivation (log CFU/g) in ................................ ................................ ................................ ................ 72 Table 5.6 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared errors (RMSE), and AIC c values for wheat kernels, meal, and flour. ................... 75 Table 5.7 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared errors (RMSE), and AIC c values for almond kernels, meal, and butter. ............... 78 xi Table 5.8 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared errors (RMSE), and AIC c values for date pieces and date paste. .......................... 81 Table 5.9 Parameter estimates (mean ± standard error) for the log - linear/Bigelow - type models (secondary models), relative error (%), root mean squared error (RMSE), and AIC c values. ..... 86 Table 5.10 Model bias for each product from the Bigelow - type models. Negative values indicate underprediction of actual lethality. ................................ ................................ ............................... 88 Table 5.11 Water acti vity values ( ± standard deviation) at 25 and 80°C for wheat, almond, and date products ................................ ................................ ................................ ................................ . 96 Table 5.12 Composition results for almo nd, wheats, and date. ................................ .................... 99 Table 5.13 Sugar profile for dates using high - performance liquid chromatography (HPLC). ..... 99 Table 5.14 DSC parameters ( ± standard deviation) for 0.25, 0.45, and 0.65 a w wheat flour, almond butter, and date paste. ................................ ................................ ................................ .................. 100 Table 5.15 Calculated D - values (± standard error) at 0.45 a w and 80°C using log - linear/Bigelow - type model. ................................ ................................ ................................ ................................ .. 106 Table A.1 Salmonella inactivation data during isothermal treatment (80 ° C) for almond meal (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. ................................ ........ 114 Table A.2 Salmonella inactivation data during isothermal treatmen t (80 ° C) for almond butter (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. ................................ ........ 115 Table A.3 Salmonella inactivation data during isothermal treatment (80 ° C) for wheat meal (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. ................................ ........ 117 Table A.4 Salmonella inactivation data during isothermal treatment (80 ° C) for wheat flour (~0.4 a w ) using pre - fabrication and post - fabrication inoculation pr otocols. ................................ ........ 118 Table A.5 Salmonella inactivation data during isothermal treatment (80 ° C) for date paste (~0.45 a w ) using pre - fabri cation and post - fabrication inoculation protocols. ................................ ........ 119 xii Table B.1 Salmonella inactivation data during isothermal treatment (80°C) for whole almonds after 0, 7, 15, 27, 68, 70, and 103 weeks of storage at room temperature. ................................ . 120 Table C.1 Size distribution for wheat meal and wheat flour. ................................ ..................... 122 Table C.2 Size distribution for almond meal . ................................ ................................ ............. 122 Table E.1 Come - up time ( ± standard deviation) for almond products. ................................ ...... 125 Table E.2 Come - up time ( ± standard deviation) for wheat products. ................................ ......... 125 Table E.3 Come - up time ( ± standard deviation) for date products. ................................ ............ 125 Table F.1 Salmonell a inactivation data for almond kernels. ................................ ...................... 126 Table F.2 Salmonella inactivation data for almond meal. ................................ .......................... 132 Table F.3 Salmonella inactivation data for almond butter. ................................ ......................... 138 Table F.4 Salmonella inactivation data for wheat kernels. ................................ ......................... 144 Table F.5 Salmonella inactivation data for wheat meal. ................................ ............................. 150 Table F.6 Salmonella inactivation data for wheat flour. ................................ ............................. 156 Table F.7 Salmonella inactivation data for date pieces. ................................ ............................. 162 Table F.8 Salmonella inactivation data for date paste. ................................ ............................... 168 Table I.1 Salmonella population ( ± standard deviation ) reduction during the thermal come - up time for almond products. ................................ ................................ ................................ ................... 183 Table I.2 Salmonella population ( ± standa rd deviation ) reduction during the thermal come - up time for wheat products. ................................ ................................ ................................ ...................... 184 Table I.3 Salmonella population ( ± standard devia tion ) reduction during the thermal come - up time for date products. ................................ ................................ ................................ ........................ 184 xiii Table M.1 The a w and moisture content (± standard deviation) of almond kernels, almond meal, and almond butter fabricated from incompletely equilibrated almonds. ................................ .... 194 Table O.1 Definition of partial and full equilibration of almond kernels, almond meal, and almond butter. ................................ ................................ ................................ ................................ .......... 197 Table O.2 The a w and moisture content (± standard deviation) before heating of almond kernels, almond meal, and almond butter after partial and full equilibration. ................................ ......... 198 Table O.3 D values (± standard deviation) determined by linear regression of the Salmonella survivor curves (Figure C.1) of the almond kernels, almond me al, and almond butter after partial and full equilibration (~0.25 a w ). ................................ ................................ ................................ 199 Table P.1 Inactivation container loading for almond k ernels, almond kernels, almond butter, wheat kernels, date pieces and date paste. ................................ ................................ ............................. 200 xiv LIST OF FIGURES Figure 3.1 Inoculation steps for pre - and post - fabrication protocols. ................................ ........... 28 Figure 3.2 Isothermal (80 °C) Salmonella survival curves and log - linear model fit after pre - fabrication and post - fabrication inoculation of: (A) almond meal and almond butter at 0.40 a w , (B) date paste at 0.45 a w , and (C) wheat meal and wheat flour at 0.40 a w . ................................ ......... 35 Figure 4.1 Survival (log CFU/g) of Salmonella Enteritidis PT 30 (mean values of triplicates ± standard deviation) on whole al monds (~0.45 a w ) after 0 (I and II), 7 (I), 15 (I), 27 (I), 68 (I), 70 (II) and 103 (II) weeks of storage at room temperature, and after reaching the come - up temperature in thermal inactivation trial (~80 C). ................................ ................................ ............................ 50 Figure 4.2 The survival (log CFU/g) of Salmonella Enteritidis PT 30 (mean values of triplicates and log - linear model) during isothermal heating (~80 C) of whole almonds (~0.45 a w ) after 0 (I and II), 7 (I), 15 (I), 2 7 ( I), 68 (I), 70 (II) and 103 (II) weeks of storage at room temperature. ... 51 Figur e 5.1 Simulated temperature and a w profiles for: (A) almond and wheat products, and (B) date products. Solid line is simulated temperature, and dashed line is simulated a w . .......................... 66 Figure 5.2 Isothermal (80°C) Salmonella survival curves and log - linear model fit for: (A) almond kernels at 0.45 a w and three different temperatures (80, 85, and 90°C), (B) almond kernels at three different a w (0.25, 0.45, and 0.65) and at 80°C, and (C) almond kernels, wheat kernels, and date pieces at 0.45 a w and 80°C. ................................ ................................ ................................ ........... 69 Figure 5.3 Is othermal (80°C) Salmonella survival curves and log - linear model fit for: (A) wheat products, (B) almond products, and (C) date products at 0.45 a w . ................................ ............... 70 Figure 5.4 Example of reference conditions for the log - linear/Bigelow - type model of almond kernels. ................................ ................................ ................................ ................................ .......... 84 Figure 5.5 Observed and predicted log (N/N 0 ) for the log - linear/Bigelow - type model for: (A) wheat kernels, (B) wheat meal, and (C) wheat flour. ................................ ................................ ... 87 Figure 5.6 Observed and predicted log (N/N 0 ) for the log - linear/Bigelow - type model for: (A) almond kernels, (B) almond meal, and (C) almond butter. ................................ .......................... 87 xv Figure 5.7 Observed and predicted log (N/N 0 ) for the log - linear/Bigelow - type model for: (A) date pieces and (B) date paste. ................................ ................................ ................................ .............. 88 Figure 5.8 Relationship of D - value, estimated from log - linear (symbols) vs log - linear/Bigelow - type (line) models, and a w for wheat kernels, wheat meal, and wheat flour. ............................... 89 Figure 5.9 Relationship of D - value, estimated from log - linear (symbols) vs log - linear/Bigelow - type (line) models, and a w for almond kernels, almond meal, and almond butter. ....................... 90 Figure 5.10 Relationship of D - value, estimated from log - linear (symbols) vs log - linear/Bigelow - type (line) models, and a w for date pieces, and date paste. ................................ ........................... 91 Figure 5.11 Example of SSC for the log - linear/Bigelow - type model of almond kernels. ............ 93 Figure 5.12 Example of SSC for the log - linear/Bigelow - type model for: (A) almond kernels and (B) almond meal. ................................ ................................ ................................ .......................... 93 Figure 5.13. Relationship o f Z aw and Z T (°C) for all products. ................................ ..................... 95 Figure 5.14 Relationship of D 80°C , estimated from log - linear (dot) vs log - linear/Bigelow - type (line) models, with a w in similar product structure. ................................ ................................ .............. 105 Figure 5.15 Relationship between (A) Z T and pro duct structure, and (B) Z aw and product structure. ................................ ................................ ................................ ................................ ..................... 107 Figure D.1 Example of almond kernels condition ing in equilibration chamber. ........................ 123 Figure D.2 Custom - designed stirrer using for equilibrating almond butter. ............................... 123 Figure D.3 Water activity meter for measuring a w at high temperature. ................................ .... 124 Figur e D.4 Example of almond products in plastic bag and aluminum test - cell for thermal inactivation studies. ................................ ................................ ................................ ..................... 124 Figure F.1 Isothermal Salmonella survival curves and log - linear model fit for almond products at (A) constant a w (0.45 a w ) with three different temperatures (80, 85, and 90 ° C), and (B) constant temperature (80 ° C) with three different a w (0.25, 0.45, and 0. 65 a w ). ................................ ........ 174 xvi Figure F.2 Isothermal Salmonella survival curves and log - linear model fit for wheat products at (A) constant a w (0.45 a w ) with three different temperatures (80, 85, and 90 ° C), and (B) constant temperature (80 ° C) with three different a w (0.25, 0.45, and 0.65 a w ). ................................ ........ 175 Figure F.3 Isothermal Salmonella survival curves and log - linear model fit for date products at (A) constant a w (0.45 a w ) with three different temperatures (70, 75, and 80 ° C), and (B) constant temperature (80 ° C) with three dif ferent a w (0.25, 0.45, and 0.65 a w ). ................................ ........ 176 Figure F.4 Isothermal Salmonella survival curves and log - linear model fit of (A) almond kernels, wheat kernels, and date pieces, (B) almond meal, wheat meal, and wheat flour, and (C) almond butter and date paste at constant a w (0.45 a w ) and temperature (80 ° C). ................................ ..... 177 Figure J.1 Relationship of Weibull shape factor with (A) temperature (B) a w for almond kernels, almond meal, and almond butter. ................................ ................................ ................................ 185 Figure J.2 Relationship of Weibull shape factor with (A) temperature (B) a w for wheat kernels, wheat meal, and wheat flour. ................................ ................................ ................................ ...... 186 Figure J.3 Relationship of Weibull shape factor with (A) temperature (B) a w for date pieces and date paste. ................................ ................................ ................................ ................................ .... 187 Figure K.1 SSC for the log - linear/Bigelow - type model of almond kernels. .............................. 188 Figure K.2 SSC for the log - linear/Bigelow - type model of almond meal. ................................ .. 188 Figure K.3 SSC for the lo g - linear/Bigelow - type model of almond butter. ................................ 189 Figure K.4 SSC for the log - linear/Bigelow - type model of wheat kernels. ................................ . 189 Figure K.5 SSC for the log - linear/Bigelow - type model of wheat meal. ................................ .... 190 Figure K.6 SSC for the log - linear/Bigelow - type model of wheat flour. ................................ .... 190 Figure K.7 SSC for the log - linear/Bigelow - type model of date pieces. ................................ ..... 191 Figure K.8 SS C for the log - linear/Bigelow - type model of date paste. ................................ ....... 191 xvii Figure L.1 DSC thermogram of (A) almond butter (0.25, 0.45, and 0.65 a w ), (B) wheat flour (0.25, 0.45, and 0.65 a w ), and date paste (0.25, 0.45, and 0.65 a w ). ................................ ...................... 192 Figure N.1 Survival (log CFU/g) o f Salmonella Enteritidis PT30 during isothermal heating (~80 C) of whole, skin - damaged, and blanched almonds (~0.40 a w ). ................................ ........ 196 Figure O.1 Survival ( log CFU/g) of Salmonella Enteritidis PT30 during isothermal heating (~80°C) of the almond kernels, meal, and butter after partial and full equilibration (~0.25 a w ). 199 Figure P.1 Survival ( log CFU/g) of Salmonella Enteritidis PT30 during isothermal heating (~80°C) of (A) almond products (~0.25 a w ), (B) wheat kernels (~0.45 a w ), and (C) date products (~0.45 a w ). ................................ ................................ ................................ ................................ ..................... 201 1 1 I NTRODUCTION 1.1 Problem Statement Outbreaks of salmonellosis and recalls associated with low - moisture foods have increased in recent years. From 1996 to 2009, Salmonella cases increased by more than 20% in the United States , and the ability of outbreak detection by the PulseNet system (Pul sed - field gel electrophoresis: PFGE) increased illne ss es reported by almost 10% (Scharff et al., 2016) . The PulseNet system has helped improve the detec tion of outbreaks, but at the same time recalls have also increased during this period. In 2015, the U.S. Food and Drug Administration (FDA) reported that over 30 Salmonella - linked recalls were attributed to low - moisture food products (U.S. Food and Drug Administration, 2016a) . Additionally, low - moisture food products , such as poppy seeds (U.S. Food and Drug Administration, 2016b) and ginger powder (U.S. Food and Drug Admi nistration, 2017) , were recalled in 2016 and 2017 due to Salmonella contamination. The Center for Disease Control and Prevention (CDC) reported multistate outbreaks of Salmonella in sprouted nut butter spreads (Centers for Disease Control and Prevention, 2016a) and pistachios (Centers for Disease Control and Prevention, 2016b) in 2016. Two people from the pistachio outbreaks were hospitalized . Outbreaks of salmonellosis linked to low - moisture foods, including almonds, have occurred throughout the world (Isaacs et al., 2005) . P eanut butter (Centers for Disease Control and Prevention, 2007) , wheat flour (McCallum et al., 2013) , unsweetened cereal (Russo et al., 2013) , and chocolate (Werber et al., 2005) are all additional exampl es. Further, 75 people in New Zealand were infected by consuming contaminated raw flour from an uncooked baking mixture from October 2008 to January 2009 (McCallum et al., 2013) . Moreover, a variety of low - moisture foods , such as bleached flour, raw macadamia nuts, pistachios, almond butter, and ginger powder were 2 recalled due to contamination from Salmonella (U.S. Food and Drug Administration, 2014b, 2015, 2016; U.S. Food and Drug Administration, 2015, 2017) . Compared to other pathogenic organisms in fo od products, Salmonella in low - moisture foods is highly resistant to lethal treatments and able to survive long periods (Blessington et al., 2012; Ki mber et al., 2012) . A lso, this pathogen can remain viable in water for up to a week and in soil for over a year (Adams and Moss, 2008) . For example , Salmonella Mon tevideo survived on red winter wheat during 28 weeks of storage at a relative humidity of 13% (Crumrine and Foltz, 1969) . I n a date paste , Salmonella decreased during storage, but was still detected after 8 months at 4 °C (Beuchat and Mann, 2014) . Standard hygiene and sanitation practices are designed to prevent and control Salmonella from contaminating incoming raw materials and ingredients . Environmental monitoring and control also are important steps to minimize pathogens in food p roducts (Chen et al., 2009b) . For example, environmental contamination and substandard sanitation were the likely origin of a Sa lmonella outbreak traced to a peanut butter factory (Sheth et al., 2011; Viazis et al., 2015) . In the case of cereal products linked to Salm onella Agona outbreaks, the pathogen was detected in environmental samples within the production facility (Russo et al., 2013) . via l oss of product in the market, recall costs to the manufacturer, lost productivity, and a decrease in sales. The U.S. Department - ERS) esti mated the medical costs due to salmonellosis at $3.7 billion per year (U.S. Department of Agriculture, 2014) . Salmonella was the leading cause of medical costs from foodb orne outbreaks in the United States in 2015 (News desk, 2015) , with an estimated cost of illness for Salmonella infection at $1,792 per case (Scharff et al., 2016) . Based on one estimate, Weise (2009) reported that the 2009 3 outbreak/recall due to Salmonella in peanut butter cost $1 billion, with the Kellogg Company a lone estimated to have lost $70 million in that recall, which was one of the largest food recalls in the history of the United States. Such recalls can also negatively affect customer confidence in food safety and therefore reduce sales of the products aff ected. Food p roduct characteristics, such as water content and water activity, are important factors that impact Salmonella surviva l [See Chapter 2 for detailed in discussion] . Salmonella thermal resistance in low - moisture food products, such as wheat flo ur and peanut paste, increases as water activity decrease s (Kataoka, 2014; Smith and Marks, 2015) . He et al. (2013) reported that Salmonella thermal resistance in peanut butter at 90°C was significantly reduced when the water activity increased. For different food products at the same water activity levels, thermal resistance also can be affected by chemical composition and the type of product. For example, Salmonella thermal resistance in all - purpose flour is significantly lower than in peanut butter at 0.45 a w (Syamaladevi et al., 2016a) ; however , slight changes of fat content in different peanut butter products did not affect the heat resistance of Salmonella (Kataoka, 2014) . Limited studies have shown that other factors can affect Salmonella thermal resistance in low - moisture foods, such as sodium chloride concentration, type of sugar, and fat content (D'Souza et al., 2012; Kataoka, 2014; Mattick et al., 2001; Shrestha and Nummer, 2016) . However, no known prior studies have evaluated the effect of varying physical structures of low - moisture foods on Salmonella thermal resistance. In contrast, there have be en several such studies with high - moisture foods (Mogollon et al., 2009 ; Tuntivanich et al., 2008; Velasquez et al., 2010) , which reported that physical structure did impact Sa lmonella thermal resistance in raw pork, beef, and turkey. Thermal resistance of Salmonella in whole - muscle beef and turkey was 50% greater than in ground products of equivalent composition at 55, 60, and 62.5°C (Mogollon et al., 2009; 4 Tuntivanich et al., 20 08) . Similarly, Salmonella thermal resistan ce in ground pork was 0.64 to 2.96 times lower than in whole muscle pork cooked at 55 to 63°C (Velasquez et al., 2010) . Although these were all high - moisture food systems, the results do suggest that product structure (given equivalent chemical composition, moisture content, and temperature) might also affect Salmo nella thermal resistance in other food materials, such as low - moisture foods. Currently, there is no known prior research regarding the effects of product structure (in combination with water activity and temperature) on Salmonella thermal resistance in low - moisture food products. Specifically, if product structure affects the thermal response of Salmonella in low - moisture foods, then this could have a significant impact on food safety, especially because product structure has not typically been consider ed as a contributing factor in inactivation models or process validations. 1.2 Research Goal, Objectives, and Hypothes e s The overall goal was to improve pasteurization validations for low - moisture foods by providing new knowledge about the effects of product structure and water activity on Salmonella thermal resistance in or on low - moisture foods (almond, date, and wheat products). The specific research objectives we re: 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 . 5 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 ) . The hypotheses of this research we re that: (1) Inoculation protocol s impact Salmonella thermal resistanc e on or in almond, date, and wheat products, (2) T hermal resistance of Salmonella on almond kernels does not change during long - term storage (up to 2 years) , (3) Salmonella thermal resistance on or in almond, date, and wheat products increases with decreasing water activity , regardless of product structure , and (4) Product structure of almond, date, and wheat products significantly affects the Salmonella thermal resistance. 6 2 L ITERATURE REVIEW Low - moisture foods come in a variety of categories , such as nuts, fruits, and wheat products . Among various factors previously assessed, three were found to have the greatest impact on Salmonella thermal resistance : w ater activity (Syamaladevi et al., 2016b) , inoculation protocol (Hildebrandt et al., 2016) , and temperature (Smith et al., 2016) . This review highlights previous literature that has conveyed basic information on low - moisture foods, and , more specifically, on product factors (water, physical, and chemical properties) that have the largest impact on both the survival and thermal inactivation of Salmonella in low - moisture systems. 2.1 Low - Moisture Foods of Interest To date , prior studies on Salmonella thermal resistance in low - moisture food s have typically include d only one specific food category such as peanut butter (Li et al., 2014a) , almonds (Abd et al., 2012) , or wheat flour (Syamaladevi et al., 2016a) , but have not encompassed comparisons on the basis of physical structure. In this dissertation, low - moisture foods (almond, date, and wheat produc ts ) were chosen to represent high - fat, high - sugar, and high - starch products , respectively, in thermal inactivation studies. 2.1.1 Almonds Almonds ( Prunus dulcis ) are an increasingly popular food in the United States. The California almond crop was valued at $ 5.2 billion during 2016 - 2017, with $4.4 billion exported in 2016 (Almond Board of California, 2017) . Almonds are harvested with mechanical tree shakers, shelled, sized, stored, and processed (Almond Board of California, 2018a) . In the United State s (California) , almond pasteurizat ion is required via an Agriculture Marketing Order , and the mandatory treatment criterion is a minimum 7 4 - log reduction of Salmonella on almonds (Federal Register/Vol. 72, No. 61/Friday, March 30, 2007/Rules and Regulations, Pages 15021 - 15036) (U.S. Department of Agriculture, 2007) . Pasteurization methods that have been approved by the FDA include oil roasting, dry roasting, blanching, stream processing, and propylene oxide (PPO) gas treatment (Almond Board of California, 2018b) . 2.1.2 Dates In the United States, dates ( Phoenix dactylifera ) are mainly produced in California and we re valued at $ 67 million in 2016 (U.S. Department of Agriculture, 2017) . During 2016 - 2017, the import value of fresh dates was $47 million and the United States also exported $52 million of fresh dates (U.S. Department of Agriculture, 2018) . Date s are harvested, cleaned, and sorted by si ze, skin condition, moisture content, and color (Riggs, 2015) . There are no requirements for date pasteurization. Dates usually are directly transported to open - air markets after processing in the Middle East and North Africa . In addition, fumigation is used to eliminate insect pests in the industry (Chao and Krueger, 2007) . 2.1.3 Wheat W heat ( Triticum ) is widely used in baked goods, such as cakes, flat breads, and cookies. In the United States, the U.S. produced and exported wheat to countries such as Japan, Mexico, and Nigeria (U.S. Wheat Associates, 2016) . The estimate wheat export value for 2017 was $896 million (U.S. Wheat Associates, 2018) . Wheat milling is a major value - added contributor to the food industry in the United States (North American Miller's Association, 2016) . Generally, the process of milling wheat includes 8 cleaning, separating, grinding, sieving, and bleaching. No requirement s exist for wheat pasteurization. In some cases, the wheat flour will be enriched w ith vitamins or other nutrients to improve its nutritional quality (North American Miller's Association, 2016) . 2.2 Salmonella Salmonella spp. is a G ram - negative , facultative rod - shaped bacterium, that can cause foodborne infections (Adams and Moss, 2008) . People infected with Salmon ella may experience fever, nausea, diarrhea, vomiting, abdominal pain, and, in severe cases, even death (U.S. Food and Drug Administration, 2014a) . In this dissertation, Salmonella Enteritidis (SE) phage type 30 ( PT 30 ) was use d for experiments. This strain was responsible for a large outbreak of salmonellosis associated with almonds that occurred in Canada during 2000 to 2001 (Isaacs et al., 2005) and was also associated with a salmonellosis outbreak linked to raw almonds in 2004 (Centers for Disea se Control and Prevention, 2004) . Thermal resistance of Salmonella is influenced by serovar (Doyle and Mazzotta, 2000; Santillana - Farakos et al., 2014a) and therefore using a single serovar/strain with a given study simplified the analyses of key treatment effects. 2.3 Salmonella Survival in a Low - Moisture System Salmonella can survive for long periods in dry locations and in low - moisture food products (Adams and Moss, 2008) . The persistence of Salmonella in dry environments can a ffect Salmonella control strategies. Even though the number of microorganisms might decline over time, in some cases the rate of reduction depends on multiple factors, such as product formulation, storage temperature, and the cleaning process (Chen et al., 2009a, b) . 9 Low - moisture foods, such as almonds, wheat flour, and peanut butter, do not support the growth of Salmonella ; however, contamination of low - moisture pr oducts can occur at multiple points during pre - and post - harvest - processing (Scott et al., 2009) . Contamination can also be caused by poor sanitation practices, poor equipment design, unsuitable maintenance procedures, and poor ingredient storage conditions (Scott et al., 2009) . As an example, wheat grain and flour from wheat mills were monitored for yeast, mold, and pathogens in b aseline testing between 2006 and 2007 in Queensland, Australia, where Salmonell a was detected in wheat that was contaminated with soil , stone , and other environmental contaminants (Eglezos, 2010) . In survival/storage studies, Salmonella population s on almonds declined 1.8 log CFU/g and 2.1 log CFU/g after 24 and 48 weeks of storage (23°C) , respectively (Abd et al., 2012; Uesugi et al., 2006) . Salmonella population s on in - shell pecan s also decreased by 2.49 log CFU/g after being stored at 21°C for 78 weeks (Beuchat and Mann, 2010) . At 25°C, the population of Salmo nella in hazelnuts, pecans, and pine nuts decreased by 1 log after 24, 34, and 52 weeks of storage , respectively (Santillana - Farakos et al., 2017) . Salmonella population s in date paste declined 2.08 log CFU/g after 242 days of storage at 4°C , and by < 1 log CFU/g after 84 days of storage ( 25°C ). In contrast , Salmonella population s in date paste homogenates with water actually increase d by 2.74 log when stored at 25°C for 2 days (Beuchat and Mann, 2014) . W ater activity (a w ) impact s the surviva l of Salmonella . In whey protein powder, Salmonella Montevideo and Salmonella Typhimurium survived better at 0.18 a w than at 0.54 a w after 6 months of storage (36°C) (Santillana - Farakos et al., 2014a) . Increasing the water activity in nut products (hazelnuts, pecans, and pine nuts) decreased the survival of Salmonella after 52 weeks of sto rage (25°C) (Santillana - Farakos et al., 20 17) . Additionally, peanut paste with a low water activity (0.3 0 a w ) led to gr eater Salmonella survival when compared to the same product at 0.6 0 a w after 12 10 months of storage at 20°C (Kataoka, 2014) . However, a 9% difference in fat content in peanut paste samples did not affect the surviv al of during long - term storage (Kataoka, 2014) . The change in a w and moisture content during storage of low - moisture foods has been reported in very few microbial studie s survived . Kimber et al. (2012) reported that the a w and moisture content (MC) of almonds in sealed plastic bag s fluctuated ( 4 - 6% MC and 0.3 0 - 0.6 0 a w ) during 7 months of storage at - 19, 4, and 24°C. The MC of pea nuts and pecan s also increased by 1.2% and 1.0%, respectively, when stored in sealed plastic bags at 4 °C (Brar et al., 2015) . When stored at ambient temperature in a sealed container for a full year, the MC of raw peanuts and pecans (Brar et al., 2015) and walnut kernels (Blessington et al., 2012) w as stable in a sealed container when stored for a full year (3.8% MC for peanuts, 2.6% MC for pecans, and 3.0% MC for walnut kernels). 2.4 Factors that impact Salmonella Thermal Resistance in Low - Moisture Foods Many factors affect Salmonella thermal resistance , including a w (He et al., 2011; He et al., 2013) , fat content (Kataoka , 2014) , and salt content (Shrestha and Nummer, 2016) . In this section, the selected factors in this dissertation (i.e. , water activity, temperature, and inoculation method) are discussed relative to the pathogen response during processing. 2.4.1 Water a ctivity By definition, the a w of a food product is the ratio between the vapor pressure of the food and vapor pressure of pure water (Barbosa - C novas, 2007) . It can be calculated by the following equation: (1) 11 where is the equilibrium partial vapor pressure in the system, is the partial equilibrium vapor pressure of pure liquid water, and T is the temperature at which the sample is measured. A w is the most important factor in controlling the growth of Salmonel la in food products. Salmonella does not grow at a w lower than 0.94 (Adams and Moss, 2008) . Scott et al. (2009) reported that the controlled a w in the industry was below 0.85 a w ; therefore, the prevention of Salmonella growth in low - moisture systems is typically based on controlled a w . A w also impacts Salmonella thermal resistance in low - moisture foods (Syamaladevi et al., 2016b) . When the water activity of a food matrix is reduced, Salmonella thermal resistance can increase greatly. Villa - Rojas et al. (2013) showed that the D 70°C for Salmonella Enteritidis PT 30 on almond kernels at 0.601 a w (15.5 min) was higher than the D 68°C at 0.946 a w (0.42 m in). He et al. (2013) reported increase d thermal resistance of Salmonella (Enteritidis, Typhimurium, and Tennessee) in peanut butter when the water activity decreased from 0.8 0 to 0.2 0 a w , when heated at 90 and 126°C. Salmonella (Agona, Montevideo, and Typhimurium) was more thermal ly resistant in inoculated whey protein powder at 0.18 a w than at 0.54 a w when both products were vacuum sealed and heated at 70°C for 48 h (Santillana - Farakos et al., 2014a) . Initial water activity (from 0.1 0 to 0.7 0 a w ) also impacted the viability of Salmonella cerevisiae i n wheat flour and skim milk powder during hot air treatment (150 and 200°C) (Laroche et al., 2005) . The d ifferent water activity levels (0.3 0 and 0.6 0 a w ) did a ppear to affect Salmonella thermal resistance in inoculated wheat flour samples. The thermal resistance of Salmonella in rapidly - desiccated flour (0.6 0 a w to 0.3 0 a w - hydrated flour (0.3 0 a w to 0.6 0 a w in 2.5 min) were similar when compared to the heat resistance in flour that was slowly equilibrated (4 - 6 days) to the same a w value (Smith and Mar ks, 2015) . Therefore, the speed of a w change did not impact Salmonella thermal resistance . 12 Few studies have report ed the impact that moisture content has on the heat tolerance of Salmonella . Beuchat and Mann (2011b) reported that Salmonella declined faster on pecan nutmeats when the initial MC was higher (2.8 vs. 11.2%) in hot air treatments (15 min at 90 and 120°C). Additionall y, MC correlated with the inactivation kinetics of Salmonella during a moist - air heating process and was reported to be a better parameter in calculating process validations (Garcés - Vega, 2017) . The relationship between a w and MC is described by moisture - sorption isotherm s . A dsorption isotherms generally yield a lower moisture content than desorption isotherms at a given water activity, possibly due to the food structure, type of food, or the process temperature (Okos et al., 2007) . The difference in the equilibrium moisture content between the adsorption and desorption isotherms at a given relative humidity or water acti (Okos et al., 2007) . This hysteres is pattern occurs in low - moisture foods such as almond kernels (Pahlevanzadeh and Yazdani, 2005) , wheat flour (Moreira et al., 2010) , and date s (Chukwu, 2010) . T herefore , the sorption stage of low - moisture foods may also need to be considered in thermal inactivation process es, since a w may not sufficiently describ e the water effect on thermal resistance (Garcés - Vega, 2017) . 2.4.2 Temperature Temperature is probably the most important parameter in thermal process validation in low - moisture foods (Chen et al., 2009b) . Numerous studies have repor t ed the effect of temperature on Salmonella inactivation in/on various of low - moisture products , such as Salmonella Typhimurium DT104 i n low - a w (high - sugar) broths (Mattick et al., 2001) , Salmonella cerevisiae on wheat flour and ski m milk powder (Laroche et al., 2005) , Salmonella cocktails in peanut butter 13 (He et al., 2013; Ma et al., 2009; Shachar and Yaron, 2006) , Salmonella cocktails on pecans (Beuchat and Mann, 2011a, b) , Salmonella Enteritidis PT 30 on almonds (Abd et al., 2012; Harris et al., 2012) , and Salmonella cocktails in dried fruits (Beuchat and Mann, 2014) . Various approaches to modeling this affect are described below in section 2.6. 2.4.3 Inoculation m ethod Previous reports of bacterial survival or inactivation in low - moisture foods are based on a range of inocu lation methods. Ideally, the inoculation methods should yield bacterial responses that reflect actual contamination and processing scenarios. For inoculum preparation, Salmonella strains have been grown in tryptic soy broth (TSB) (Danyluk et al., 2005; Ma et al., 2009; Smith and Marks, 2015) or brain heart infusion (BHI) broth (He et al., 2011; He et al., 2013) . Bacteria were harvested and re - suspended in peptone water (Laroche et al., 2005) , a binary water/glycerol solution (Smith and Marks, 2015) , or peanut oil (for peanut butter) (He et al., 2011; He et al., 2013) . The means of dispersing the inoculum in the food matrix was product dependent and included hand mixing (Syamaladevi et al., 2016a) , machine stomaching (Smith and Marks, 2015) , misting or transfer from sand (Beuchat and Mann, 2014) , a mortar for food powder (Laroche et al., 2005) , or a sterile wooden tongue depressor for nut butter products (Burnett et al., 2000; Ma et al., 2009) . Initial pathogen inoculation levels in samples have been highly variable, ranging from 4.5 to 9.0 log CFU/g for peanut butter (Kataoka, 2014; Keller et al., 2012; Ma et al., 2009; Shachar and Yaron, 2006) . The impact of certain aspects of inoculation protocols on Salmonella thermal resistance in low - moisture foods has been reported in few studies. Ma et al. (2009) found that Salmonella thermal resistance in peanut butter increased with increasing incubation time during inoculum 14 preparation. Similarly, Keller et al. (2012) reported that Salmonella growth procedures, including temperature and growth media, also impacted Salmonella thermal resistance in peanut butter. More recently, Hildebrandt et al. (2016) used five different methods for inoculating Salmonella into wheat flour , with significantly different survival kinetics obtained at the same water activity (0.45 a w ) and temperature (80°C) . Additionally, a mist - inoculation procedure was shown to result in lower Salmonella survival than did a sand - inoculation procedure for stored dried fruit (Beuchat and Mann, 2014) . Many studies have examined thermal resi stance of bacteria; however, very few have assessed the thermal resistance of the same Salmonella strains in the same product at the same a w and temperature , such that variability in inoculation methods can affect results and imp ede cross study comparison. For example, a t the same a w (~0.40 - 0.45) and temperature (90°C), the thermal resistance of Salmonella Tennessee in peanut butter, when inoculated by adding strains directly into the matrix (He et al., 2013) , was five times lower than in another s tudy (Li et al., 2014a) where the strains were suspended in peanut oil prior to introduction into the peanut butter. Consequently, increasing evidence suggests that the inoculum preparation and methodologies are likely key factors affecting thermal resistance and therefore any pr ocess validation relying on the resulting inactivation data or parameters. 15 2.5 Changes in Physical and Chemical Properties during Fabrication and Heating 2.5.1 Microenvironment Fabrication change s the product structure , potentially leading to different microenvironment s . F or example, fabricating almond into almond butter may form a two - phase system (oil and water) (Li et al., 2014b) . Salmonella may survive and exhibit different thermal resistance in each phase. Shachar and Yaron (2006) reported that Salmonella was less thermal ly resistant in water than in the oil phase , probably because t he high fat content protects Salmonella cells at high temperature . Li et al. (2014b) reported Salmonella survival and thermal resistance in peanut butter and nonfat dry milk powder mixture. Salmonella was inoculated in to peanut butter and milk powder before mixing. Salmonella population s in milk powder declined faster than in peanut butter (4 - log reduction) after 5 weeks of storage at 25 °C . Salmonella population s also had higher rate of reduction in milk powder as compa red to peanut butter after heating at 90 °C for 10 min (3 - and 5 - log reduction) (Li et al., 2014b) , indicating that the microenvironment impacted Salmonella behavior. Also, the attachment and adherence of cells on/in selected low - moisture product s after fabrication may have impact ed pathogen behavior and thermal processing (Gurtler et al., 2014) . These results indicate that the microenvironment around Salmonella (i.e., location, or attachment ) may be one reason for different thermal resistance s after product fabrication. 16 2.5.2 Thermal t reatment 2.5.2.1 Calorimetry for h eat t ransfer Differential scanning calorimetry (DSC) can be used to evaluate many of the thermally induced physical changes that take place , such as fat crystallization in edible oil (Tan and Man, 2000) , phase transitions of date palm (Zaitoon et al., 2016) , or wheat grain cooking (Jankowski and Rha, 1986) . During heating , the physical state of some products may change due to , for example, the denaturation of proteins. Amirshaghaghi et al . (2017) reported an irreversible denaturation of almond proteins after heating above 80°C. According to Jankowski and Rha (1986) , dry grain and starch show ed similar biphasic thermal transition s characterized by peak s at 64.5°C and 86°C. DSC also was used to determine phase transitions of dates for improving storage conditions to extend she lf - life (Zaitoon et al., 2016) . Glass transition temperature decreased as the moisture content of dates increased. Physical changes during thermal treatment may also have impact Salmonella thermal resistance. 2.5.2.2 Water a ctivity at h igh t emperatures Water activity plays an important role in the heating process. Syamaladevi et al. (2016a) reported that the relationship between a w and temperature varied widely among different low - moisture products. When the temperature increased from 25 to 80°C, the a w of wheat flour increased from 0.45 to 0.80 a w , but the peanut butter a w decreased from 0.45 to 0.04 a w . The D 80°C of Salmonella in wheat flour and peanut butter was 6.9 and 17 min, which correspond s with a w changes at high temperature. Therefore, these a w effects likely have an impact on the tested and reported thermal resistance of Salmonella . 17 2.5.3 Primary model Various mathematical models have been developed to describe microbial growth and inactivation processes. During thermal processing, the purpose of the inactivation model is to understand and predict the thermal resistance and survival of bacteria (McKellar and Lu, 2 004) . The first - order, log - linear model is a well - known primary model that describes first - order reaction kinetics in heat processing. Log - linear inactivation kinetics have been used to estimate D - values (time required for a log reduction), by the follo wing equation: ( 2 ) where N and N 0 are the bacterial populations (CFU/g) at times t and 0, respectively; t is the period of time of the isothermal treatment; and D(T) is the time (min) required to reduce the microbial population by 90% (1 - log reduction) at a specified temperature (McKellar and Lu, 2004) . However, some studies of thermal inactivation in low - moisture foods reported survival curves that did not follow log - linear kinetics (Abd et al., 2012; Ma et al., 2009) . In those cases, th e Weibull model has been applied as shown below: ( 3 ) where N and N 0 are the populations (CFU/g) at times t and 0, respectively; t is the time of the isothermal treatment; p is the shape factor, and factor (Peleg, 2006) . 2.5.4 Secondary models Secondary models have been developed to account for the effects of environmental factors such as temperature, pH, and product a w , on primary model parameters (Gaillard et al., 1998) . For 18 this research, product structure, water activity, and temperature are factors that may impact Salmonella thermal resistance; therefore, the secondary model will be applied to evaluate the combined effects of temperature, water activity, and product structure. However, product structure is not a continuous variable that can be applied within secondary models. The Bigelow - type model is a common secondar y model that has been used to describe the effects of temperature and water activity on the D - value. The Bigelow - type model is based on the model structure of Gaillard et al. (1998) and can be written as: ( 4 ) where D ref is the time required to reduce the microbial population by 90% (1 log reduction) at T=T ref and a w = a w,ref ; T is temperature (°C); T ref is the optimized reference temperature (°C); a w,ref is the optimized refe rence water activity (a w is between 0 to 1); z T and z aw are temperature (°C) and water activity changes required for increasing or decreasing the D - value by a log cycle. 2.5.5 Model selection Error measurements have been used to evaluate the appropriateness and effectiveness of models. Model errors can be described by the coefficient of determination (R 2 ) or, root mean squared error (RMSE): ( 5 ) ( 6 ) 19 where N predicted and N observe d are the bacterial populations (CFU/g) at predicted and observed times; N average is the average population (CFU/g) from time 0 to t; n is the number of observation points ; and m is the number of model parameters. Additionally, models applied to a single data set can be compared via the Corrected Akaike Information Criterion (AIC c ) (Motulsky and Christopoulos, 2004) : ( 7 ) where n is the number of data; SS is the sum of squared residuals; and K is the number of parameters plus 1. A lower AIC c indicates the more - likely - correct model. 2.5.6 Modeling Salmonella inactivation in low - moisture foods Models for thermal inactivation of bacteria in low - moisture foods have been developed for specific conditions, such as sucrose solution effects of a w on the th ermal inactivation of Listeria monocytogenes (Sanchez - Zapata et al., 2011) , the inactivation kinetics of Salmonella Enteritidis PT 30 on ground almond kernels under dry conditions (Villa - Rojas et al., 2013) , or the combined effects of temperature, pH, and water activity on heat resistance of Bacillus cereus spores (Gaillard et al., 1998) . A w is one of the most influential factors used in model development f or low - moisture food s . Smith and Marks (2015) assessed the thermal resistance of Salmonella Enteritidis PT 30 on wheat flour subjected to rapid a w changes (Smith and M arks, 2015) . In this study, the a w of the samples was rapidly decreased from 0.6 0 a w to 0.3 0 a w or increased from 0.3 0 a w to 0.6 0 a w . A log - linear model was used to estimate the parameters that described Salmonella inactivation. However, for each of the models (0.6 0 a w to 0.3 0 a w and 0.3 0 a w to 0.6 0 a w ), the R 2 values were low (0.42 20 and 0.73), respectively. The study by Smith et al. (2016) also support s the importance of water activity by evaluating thermal resistance of Salmonella Enteritidis PT 30 on whe at flour with primary (log - linear and Weibull type) and secondary model s (second - order response surface, modified Bigelow, and combined effects) . The log - linear and the modified Bigelow - type models were the best models , based on RMSE and AIC c values. Mattick et al. (2001) reported the heat tolerance of Salmonella serovars at 50 to 80°C and water activities of 0.65 to 0.90 a w using a Weibull model. Secondary inactivation models were evaluated by comparing regression coefficients and analyzing P values. However, the generated thermal inactivation models underpredict ed the thermal death rate in low - moisture foods , suggest ing that additional factors should be included. Villa - Rojas et al. (2013) used a polynomial secondary model to assess the effect of temperature and a w on Salmonella thermal inactiv ation. The first - order kinetics model had good correlation coefficients (0.82 to 0.92), but the Weibull model was better (0.93 to 0.99). Use of 2 = 0.927 and 0.8 18). Therefore, the thermal inactivation of Salmonella Enteritidis PT 30 on almond kernels could be described by the Weibull distribution model. Santillana - Farakos et al. (2013) evaluated the log - linear model, Geeraerd - tail model, Weibull model, Biphasic - linear model, and Baranyi model by using the F - value, the root mean squared error (RMSE), and the adjusted coefficient of determination ( ). The Weibull model best described Salmonella survival kinetics in low - moisture foods. Additionally, Santillana - Farakos et al. (2013) developed a secondary model for predicting the effects of a w , water mobility , and temperature on the survival of Salmonella in whe y protein powder . This secondary model was 21 also evaluate d for other low - moisture foods, such as cereal, nuts, and peanut butter (Santillana - Farakos et al., 2014b) . Li et al. (2014a) also used the Weibull model to a ss ess Salmonella thermal inactivation i n foods of modified composition - peanut butter and peanut butter spread. Results suggested that the effect of temperature can be described by a log - linear model, and the survival curves can be described by the Weibull model. 2.6 Conclusion The impact of the inoculation procedure on Salmonel la thermal resistance for fabricated products (such as powders and pastes ) has already been demonstrated, but has never been evaluated with differing inoculation steps (before and after fabrication process). Factors, such as water activity, product structu re, and temperature, have an impact on Salmonella thermal resistance in low - moisture foods, and must also be evaluated to understand bacterial behavior during thermal pasteurization processes. Based on the overall literature review, this dissertation repre sents the first study known to quantify and report how product structure a ffects the thermal resistance of Salmonella. Lastly, the evaluation of Salmonella thermal resistance during long - term storage periods will be important to confirm the relevance of th ermal inactivation parameters to real - world process validations. Ultimately, thermal inactivation models can be used to improve food safety validation methods for low - moisture foods. Moreover, the behavior of Salmonella in low - moisture foods, at different temperatures, product structures, and water activities, can be us ed to improve current inactivation processes and process validation methodologies. 22 3 I NOCULATION PROTOCOLS INFLUENCE THE THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 IN FABRICATED ALMOND, WHEAT, AND DATE PRODUCTS Inoculation methods represent ing two contamination scenarios were assessed . Surface contamination can occur before, after, or even during processing and fabrication of low - moisture products. This experiment was designed to quantify the effect of inoculation protocol (pre - and post - fabrication) on the thermal resistance of Salmonella Enteritidis PT 30 in fabricated low - moisture foods (almond, wheat, and date products). This chapter was accepted for publication by the Journal of Food Protection . 3.1 Materials and Methods Overall, the experimental design consisted of inoculating almond meal, almond butter, wheat meal, wheat flour, and date paste via two different inoculation protocols (pre - fabrication and post - fabrication). Thereafter, the thermal resistances of Salmonella in these samples were compared by performing isothermal heat treatments in triplicate. In general terms, the pre - fabrication protocols entailed inoculation of intact natur al products (i.e., whole almond kernels, wheat kernels, and date pieces), which would correspond to environmental, in - field, or preprocessing contamination, and then those products were processed to produce meal, flour, butter, or paste. In contrast, the p ost - fabrication protocols entailed inoculation of the fabricated products after they were already produced, which would correspond to an in - plant or postprocessing contamination event. All known prior thermal inactivation studies with fabricated low - moistu re products have been conducted using post - fabrication inoculation protocol s . 23 3.1.1 Almond meal and almond butter Almonds (Nonpareil, size 27/30, Select Harvest, Turlock, CA) were sourced from a retail supplier, vacuum - packed (350 g per bag), and stored at ~2.5 °C for up to a year. Almond meal and almond butter were fabricated using a food processor (model FP21, Hamilton Beach Brands, Inc., Glen Allen, VA). To produce almond meal, whole almonds (100 g) were ground at the lowest speed setting for 45 s and sieved t hrough US standard sieves no. 20 and 80 (W.S. Tyler, Inc., Mentor, OH), capturing the material between the two sieves as the meal. Almond butter was produced by similarly grinding 200 g of almonds for 15 min total, while adding dry ice pellets (~30 mL) eve ry 2 min to maintain product temperature below 40°C (confirmed via a handheld infrared thermometer, model 566, Fluke Corporation, Everett, WA). 3.1.2 Wheat meal and wheat flour Organic soft white whole wheat kernels ( Triticum aestivum , Eden Foods Inc., Clinton, MI) were stored in their original package at room t emperature (~20°C) for up to 6 months. Wheat meal and wheat flour were fabricated by milling whole wheat kernels (50 g) for 45 s in a coffee mill (model 501, Jura - Capresso Inc., Montvale, NJ). Fabricated wheat samples were sieved through US standard sieves no. 20, 80, and 200. Ground product passing through a no. 20 sieve, but not through a no. 80 sieve, was called wheat meal, whereas ground product passing through a no. 80 sieve, but not through a no. 200 sieve, was termed wheat flour. 3.1.3 Date paste Dates (medjool, jumbo) were purchased from a retail supplier (Nuts.com, Cranford, NJ) and stored in their original package at ~2.5 ° C for up to a year. Date paste was fabricated by feeding 24 dates through a meat grinder plate with holes 1 cm in diameter (model K5 - A, KitchenAid, Benton Harbor, MI). The resulting paste was then fed through the grinder two more times to ensure homogeneity, which was determined by sampling inoculated date paste and enumerating for Sal monella survivors in five subsamples per replication (~1 g each) . 3.1.4 Inoculation and equilibration The general inoculation preparation method was derived from the procedures of Danyluk et al. (Danyluk et al., 2005) . Salmonella enterica serovar Enteritidis PT 30, previously obtained from Dr. Linda Harris (University of California, Davis), was kept frozen at - 80 ° C in a c oncentrated culture containing 20% glycerol. The frozen culture was subjected to two successive 24 h (37 ° C) transfers in TSB (Difco, BD, Franklin Lakes, NJ) 17% (m/m) containing 0.6% yeast extract (Difco, BD). Thereafter, a plate (150 by 15 mm) of Trypticase soy agar (TSA; Difco, BD) conta ining 0.6% yeast extract (TSAYE) was spread for confluent growth and incubated for 24 h (37 ° C). For pre - fabrication inoculation of almond and wheat products, the lawn cultures were each harvested in 10 mL of 0.1% peptone water. Thereafter, 8 mL of the liquid suspension ( ~ 10 7.5 to 10 9 CFU/mL) was added directly to 100 g of either almond or wheat kernels and mixed in a s terile plastic bag for 1 min. These wet inoculated samples were placed on filter paper (P8, Fisher Scientific, Pittsburgh, PA) in an open plastic container, dried ( ~ 3 h) in a biosafety cabinet, and then placed in an equilibration chamber (described in the Equilibration section) until they reached the target a w (0.40 ± 0.02). After equilibration, the samples were processed into meal, flour, or butter and were re - equilibrated as described below. 25 For post - fabrication inoculation of almond and wheat sample s, the Salmonella inocula (8 mL, grown and harvested as described above) were pelleted by centrifugation (model Sorvall RC 6 plus, SS - 34 rotor, Thermo Fisher Scientific, Waltham, MA) at 2,988 × g for 15 min. To minimize the change in a w during inoculation (and to prevent physical changes caused by the addition of water to the meals and powder), the Salmonella pellet was introduced into 50 g of almond meal, almond butter, wheat meal, or wheat flour and hand - mixed for 3 min in a sterile 24 - oz (710 - mL) plastic bag (Nasco, Fort Atkinson, WI). Inoculated samples then were equilibrated (as described below) until they reached the target a w (0.40 ± 0.02). In the pre - and post - fabrication protocols, date samples were inoculated using cell pellets that were produced by the same method as the post - fabrication protocol (described above) for almond and wheat samples. Based on preliminary tests, the inoculum was nonhomogeneous distributed by di rectly introducing the pellet into the date paste, because the highly viscous or semisolid structure of the paste impeded uniform distribution of the solid pellet. Therefore, the pellets were resuspended in 2 mL of 0.1% peptone water and homogenized using a vortex (model G - 560, Scientific Industries Inc., Bohemia, NY). This highly concentrated suspension for inoculation contained ~ 10 11 CFU/mL. For pre - fabrication inoculation, whole dates were each cut into 12 pieces ( ~ 1.8 g each) for faster equilibration. E ach date piece was spot - inoculated (200 µ L of total inoculum across 12 pieces) on the date skin, dried for 20 min in a biosafety cabinet, and then conditioned to ~ 0.45 a w in an equilibration chamber (described below) for up to 1 week. Date paste was fabri cated by grinding the inoculated date pieces, as previously described. If the a w after grinding was not 0.45 26 ± 0.02, the paste was returned to the chamber and re - equilibrated to the target a w (0.45). However, if the number of days the product spent re - equi librating as paste exceeded the number of days spent originally equilibrating as inoculated pieces, the product was considered unusable and was discarded, in order to control the overall treatment for both the intact date s and paste. For post - fabrication inoculation, dates were passed once through the grinder (previously described), after which 600 µ L of the concentrated Salmonella suspension was added to 60 g of ground dates. The inoculated date paste was then passed through the grinder four more times to evenly distribute the inoculum prior to equilibration to 0.45 ± 0.02 a w in the equilibration chamber. 3.1.5 Equilibration Samples were placed in custom - designed equilibration chambers (Smith and Marks, 2015) to adjust and control the sample a w . Controlled - humidity air (± 0 . 2%) obtained by mixing air passed through a desiccant column (dry air) or a water column (wet air) was monitored and controlled by a humidity sensor (DHT 22 , Adafruit Industries, New York, NY) and a microcomputer. Batches of samples (~300 g of almonds, 100 g of wheat, and 50 g of dates) were equilibrated to 0.40 ± 0.02 (almonds and wheat) or 0.45 ± 0.02 (dates) a w . Total equilibration times were 6 - 9 days for the almond meal, wheat meal, and wheat flour, and 11 - 14 days for the almond butter and date paste. 3.1.6 Water activi ty measurement Water activity of representative samples (pulled after mixing the bulk inside the equilibration chamber) was measured daily using a water activity meter (AquaLab 3TE, Decagon Devices, Pullman, WA) to confirm that the target a w was reached. 27 3.1.7 Thermal treatment After equilibration to the target a w , samples (~0.7 g of almond meal, 1.2 g of almond butter, 0.6 g of wheat meal, 0.5 g of wheat flour, and 1.2 g of date paste) w ere loaded into sealed aluminum test cells (Chung et al., 2008) in the equilib ration chamber to prevent a w changes . Sample thickness in the aluminum test cells was less than 1 mm. Samples were heated in an isothermal water bath set at 80.5 ° C ( GP - 400, Neslab, Newington, NH) . Com e - up time for the product to reach the target temperature (79.5 ° C) was measured in six replicates for each sample type, using a test cell with a T - type thermocouple probe positioned at the geometric center of the sample, and was averaged for use in all fur ther experiments. After reaching the come - up time (2.0 ±0.1 min for almond meal, 2.8 ±0. 1 min for almond butter, 1.3 ±0.1 m in for wheat meal, 1.4 ±0.3 m in for wheat flour, and 2.5 ±0. 1 min for date paste) , the initial (time zero) sample was removed, and subsequent samples were pulled at pre - determined time points and immediate ly cooled in an ice bath to halt further bacterial inactivation . 3.1.8 Recovery and enumeration Samples were aseptically removed from the te st cells , diluted (1:10 dilution) in 0.1% peptone water , and homogenized by stomach ing for 3 min (Model 1381/471, NEU - TEC Group Inc., Farmingdale, NY). Serial dilutions in 0.1% peptone wat er were plated in duplicate on mTSA YE (TSAYE supplemented with 0.05% of ammonium ferric citrate and 0.03% of sodium thiosulfate pentahydrate ; Fisher Chemical, Fair Lawn, NJ ) , which was a non - selective differential medium. The plates were incubated for 48 h at 37°C prior to counting the black colonies as Salmonella . Preliminary tests with uninoculated samples yielded no such colonies for any of the materials used in this study. 28 Pre - fabrication protocol Post - fabrication protocol Figure 3 . 1 Inoculation steps for pre - and post - fabrication protocols. Re - equilibration Fabrication Equilibration Inoculation Samples Equilibration Inoculation Fabrication Samples Thermal treatment 29 3.1.9 Statistical analyses Initial Salmonella populations and initial a w values from the pre - fabrication and post - fabrication methods were compared using the paired t - test (Microsoft Excel 2013 software, Microsoft Inc . , Seattle, WA ). For the pre - fabrication method, a w and Salmonella populations on the initial inoculated samples (kernels/fruits) and final samples (meal/butter/paste/flour) also were compared via a paired t - test. Reproducibility for each product was determined by calculating the standard error of replication as follows: (8 ) where m is the number of data points over time for each survival curve, n is the number of replications for each observation point, and y is the Salmonella population (log CFU/g). After pooling all triplicate data ( Appendix A ) within each treatment , the inactivation model parameters were estimated using nlinfit (nonlinear regression routine in the statistical toolbox) in MATLAB (version R2016a, MathWorks Inc., Natick, MA ) for the log - linear and Weibull models. The log - linear model was estimated by the following equation: ( 9 ) where N and N 0 are the populations (CFU/g) at times t and 0, respectively , t is the time of the isothermal treatment (min), and D(T) is the time (min) required to reduce the microbial population by 90% at a specified temperature ( T , °C). 30 The Weibull model parameters were estimated , according to the following equation (Peleg, 2006) : ( 10 ) where p is the shape factor , and is the location factor ( min). The estimated time for a 1 log - reduction (min) in each sample was calculated by the following equation (van Boekel, 2002) : ( 11 ) where d is the number of decimal reductions (i.e., d =1 for a 1 log reduction) . The Corrected Akaike Information Criterion (AIC c ) (Motulsky and Christopou los, 2004) was calculated to select the most - likely - correct model, with t he lower AIC c indicating the more - likely - correct model: ( 12 ) where n is the number of data points ; SS is the sum of square s of residuals , and K is the number of parameters plus 1. The relative probability of each model being the correct model also was calculated as follows (Motulsky and Christopoulos, 2004) : (13) Model parameters for pre - and post - fabrication samples of each product were also compared using the paired t - test (Microsoft Excel 2013). 31 3.2 Results and Discussion 3.2.1 Sample preparation and water activity control For the pre - fabrication methods, Salmonella Enteritidis PT 30 populations on the products after fabrication (i.e., meal, butter, flour, paste) were not significantly different from the p opulations on the intact products prior to fabrication (i.e., almonds , wheat kernels, date pieces) ( P > 0.05). Additionally, the pre - and post - fabrication products had similar a w values ( P > 0.05). In a comparison of initial Salmonella Enteritidis PT 30 populations between the pre - and post - fabrication protocols before hea ting ( Table 3 . 1 ), initial populations in date paste were statistically equivalent for the pre - and post - fabrication methods ( P > 0.05, 7.6 to 7.7 log CFU per sample). Additionally, separate subsampling tests yielded good homogeneity for both date preparation methods ( ± 0.2 and ± 0.3 log CFU/g for pre - and post - fabrication , respectively). Salmonella populations for almond and wheat products in the pre - fabrication method were significantly lower ( P < 0.05) than those for the post - fabrication method ( Table 3 . 1 ), because the Salmonella Enteritidis PT 30 concentration in the pellet inoculum for post - fabrication was higher than in the liquid inoculum for pre - fabrication . For date paste, the initial pre - and post - fabrication populations of Salmonella Enteritidis PT 30 were similar ( P > 0.05) and were l ower than the other product types because the inoculum contained fewer cells. However, prior results have shown that initial inoculation level does not affect thermal resistance of Salmonella Enteritidis PT 30 in low - moisture products (Hildebrandt et al., 2016) ; therefore, comparisons of thermal resistance between pre - and post - fabrication sam ples should not be affected by these differences in initial population. 32 Table 3 . 1 Salmonella population (± standard deviation) and water activity (± standard deviation) of almond meal, almond butter, date paste, wheat meal, and wheat flour subjected to pre - fabrication and post - fabrication inoculation protocols before heating Products Salmonella population (log CFU/g) Water activity Pre - fabrication Protocol Post - fabricat ion Protocol Pre - fabrication Post - fabrication Almond meal 8.0 ± 0.3 A 9.2 ± 0.2 B 0.410 ± 0.014 A 0.393 ± 0.003 A Almond butter 7.7 ± 0.2 A 9.3 ± 0.3 B 0.414 ± 0.012 A 0.406 ± 0.004 A Date paste 7.7 ± 0.2 A 7.6 ± 0.2 A 0.450 ± 0.015 A 0.456 ± 0.019 A Wheat meal 8.8 ± 0.1 A 9.7 ± 0.1 B 0.406 ± 0.009 A 0.405 ± 0.005 A Wheat flour 9.0 ± 0.1 A 9.7 ± 0.1 B 0.392 ± 0.017 A 0.400 ± 0.012 A Within a row (and same measurement) , means with a common superscript letter were not significantly different ( = 0.05). 3.2.2 Model selection Model parameters ( Table 3 . 2 ) for the log - linear and Weibull models were estimated using Salmonella Enteritidis PT 30 survival data ( Figure 3 . 2 ). AIC c analysis ( Table 3 . 2 ) gave the most - likely - correct model for each product type. The Weibull mode l was more likely correct for pre - fabrication almond meal (% likelihood > 99.99%), pre - fabrication almond butter (% likelihood > 99.99%), post - fabrication almond butter (% likelihood > 99.99%), pre - fabrication wheat meal (% likelihood > 90%), pre - fabricati on wheat flour (% likelihood > 96%), and post - fabrication wheat flour (% likelihood > 84%). However, the log - linear model was more likely correct for post - 33 fabrication almond meal, pre - and post - fabrication date paste, and post - fabrication wheat meal (% likelihood, ~70 to 98%). Because the Weibull model was not the most - likely - correct model for all products and was dependent on product type and inoculation protocol, both the D 80°C value and the Weibull - estimated time for a 1 - log reduction w ere calculated and compared for all products ( Table 3 . 2 ). 34 Table 3 . 2 Standard error s of replications , D C values (± standard error ) determined by non - linear regression of the Salmonella survivor c (± standard error ) and p (± standard error ) Weibull parameters for the almond meal, almond butter, date paste, wheat meal, and wheat flour (~0.4 0 0.45 a w ) subjected to pre - fabrication and post - fabrication inoculation protocols . Products Standard error of replications (log CFU/g) Log - linear model Weibull model D - value (min) RMSE (log CFU/ g) AIC c (min) p RMSE (log CFU/ g) AIC c Estimated time for one - log reduction (min) Relative likelihood of log - linear over Weibull model (per AIC) Almond meal Pre - fabrication 0.33 49.8 ± 2.1 A 0.418 - 54.1 29.6 ± 4.5 A 0.61 ± 0.07 A 0.308 - 72.9 29.6 ± 5.3 A 0.0001 Post - fabrication 0.85 33.4 ± 1.7 B 0.72 9 - 18.6 34.1 ± 6.4 A 1.02 ± 0.15 B 0.740 - 14.4 34.1 ± 6.8 A 0.8870 Almond butter Pre - fabrication 0.90 42.9 ± 2.6 A 0.694 - 20. 7 8.5 ± 3.5 A 0.37 ± 0.06 A 0.390 - 57. 4 8.5 ± 3.0 A ~0.0000 Post - fabrication 0.49 18.3 ± 1.0 B 1.13 2 2.0 4.7 ± 1.5 A 0.57 ± 0.06 A 0.47 7 - 3 2.0 3.4 ± 0.9 A ~0.0000 Date paste Pre - fabrication 0.31 3.5 ± 0.5 A 0.32 2 - 72. 5 3.3 ± 0.6 A 1.11 ± 0.38 A 0.32 7 - 66.8 3.3 ± 0.4 A 0.9436 Post - fabrication 0.79 1.2 ± 0.1 B 0.696 - 20.5 1.1 ± 0.2 B 1.30 ± 0.37 A 0.699 - 18.8 1.4 ± 0.3 B 0.6995 Wheat meal Pre - fabrication 0.80 10.3 ± 0.3 A 0.42 2 - 59.7 5.8 ± 0.7 A 0.69 ± 0.05 A 0.27 9 - 64.0 5.8 ± 0.8 A 0.1043 Post - fabrication 0.33 19.5 ± 0.8 B 0.652 - 35.7 7.5 ± 1.6 A 0.60 ± 0.05 A 0.37 3 - 42. 3 7.5 ± 1.6 A 0.0367 Wheat flour Pre - fabrication 0.54 8.9 ± 0.4 A 0.61 9 - 36.4 5.1 ± 1.2 A 0.71 ± 0.09 A 0.52 4 - 28.6 5.1 ± 1.2 A 0.9802 Post - fabrication 0.74 15.1 ± 0.7 B 0.978 - 7. 6 4.7 ± 1.7 A 0.59 ± 0.08 A 0.726 - 10.9 4.7 ± 1.4 A 0.1572 Within a column (and within the same product) , means with common superscript letters were not significantly different ( = 0.05). 35 Figure 3 . 2 Isothermal (80°C) Salmonella survival curves and log - linear model fit after pre - fabrication and post - fabrication inoculation of: (A) almond meal and almond butter at 0.4 0 a w , (B) date paste at 0.45 a w , and (C) wheat meal and wheat flour at 0.4 0 a w . 36 3.2.3 Replication error Replication error s ( Table 3 . 2 ) for each product were calculated to q uantify consistency of the experiments. The highest standard error of replication (0.90 log CFU/g) was for pre - fabrication almond butter, which may have been affected by oil separation during the equilibration process. 3.2.4 Product effects Based on the pre - fabrication D 80°C values, S almonella Enteritidis PT 30 thermal resistance in almond products was approximately four times greater ( P < 0.05) than in wheat products, which was approximately three times greater ( P < 0.05) than in date products. For the post - fabrication results, the same general rank ordering was true ( P < 0.05), except for a smaller difference betwee n almond and wheat products. This observation is consistent with prior S almonella Enteritidis PT 30 studies, which have reported larger D - values for high fat products (e.g., D 83 ° C of 16 min for peanut butter (M a et al., 2009) as compared to a D 80 ° C of 5 min for wheat flour (Smith et al., 2016) . 3.2.5 Structure effects S almonella Enteri tidis PT 30 thermal resistance was significantly greater in almond butter than in almond meal ( P < 0.05) for the post - fabrication protocol. In addition, S almonella Enteritidis PT 30 thermal resistance in wheat meal was significantly ( P < 0.05) greater than in wheat flour for both inoculation protocols. Surface interactions between product particles and Salmonella cells during fabrication may have impacted S almonella Enteritidis PT 30 attachment differently in almond and wheat products ( due to significantly different composition between these products), resulting in different impacts on thermal resistance; however, the fundamental mechanisms causing these differences are not yet conclusively known. 37 3.2.6 Almond products The D 80°C for pre - fabric ation almond meal (49.8 min) was higher ( P < 0.05) than that for post - fabrication almond meal (33.4 min). Villa - Rojas et al. (2013) reported a much lower D 80°C of 1.63 min for almond meal at 0.6 0 a w compared to this study, which would be expected to be due to the differences in a w . Additionally, this may have been impacted by differences in inocul um preparation, in that the prior study used phosphate buffer as the liquid suspension. In almond butter, the pre - fabrication D 80°C for Salmonella Enteritidis PT 30 (42.9 min) was two times greater than for post - fabrication (18.3 min). During the milling process, almond oil was expressed, and bacteria were presumably forced into the oil droplets. It can be assumed that the internal shear f orce during hand mixing ( post - fabrication ) was much lower than for mechanical stomaching ( pre - fabrication ); therefore, the fraction of Salmonella Enteritidis PT 30 cells entrained in the oil phase likely increased during fabrication, leading to greater the rmal resistance in pre - as opposed to post - fabrication almond butter. This enhanced survival is supported by the published literature indicating that high fat content protects bacterial cells at high temperature (Shachar and Yaron, 2006) . Thermal resistance of Salmonella has been assessed in peanut butter, but not in almond butter . Based on the log - linear model, Ma et al. (2009) and He et al. (2011) and (2013) reported a D 83°C of Salmonella Tennessee i n regular peanut butter of 16 min at 0.45 a w , and a D 90°C for a Salmonella cocktail on regular and low - fat peanut butter of 3.5 and 2.6 min, respectively, at 0.4 0 a w . Therefore , Salmonella strain, temperature, and fat content can be assumed to affect thermal resistance of Salmonella in nut butter products during processing (He et al., 2011; Ma et al., 2009; Shac har and Yaron, 2006) . 38 The Weibull distribution also has been previously used to model Salmonella inactivation in peanut butter. Ma et al. (2009) and He et al. (2013) reported the Weibull parameters and estimated times for 1 log - reduction of 1.92 min at 83°C, and 6.62 min at 90°C. Weibull parameters from Li et al. (2014a) yield ed an estimated time for one log - reduction (80°C) of a Salmonella cocktail (Thompson , Newport , Typhimurium , Copenhagen, Montevideo, and Heidelberg ) in regular peanut butter (0.45 a w ) of 1.9 min, which was lower than in pre - fabrication ( 8.5 min) and in post - fabrication ( 3.4 min) almond butter in this study. 3.2.7 Date products Thermal resistance of Salmonella Enteritidis PT 30 in post - fabrication inoculated date paste was the lowest amongst all the products (D 80 ° C ~ 1.2 min). Salmonella cells o riginally inoculated onto the date surface (pre - fabrication protocol) were more thermally resistant than those inoculated directly into the date paste (post - fabrication protocol). In the pre - fabrication method, the inoculated dates were equilibrated before grinding and re - equilibrated after grinding, but the post - fabrication samples were equilibrated in paste form. This difference in equilibration procedures, necessitated by the different fabrication procedures, may partially explain the observed differences in Salmonella Enteritidis PT 30 thermal resistance. Date paste also has a very high sugar cont ent (~66%) (U.S. Department of Agriculture, 2016) . Although previous studies on Salmonella thermal resistance in date paste are lacking, Mattick et al. (2001) reported the Weibull parameters for high sugar content broths (0.65 a w ) at 80°C. The ir estimated time f or a 1 - log reduction of Salmonella Typhimurium was 3.6 min, which was higher than that for post - fabrication inoculated date paste (1.5 min) in this study, but on the same order of magnitude. 39 3.2.8 Wheat products Thermal resistance of Salmonella Enteritidis PT 30 in wheat meal and wheat flour showed an opposite result from the almond and date products, with resistance greater in post - as opposed to pre - fabrication samples. In the pre - fabrication protocol, wheat meal and flour particle surfaces tha t previously were internal in the intact wheat kernel would have been cross - contaminated from the inoculated external surfaces during grinding and handling. However, in the post - fabrication protocol, all wheat meal and flour particle surfaces had equal pro bability of being contaminated when the inoculum was added to the powders and mixed. This difference between the two protocols therefore may have influenced the extent of Salmonella Enteritidis PT 30 attachment to any given particle surface, which could ha ve affected thermal resistance in a manner that would have been different than in the almond products, given the significantly different compositions. According to Smith et al. (2016) , Salmonella Enteritidis PT 30, which was inoculated via a similar method as the present post - fabrication protocol, exhibited a D 80°C of 5.5 min in wheat flour at 0. 43 a w , which was lower than that for post - inoculation wheat flour at 0.4 a w (15.1 min). They also used commercial white wheat flour , which may have altered the heat resistance , due to differences in composition (i.e., lower lipids content) and particle - cell interactions (Smith et al., 2016) . Syamaladevi et al. (2016a) also assessed thermal inactivation of a Salmonella cocktail in wheat flour at 80°C (inoculated post - fabrication). At 0.45 a w , the D 80° C was 6.9 min, which was lower than for the post - fabrication method used in this study (15.1 min). The Syamaladevi et al. (2016a) experiment was similar to this study, except for the inoculum preparation. These results sup port the premise that inoculation procedures impact thermal resistance of Salmonella in wheat flour (Hildebrandt et al., 2016) . 40 3.3 Conclusion The results have shown that thermal resistance of S almonella Enteritidis PT 30 depends on t he inoculation protocol, product type, and product structure. In all known prior studies with fabricated products (e.g., peanut butter (He et al., 2013; Ma et al., 2009) , wheat flour (Hildebrandt et al., 2016; Smith and Marks, 2015) , and dried fruits (Beuchat and Mann, 2014) ), post - f abrication inoculation protocols were applied to inoculate products, determine inactivation kinetics, and validate the processes. This suggests that some published data may not accurately reflect actual scenarios where a raw material is contaminated and th en fabricated into an ingredient or finished product, which may influence thermal resistance. The se results also suggest that pre - fabrication contamination events may be of greater concern in process validation. Additional tests are being conducted to quan tify Salmonella thermal resistance in different product matrices at various a w levels and to model Salmonella behavior in a range of low - moisture foods. 41 4 S URVIVAL AND THERMAL RESISTANCE OF SALMONELLA ENTERITIDIS PT 30 ON ALMONDS AFTER LONG - TERM STORAGE S almonella in low - moisture foods can survive for long periods. However, the thermal resistance of Salmonella on almonds after long - term storage has been reported for only one thermal process (hot oil treatment). In this study, the effects of long - term storage on the survival and thermal resistance of Salmonella Enteritidis PT 30 on almonds were evaluated. 4.1 Materials and Methods 4.1.1 Almond kernels Nonpareil almond kernels ( size 27/30, Select Harvest, Turlock, CA ) were vacuum packaged (350 g/bag) and stored at ~2.5°C . 4.1.2 Inoculation preparation The Danyluk et al. (Danyluk et al., 2005) inoculation procedure was followed with slight modifications descr ibed below. Salmonella enterica serovar Enteritidis phage type 30 (obtained from Dr. Linda Harris, University of California, Davis) was stored at - 80 °C in Trypticase Soy Broth (TSB; Difco, BD , Franklin Lakes, NJ) supplemented with 20% (vol/vol) glycerol. The original culture was transferred to a tube of T SB containing with 0.6% yeast extract (TSYBE) (Difco, BD) for 24 h (37ºC) , transferred to another tube of TS Y B E and incubated for an additional 24 h (37ºC), and then transferred to a plate (150 by 15 mm) o f Trypticase Soy Agar (TS A ; Difco, BD) containing 0.6% yeast extract (TSAYE) to obtain confluent growth after 24 h (37ºC). The lawn culture was harvested using 10 ml of 0.1% peptone water (Buffered P eptone W ater; Difco, 42 BD) per lawn plate for 5 plates in totals, and the inoculum was collected in a sterile plastic bottle before inoculating the almond kernels. 4.1.3 Almond inocualtion Prior to inoculation, the refrigerated almond kernels were held at room temperature for 30 min. The almonds (500 g) were hand - mixed with 40 ml of the inoculum (~10 7.5 to 10 9 CFU/ml) in a sterile plastic bag for 1 min , removed and placed in a single layer on filter paper (P8, Fisher Scientific, Pittsburgh, PA) , and dried for ~3 h in a biosafety cabinet before being moved into a humidity - controlled equilibration chamber. 4.1.4 Water activity equilibration Custom - designed equilibration chambers were used to maintain the humidity conditions during equilibration of the inoculated almonds prior to long - term storag e and thermal treatment (Smith and Marks, 2015) . The humidity ( 45 ± 0 . 2%) was maintained by passing air through either a dry column ( desiccant beads) or wet column (DI water), monitoring the chamber with a humidity sensor (DHT 22, Adafruit Industries, New York, NY) , and controlling the mix via solenoid values controlled by a microcomputer (Arduino Mega 2560, Turin, Italy). In the chamber, the almonds were spread in a single layer on perforated metal shelves and equilibrated for ~7 days to 0.45 ± 0.02 a w , which was confirmed by a water activity meter (AquaLab 4 TE, Decagon Devices, Pullman, WA) . 43 4.1.5 Long - term storage I noculated almonds from the same batch were randomly separated into two groups (I and II) of 250 g each, placed into s teel cans (16 oz., Uline, Pleasant Prairie , WI), sealed with electrical tape (3M Co., Ltd, Two Harbors, MN ), and stored in an i n s u lated container at room temperature (23 ± 0 . 2°C ). Group I subsamples were removed at 0, 7, 15, 27, and 68 weeks to quantify Salmonella survival and thermal resistance ( described below). Group II subsamples were removed at 70 and 103 weeks for the same analyses. Each group consisted of samples from three different initial inoculations . For group I, after each storage period , a random subsample was removed from each replic ate to measure a w . If the a w was out of the target range for testing ( 0.45 ± 0.02 a w ) , the entire group I sample was unpacked from the storage container and placed in the equilibration chamber (5 - 7 days) until the target a w was achieved. Then, a subsample ( ~ 15 g/replicate) was randomly removed for the thermal inactivation test, and the remaining unused almonds were placed back into storage as described above. For group II, the almonds remained in the sealed steel cans, which were not open ed until weeks 70 and 103 . The group II samples were tested using the same methodology as group I, but they were not re - equilibrated in the chambers prior to thermal treatment. 4.1.6 T hermal treatment Single almond s were vacuum - pack ag ed as a thin layer ( < 1 mm ) in plastic bag s (4 oz., Nasco, Fort Atkinson, WI), with a total of 9 bags per replicate (1 bag for 1 experimental time - point in each treatment). Before performing the thermal inactivation experiments, the thermal come - up time was established by inserting t hin - wire thermocouple s ( T - type, 36 gauge, OMEGA Engineering Inc., Stanford, CT) underneath the skin of six replicates of individually vacuum - 44 packaged almonds, to determine the time for the almond surface temperature to reach within 0.5 °C of the 80 ° C target temperature in a water bath ( GP - 400, Neslab, Newington, NH) . For the experiments, the initial (time zero) samples were removed from the water bath after the come - up time (2.7 ± 0.4 min) had been reached. Subsequently, almonds were removed at 8 additional time points up to 96 min of heating , and the bags were immediately submerged in an ice bath for >1 min. 4.1.7 Enumeration The c ooled samp les were aseptically unpacked and diluted (1:10) in 0.1% peptone water, stomached for 3 min ( Model 1381/471, NEU - TEC Group Inc, Farmingdale, NY) , serially diluted, and plated on modified TSAYE supplemented with 0.05% of ammonium ferric citrate and 0.03% of sodium thiosulfate pentahydrate ( Fisher Chemical, Fair Lawn, NJ ). The Salmonella survivors (differentiated by black colonies) were enumerated after incubating for 48 h at 37°C. Salmonella s urvivor data points w ere omitted if the average count of the duplicate plates was not within 25 - 250 colonies (Tomasiewicz et al., 1980) . 4.1.8 Statistical analyses The Salmonella survival data from group s I and II were compared within each group. The variation in a w and Salmonella survival (log CFU/g) during storage was evaluated by analyses of variance (A NOVA ) with Tukey means comparison, using Minitab (version 18, Minitab Inc., State College , PA ). Survivor data at the different storage period s were also compared by using analysis of covariance (A NCOVA ) in MATLAB within a group. 45 Additionally, log - linear and Weibull models were fit to pooled triplicate survivor data ( Appendix B ) by using nlinfit (nonlinear regression routine in the statistical toolbox) in MATLAB (version R2016a, MathWorks Inc., Natick, MA ) . The log - linear model parameters were estimate d by using the following equation ; (1 4 ) where N and N 0 are the populations (CFU/g) at times t and 0, respectively; t is the time of the isothermal treatment (min); and D(T) is the time (min) required to reduce the microbial population by 90% at a specified temperature (T, °C). The Weibull model parameters were estimated according to the following equation (Peleg, 2006) : ( 15 ) where p is the shape factor , and is the location factor (min). The Weibull model parameters were also compared (for different storage time within each group) using the 95% confidence interval (CI) results. The Corrected Akaike Information Criterion (AIC c ) (Motulsky and Christopoulos, 2004) was calculated to select the most - likely - correct model, with t he lower AIC c value indicating the more - likely - correct model: ( 16 ) 46 where n is the number of data points, SS is the sum of square s of residuals , and K is the number of parameters plus 1. The relative probability of each model being the correct model also was calculated as follows (Motulsky and Christopoulos, 2004) : ( 17 ) 4.2 Results and Discussion 4.2.1 Water activity and moisture content of stored almonds After 6 weeks of storage, the a w of some of the group I replicates ( Table 4 . 1 ) were lower than the initial range (0.45 ± 0.02 a w ), giving an average a w of 0.43 a w ( P < 0.05); therefore, the samples were re - equilibrated at 45% RH for 5 - 7 days, and the a w was measured again before performing any further thermal treatments. All of the group I sample replicates were re - equilibrated (0.45 ±0.02 a w ) prior to running any the rmal treatments. In contrast, the group II samples at 70 weeks were in the target a w range (0.45 ±0.02 ) and were not different ( P > 0.05) from the initial a w value (week 0). At week 103, the a w of the almonds ( ~ 0.471 a w ) was higher ( P < 0.05) than week 0 ( ~ 0.452 a w ), but not significantly different ( P > 0.05) to week 70 ( ~ 0.460 a w ). Therefore, the stored almonds from week 103 were thermally treated without re - equilibration, similar to the week 70 samples. 47 Table 4 . 1 The a w ( ± standard deviation ) , and Salmonella Enteritidis PT 30 survival ( ± standard deviation ) for whole almonds after 0 (groups I and II) , 7 (I) , 15 (I) , 2 7 ( I), 68 (I), 70 (II) and 103 (II) week s of storage at room temperature (and prior to re - equilibration) . Storage time (weeks) a w Salmonella survival (log CFU/g) 0 (I and II) 0.452 ± 0.00 5 A 8.5 ± 0.2 A 7 ( I ) 0.428 ± 0.002 B 8.5 ± 0.2 A 15 ( I ) 0.417 ± 0.001 B 8.5 ± 0.1 A 2 7 ( I ) 0.417 ± 0.003 B 7.8 ± 0.2 B 68 ( I ) 0.463 ± 0.003 A 6.2 ± 0.3 C 70 ( II ) 0.460 ± 0.002 A , B 7.3 ± 0.1 B 10 3 ( II ) 0.471 ± 0.002 B 6.2 ± 0.3 C Within a column (and within the same group) , means with same superscript were not significantly different ( = 0.05). Although the a w of the stored samples did change significantly in a few cases in this study, the change s were relatively small (< 0.04 a w ). In contrast, prior studies involving unsealed and sealed storage reported a w changes of 0.3 0 (Zhang et al ., 2017) to 0.4 0 a w (Keller et al., 2013) and 0.2 0 a w (Kimber et al., 2012) , respectively. T he t ype of containers and storage conditions (sealed or unsealed) clearly impacted a w changes during storage. In terms of moisture content, the two storage groups were not significantly different ( P > 0.05) at week 68(I) and 70(II) ( 3.9 and 3.8% MC, respectively ) , and the re - equilibration process di d not have an impact on the moisture content of the almonds. Brar et al . (2015) reported that the moisture content of raw peanuts and pecan kernels remained stable after 52 weeks of storage at 22°C in sealed container s , whereas Kimber et al. (2012) reported a slight change in moisture 48 content during 28 weeks of storage (±1% MC). As expected, the moisture content of the stored group II almonds at week 70 and 103 w as also stabl e at 3.8% ( P > 0.05) . 4.2.2 Survival of Salmonella Enteritidis PT 30 after storage at room temperature For the group I samples, Salmonella Enteritidis PT 30 populations ( Table 4 . 1 ) were stable until 15 weeks of storage , but then decreased after 27, and 68 weeks of storage . For the group II samples, the Salmonella Enteritidis PT 30 populations also decreased ( P < 0.05 ) after 70 and 103 weeks of storage. Salmonella Enteritidis PT 30 populations were higher for the group II samples at 70 weeks of storage ( P < 0.05) than for group I samples at 68 weeks of storage. A fter 68 weeks of storage , the Salmonella Enteritidis PT 30 populations in the group I samples decreased 2. 3 log CFU/g from initial counts , and at 70 weeks the group II samples decreased by 1.2 log CFU/g. In previous studies, the reduction of Salmonella Enteritidis PT 30 on almond kernels in sealed plastic ba gs (primary) and plastic tubs (secondary) (Abd et al., 2012) and sealed plastic bags (Kimber et al., 2012) were similar at 48 (2.1 log CFU/g; 23°C) and 50 weeks (2.3 log CFU/g; 24°C), respec tively. However, Uesugi et a l . (2006) reported a 3.4 log CFU/g reduction of Salmonella Enteritidis PT 30 on almond kernels after 68 weeks of storage at 23°C in sealed plastic bags, which was greater than for group I in the present study at 68 weeks (2 .3 log CFU/g reduction) and group II at 70 weeks (1.2 log CFU/g reduction), which were in sealed tin cans. In addition, Brar et al. (2015) reported that Salmonella cocktail populations on pecans decreased by 0.4 log CFU/g after 10 weeks of storage in sealed plastic bags (22°C) and in another study were 1 log reduction lower after 10 weeks of storage in controlled glass or plastic desiccator jars (0.57 a w ; 25°C ) (Santillana - Farakos et al., 2017) . These results indicate that storage conditions 49 and container types are factors that likely impact Salmonella survival during long - term storage (Abd et al., 2012; Brar et al., 2015; Kimber et al., 2012) . In addition, homogeneity of the Salmonella Enteritidis PT 30 population after long - term storage was tested at week 74. Ten almonds from each replicate (30 samples in total per group) were randomly pulled from both groups. The mean populations for the group I (5.8 ± 0.7, 6.4 ± 0.8, and 5.6 ±1.0 log CFU/g; means ± SD of three replicates ) w ere lower ( P < 0.05) than in the group II samples (6.1 ± 0.9, 6.8 ± 0.3, 7.0 ± 0.2 log CFU/g). The environmental condition of group I was modified several times during the re - equilibration process , leading to a difference in a w values between the two groups which may have affected Salmonella survival (Finn et al., 2013) . 4.2.3 Reduction of Salmonella Enteritidis PT 30 during thermal come - up Salmonella Enteritidis PT 30 populations decreased ( P < 0.05) after thermal come - up at week 0, 15, and 103 ( Figure 4 . 1 ); however, the reduction of Salmonella populations during thermal come - up did not cha nge ( P > 0.05) with increasing storage time (0.9 ± 0.4 log CFU/g). 50 Figure 4 . 1 S urvival (log CFU/g) of Salmonella Enteritidis PT 30 (mean values of triplicates ± standard deviation) on whole almonds (~0.45 a w ) after 0 (I and II), 7 (I), 15 (I), 27 (I), 68 (I), 70 (II) and 103 (II) weeks of storage at room temperature, and after reaching the come - up t emperature in thermal inactivation trial (~80 C). 4.2.4 Thermal resistance of Salmonella Enteritidis PT 30 heated at 80°C Salmonella Enteritidis PT 30 inactivation data ( Figure 4 . 2 ) w ere used to estimate parameters of the log - linear and Weibull models. Results ( Table 4 . 2 ) indicate t hat the log - linear model was the more - likely - correct model for 5 out of 7 data sets, but the relative likelihood was fairly low ( 5 4 8 1 %) ; therefore, both models are presented. However, the shape factor (p - value) at week 27 and 103 were not significantly different ( P > 0.05 ) than zero , indicating the Weibull model was not a good choice in these cases. 51 Figure 4 . 2 The survival (log CFU/g) of Salmonella Enteritidis PT 30 (mean values of triplicates and log - linear model) during isothermal heating (~80 C) of whole almonds (~0.45 a w ) after 0 (I and II), 7 (I), 15 (I), 2 7 ( I), 68 (I), 70 (II) and 103 (II) weeks of storage at room temperature. When t he slope of the Salmonella Enteritidis PT 30 survival data ( Figure 4 . 2 ) was compared using ANCOVA within the same group, thermal resistance of Salmonella Enteritidis PT 30 did not change for the group I samples during the entire storage period ( P > 0.05). In the group II samples, thermal resistance of Salmonella Enteritidis PT 30 was lower ( P < 0.05 ) in 70 week as compared to 0 week samples, but the 103 week samples were not different ( P > 0.05) compared to the 70 week samples. It should be noted that the raw Salmonella populations data were determined from single kernel s , which affects variability . The variances of group II individual kernel population data at 74 weeks were 0.2 to 0.9 log CFU/g , and the standard error of D 80 ° C at 103 weeks was ± 8.4 min (37.1% of D 80 ° C ). 52 Table 4 . 2 The D , and and p Weibull parameter values ( ± standard errors) from the Salmonella Enteritidis PT 30 survivor c urves for whole almonds (~0.45 a w ) after 0 (groups I and II) , 7 (I) , 15 (I) , 2 7 ( I), 68 (I), 70 (II) and 103 (II) week s storage at room temperature. Storage time (weeks) Log - linear model Weibull model D - value * (min) RMSE (log CFU/g) AIC c * (min) p * RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) 0 (I and II) 27.0 ± 4.0 A 0.77 - 4.9 9.9 ± 7.2 A 0.60 ± 0.17 A 0.72 - 5.8 0. 3 9 7 (I) 24.2 ± 4.2 A 0.58 - 14.3 11.8 ± 8.0 A 0.59 ± 0.22 B 0.55 - 1 4 . 0 0.5 4 15 (I) 22.4 ± 4.0 A 0.71 - 4.9 5.3 ± 4.6 A 0.48 ± 0.15 A , B 0.59 - 8.7 0. 1 3 27 (I) 26.1 ± 9.7 A 0.89 2.5 4.9 ± 10.1 A 0.44 ± 0.30 A , B, * * 0.85 3. 8 0. 6 6 68 (I) 20.9 ± 3.3 A 0.38 - 21.2 15.0 ± 6.2 A 0.71 ± 0.23 A, B 0.37 - 18. 8 0.77 70 (II) 13.5 ± 2.2 B 0.76 - 6.0 10.7 ± 6.2 A 0.85 ± 0.30 A 0.78 - 3. 2 0. 8 1 103 (II) 22.6 ± 8.4 A , B 0.92 1.4 5.8 ± 11.4 A 0.35 ± 0.30 A, * * 0.92 3.3 0.72 * Within a column (and within the same group), means with common superscript letters were not significantly different ( = 0.05). ** This value is not significantly different ( = 0.05) from zero. 53 When comparing the two groups, the thermal resistance of Salmonella Enteritidis PT 30 was higher ( P < 0.05) for group I sample s at 68 weeks compared to group II sample s at 70 weeks. In prior studies, Abd et al. ( 2012) reported that the thermal resistance of Salmonella Enteritidis PT 30 during oil roasting of almonds (121 ° C) did not change after 48 weeks of storage at 23 ° C . While the relative humidity during storage was < 40%, the moisture content and water activity of the samples were not monitored. However, the thermal resistance of Salmonella Enteritidis PT 30 on almonds after long - term storage at room temperature remained unchaged overall ( P > 0.05) in both studies . 4.3 Co nclusion T his study suggests that re - equilibrating almonds (group I) multiple times may have increased the rate of reduction of Salmonella populations during long - term storage. Overall, th e findings support the hypothesis that thermal resistance of Salmone lla on almonds does not change during storage, even after approximately two years. These results indicate that the validation of thermal pasteurization processes for almonds should not be affected by storage age of the almonds subjected to the process , which is important information for commercial operations. 54 5 EFFECTS OF PRODUCT STRUCTURE , TEMPERATURE, AND WATER ACTIVITY ON THE THERMAL RESIS T ANCE OF SALMONELLA ENTERITIDIS PT 30 Factors that have an impact on Salmonella thermal resistance in low - moi sture foods, such as temperature and a w , have never been compared for multiple product structures within the same type of product, such as almond meal and almond butter. In order to account for the effects that product structure, temperature, and a w have o n Salmonella thermal inactivation , multiple primary and secondary models were fit to inactivation data from almond, date, and wheat products. 5.1 Materials and Methods The experimental design consisted of almond, date, and wheat products that were inoculated with Salmonella, fabricated into different structural forms after equilibrat ion to 0.25, 0.45, and 0.65 a w , and iso thermally processed at three temperatures between 70 - 90°C . Salmonella thermal inactivation models then were developed from t he data. 5.1.1 Wheat products Organic soft white whole wheat kernels ( Triticum aestivum , Eden Foods Inc., Clinton, MI) were stored in paper bags at room temperature (~20°C) for up to a year. Wheat meal and wheat flour were also produced from these wheat kernels after inoculat ion (See inoculation below) and equilibrat ion (See equilibration below) to 0.25, 0.45, and 0.65 a w . The wheat meal and flour products were produced from the inoculated and equilibrated kernels using a coffee grinder (model BCG1110B, KitchenAid, Benton Harbor, MI) inside an equilibration chamber (describe below) at the corresponding a w setpoint , in order to prevent a w changes during grinding. W heat meal was produced by gr i nding the wheat kernels (50 g) for 25 s, with a pause every 10 s, to limit increases in product temperature . Wheat flour was produced using the same method as for wheat meal , but 55 was processed for 60 s of total time instead of 25 s . The size distribution for wheat meal and wheat flour were analyzed by using the American Society of Agricultural and Biological Engineers (ASABE) standard S319.2 method of determining and expressing fineness of feed materials by sieving ( Appendix C ; Table C . 1 ). 5.1.2 Almond products Almonds ( Nonpareil almonds, size 27/30 ) were sourced from a wholesale distributor ( Select Harvest, Turlock, CA) , vacuum - pac ked (350 g/bag) and stored at ~2.5°C for up to two years. Prior to fabricating almond meal and butter, the almond kernels were inoculated (See inoculation below) and equilibrated (See equilibration below). Once the almond kernels were at equilibrium (0.25, 0.45, or 0.65 a w ) , almond meal and butter w ere fabricated in a food processor (model FP21, Hamilton Beach Brands, Inc., Glen Allen, VA) that was also placed in an equilibration chamber to prevent a change in a w after fabrication. The almond kernels (100 g) were processed for 45 s at the lowest speed setting into almond meal. For almond butter, 200 g of almond kernels were ground similarly, but in 2 min time intervals, for a total of 16 min. Dry ice (~30 ml) was added every 2 min to maintain the product te mperature below 40°C , which was monitored with a handheld infrared thermometer (Fluke IR 566, Everett, Washington). The size distribution of almond meal ( Appendix C ; Table C . 2 ) was analyzed by Microtrac Laser light scattering (model S3500, Micotrac Inc, Montgomeryville, PA ) . 5.1.3 Date products Dates (Medjool, jumbo) were purchased from a retail supplier (Nuts.com, Cranford, NJ) and stored in plastic bags at ~2.5°C for up to two years. The whole dates were cut into pieces ( 10 56 x 10 x 0.5 mm ), each of which consisted of a 10 x 10 mm piece of date skin , which were called date pieces. To produce the date paste, whole dates were pitted, cut into smaller chunks, inoculated (See inoculation below), and equilibrated (See equilibration below) before processing three consecutive times through a meat grinder with holes 1 cm in diameter (Kitch enAid, model K5 - A, Benton Harbor, MI) to ensure date paste homogeneity. 5.1.4 Inoculation The inoculation procedures of Danyluk et al. (2005) were used to almond and wheat products. Salmonella enterica serovar Enteritidis PT 30, previously obtained from Dr. Linda Harris (University of California, Davis), was kept frozen ( - 80°C) as a concentrated culture in Trypticase Soy Broth (TSB; Difco, BD , Franklin Lakes, NJ) containing 20% (vol/vol) glycerol . One loopful (10 µl) of frozen culture was subjected to two successive 24 h (37ºC) transfers in 10 ml of 17% (m/m) TSB containi ng 0.6% yeast extract (Difco, BD) (TSBYE) . Thereafter, a 150 by 15 mm plate of Trypticase Soy Agar (TSA; Difco, BD) containing 0.6% yeast extract (TSAYE) was spread with 1 ml of inoculum to obtain confluent growth after 24 h incubat ion (37°C). For wheat an d almond products, the Salmonella lawn culture from the TSAYE plate was harvested in 10 ml of 0.1% peptone water using a L - shaped spreader. The resulting 8 ml Salmonella suspension (~10 7.5 to 10 9 CFU/ml) was added directly to 100 g of either wheat or almond kernels and mixed in a sterile plastic bag by hand for 1 min, placed on filter paper (P8, Fisher Scientific, Pittsburgh, PA) in an open plastic container, dried (~3 h) in a biosafety cabinet, a nd then placed in an equilibration chamber (See equilibration below). After reaching their prescribed target a w , the inoculated wheat and almond products were processed as described in their respective material sections (See almond products and wheat produ cts above). 57 The date products were inoculated differently d ue to the ir size. As described by Danyluk et al. (2005) , the Salmonella was grown in lawn plates ; however, rather than 10 ml, 20 ml of 0.1% peptone water was used for harvest ing with a L - shape spreader , after which the cell suspension was centrifuged (model Sorvall RC 6 plus, SS - 34 rotor, Thermo Fisher Scientific, MA) at 2,988 × g for 15 min. The resulting Salmonella pellet was r esuspended in 2 ml of 0.1% peptone water and subsequently homogenized using a vortex mixer (model G - 560, Scientific Industries Inc., Bohemia, NY) , which yielded a liquid suspension containing ~10 11 CFU/ml. For the date pieces, 50 µl was pipetted onto the o uter skin of each piece then placed into an equilibration chamber. Production of the date paste began by cutting whole dates into 12 equally sized pieces (~1.8 g each) for more homogenous equilibration. Each d ate piece was spot inoculated (200 µl total placed in an equilibration chamber before grinding as described in the material section (See date products above). 5.1.5 Equilibration The target water activities for all sample s were 0.25±0.02, 0.45±0.02, and 0.65±0.02 a w . Custom - designed equilibration chambers (S mith and Marks, 2015) were used to modify and control the a w of all samples . R elative humidity in the chambers was monitored and controlled by a humidity sensor (DHT 22, Adafruit Industries, New York, NY) and a microcomputer (Mega 2560, Arduino, Italy) , with humidity - controlled air (±0.2% R.H. of chamber setpoint) obtained by circulating air through a desiccant (dry air) or water column (moist air). All of the samples, except for a lmond butter, were spread in a thin (< 5 mm) or single layer on open - mesh metal trays (almond 58 kernels ; Appendix D; Figure D . 1 ) or filter paper trays (w heat kernels, wheat meal, wheat flour, almond meal, date pieces and date paste; date paste was shaped into a 15 mm diameter sphere to reduce equilibration time). The almond butter was placed into a 16 oz. tin can and continuously stirred with a stainless s teel rod , controlled by a motor ( Mini 12 V ., 60 rpm, h igh t orque g ear b ox e lectric m otor , Nextrox , Newark, DE ) ( Figure D . 2 ) , to mitigate oil - water separation in the butter. Equilibration times for each of the sample types were dependent on their respective adsorption/desorption characteristics ( Table 5 . 1 ). Table 5 . 1 Equilibration time for wheat, almond, and date products. Product Equilibration time Wheat kernels 5 - 7 days Wheat meal 1 - 3 days after fabrication Wheat flour 1 - 3 days after fabrication Almond kernels 5 - 7 days Almond meal 1 - 3 days after fabrication Almond butter 5 - 7 days after fabrication Date pieces 5 - 7 days Date paste 5 - 7 days after fabrication 59 5.1.6 Water activity measurement The a w of all samples w as measured using a a w meter (AquaLab 4TE, Decagon Devices, Pullman, WA) to confirm that the target a w was achieved. 5.1.7 Water activity measurement at 80°C The samples were equilibrated until they reached the target a w at 25 ° C (0.25, 0.45, and 0.65 a w ) , confirmed via the a w meter before heating. The equilibrated samples were then placed in a custom - designed high temperature a w meter ( Figure D . 3 ; Decagon Devices, Pullman, WA) , which was placed in a hot air oven (model 725F, Thermo Fisher Scientific, MA) that was set at 83°C. The a w and temperature were recorded every 10 s. Once the s ample temperature reached 79.5°C, the samples were held for 10 min (T < 80.5°C) in the oven, and the average a w value from the last 2 min was calculated ( the change in a w value < 0.01 during the 10 min holding time ) . The a w values reported at 25°C and 80°C were averages from duplicate experimental trials. 5.1.8 Differential scanning calorimetry Wheat flour, almond butter, and date paste were equilibrated at 0.25, 0.45, 0.65 a w (25 ° C) . After equilibration , the samples (10 mg for wheat flour, 15 mg for almond butter, and 20 mg for date paste) were placed in a sealed aluminum pan and heated at 0.5°C/min from 20°C to 100°C in a differential scanning calorimeter (DSC) (model 2000, TA instruments, New Castle, DE). The glass transition temperature (T g ) was assigned an inflection point based upon the transition temperature span. The characteristic temperature (peak temperature, T p ) and total heat of transition e heating curve , where 60 characteristic transition occurred . Samples were measured twice, and the results were calculated using Universal Analysis 2000 software (TA instruments). 5.1.9 Thermal treatment After equilibration to the target a w , almond kernels (1 kerne l , ~1.2 g), wheat kernels (7 kernel s , ~0.4 g), and date pieces (1 piece, ~0.9 g) were vacuum sealed in a single layer (< 1 mm) in plastic bags (4 oz., Nasco, Fort Atkinson, WI). The fabricated products (0.7 g of almond meal, 1.2 g of almond butter, 0.6 g o f wheat meal, 0.5 g of wheat flour, and 1.2 g of date paste) were loaded into aluminum test cells ( sample thickness < 1 mm) and then sealed (Chung et al., 2008) . All of the sample containers were packed in side the equilibration chambers to prevent any change in a w, which could occur if packaged in the non - humidity - controlled laboratory environment. For all isothermal treatments ( Table 5 . 2 ) , water baths ( GP - 400, Neslab, Newington, NH) were set 0.5 ° C above the target temperature (70, 75, and 80 ° C for date products, 80, 85, and 90 ° C for almond and wheat products ; Table 5 . 2 ). The come - up time was established by immersing a sample into the water bath and removing it when the temperature was 0.5 ° C below the target temperature. C ome - up times for each product type were computed from the average of six replicate samples at each of the target temperatures (See Appendix E ). After the samples had reached the come - up time, the initial (time zero) sample was removed and immediately coole d in an ice bath. The remaining samples for each trial were removed at pre - determined time points (9 points total for each trial) and cooled prior to microbial analysis . 61 Table 5 . 2 Experimental design for the thermal inactivation of Salmonella Enteritidis PT30 on almond, wheat, and date products at 0.25, 0.45, and 0.65 a w between 70 - 90 ° C. Sample Low T emperature Medium T emperature H igh T emperature 0.25 a w 0.45 a w 0.65 a w 0.25 a w 0.45 a w 0.65 a w 0.25 a w 0.45 a w 0.65 a w Almond K ernels 80 °C 85 °C 90 °C M eal B utter Wheat K ernels 80 °C 85 °C 90 °C Meal Flour Date P ieces 70 °C 75 °C 80 °C Paste 5.1.10 Recovery and enumeration After cooling , the samples were aseptically removed from their containers , before being dilut ed 1:10 dilution in 0.1% peptone water and homogeniz ed in a stomacher for 3 min (Model 1381/471, NEU - TEC Group Inc., Farmingdale, NY). From the initial dilution, multiple serial dilutions were prepared and dispensed onto mTSAYE (TSAYE supplemented with 0.05% of ammonium ferric citrate and 0.03% of sodium thiosulfate pentahydrate; Fisher Chemical, Fair Lawn, NJ) in duplicate. The plates were incubated for 48 h at 37°C prior to cou nting the Salmonella 62 colonies , which appeared black on this non - selective , differential medium . Testing of uninoculated samples with this medium revealed no Salmonella - like colonies (< 2 log CFU/g). 5.1.11 Statistical analyses of properties Water activity values at 25 and 80°C were compared using Analysis of Variance (ANOVA) The DSC parameters were 5.1.12 Generalized linear model for testing factors affecting Salmonella inactivation The effect s of product structure, temperature, and a w o n the Salmonella inactivation data (Appendix F ) were analyzed using the generalized linear model via MATLAB (version R2017b, MathWorks Inc., Natick, MA ; Appendix G ). Time, product structure, temperature, and a w were all evaluated. The main effects, two - way interactions, and three - way interactions of these variables were included in the model. log N = 0 1 ×t 2 × T 3 ×a w 4 × S 5 ×t× T + 6 ×t ×a w 7 ×t ×S + 8 × T×a w + 9 ×T×S + 1 0 × a w ×S 1 1 ×t×T×a w 1 2 ×t ×T×S + 13 ×t×a w ×S 14 ×T×a w ×S ( 18 ) where N is the population (CFU/g) at t , T , a w , and S , t is the time of the isothermal treatment (min) , T is the temperature of the isothermal treatment (°C), a w is the initial a w of the sample, and S is the product structure of sample. 63 5.1.13 Primary models Salmonella survival data (triplicates) within each treatment were used to estimate the log - linear and Weibull parameters using nlinfit (nonlinear regression routine in the statistical toolbox) in MATLAB ( Appendix H ) . The log - linear model was estimated by the followi ng equation: ( 19 ) where N and N 0 are the populations (CFU/g) at times t and 0, respectively, t is the time of the isothermal treatment (min), and D(T) is the time (min) required to reduce the microbial populations by 90% at a specified temperature ( T , °C). The Weibull model parameters were estimated , according to the following equation (Peleg, 2006) : ( 2 0 ) the location factor (min). 5.1.14 Secondary model A preferred secondary model was developed after evaluating the primary models by the root mean square error (RMSE) and Corrected Akaike Information Criterion (AIC c ) (See model performance and selection below). In this study, product structure was not applied within the secondary model because it was a discrete class variable, unlike the continuous variables of temperature and a w . 64 The Salmonella survivor date (log N/ N 0 ) were used to estimate parameters for a Bigelow - type model (Gaillard et al., 1998) with modifications to account for the effects of temperature and a w on the D - value: ( 21 ) where D ref is the time required to reduce the microbial populations by 90% (1 log reduction) at T = T ref and a w = a w,ref ; T is the temperature (°C); T ref is the optimized reference temperature (°C); a w,ref is the optimized reference for a w (a w is be tween 0 to 1); Z T and Z aw are the temperature (°C) and a w change s required to increas e or decreas e the D - value by 1 log. The reference temperature (T ref ) and reference a w (a w,ref ) were optimized to minimize the correlation between parameters for the smallest relative errors of estimated parameters (Dolan et al., 2013) . To estimate parameters in the secondary models, the reference a w was held constant wh ile the reference temperature was var ied between a minimum and maximum process temperature. The correlation coefficient between D ref , and Z T , w as plotted , which yielded an optimized reference temperature at the value with the smallest correlation coefficient . Next, the T ref was held constant at the optimized value, and the procedure above was repeated to determine an optimized a w , ref . This two - step optimization procedure for T ref and a w,ref was iterated two additional times to yield the final optimized reference conditions. 5.1.15 Model performance and selection Model p erformance was evaluated based on RMSE (Motulsky and Christopoulos, 2004) , the AIC c (Motulsky and Christopoulos, 2004) , and the scaled sensitivity coefficient (SSC) (Dolan 65 and Mishra, 2013) . Additionally, the estimated parameters were evaluated using the 95% confidence intervals (CI) and their relative error s. The RMSE was calculated using the following equation (Motulsky and Christopoulos, 2004) : ( 22 ) where N predicted and N observed are the predicted and observed Salmonella populations (CFU/g) at each time ; n is the number of observation points ; and m is the number of model parameters. The AIC c was used to select the most - likely - correct model for primary and secondary model evaluation. A lower AIC c value indicates the more - likely - correct model: ( 23 ) where n is the number of data points; SS is the sum of squares of residuals, and K is the number of parameters plus 1. The relative likelihood was also calculated for each treatment for selection of the correct model (Motulsky and Christopoulos, 2004) : ( 24 ) SSC was calculated to test the correlation between model parameters and the unique estimability of each parameter. L arge SSCs indicated low c orrelation of parameters (Beck and Arnold, 1997) : ( 25 ) 66 For the secondary models, simulat ed temperature and a w profiles ( Figure 5 . 1 ) were generated to represent arbitrary and varying experimental conditions across the range of temperature and a w values for all experiments (which were isothermal and iso - a w ); however , the estimated parameter cannot be calculate d without variance. Therefore, temperature and a w (increasing and decreasing , respectively ) were used to determi ne the variance of the simulated experiments. The temperature profile increased linear ly from 70 to 80°C for date products, and from 80 to 90°C for almond and wheat products. Similarly, the a w profile decreased linearly from 0.65 to 0.25. Figure 5 . 1 Simulated temperature and a w profile s for : (A) almond and wheat products, and (B) date products. Solid line is simulated temperature , and dashed line is simulated a w . 67 5.2 Results and discussion 5.2.1 Initial inoculation H omogeneity of the Salmonella populations were calculated for each product structure and type across all temperature s and a w value . The mean (± standard deviation) initial Salmonella population s for inoculated wheat kernels, meal, and f lour were 8.94 ± 0.23, 8.56 ± 0.31, and 8.49 ± 0.18 log CFU/g, respectively. For almond kernels, meal, and butter, the initial Salmonella populations were 8.41 ± 0.24, 8.19 ± 0.18, and 8.25 ± 0.40 log CFU/g, respectively. For the date pieces and paste , the mean Salmonella population s were 9.04 ± 0.41 and 8.06 ± 0.34 , respectively. After reaching the come - up temperature at time 0 , the Salmonella populations had decreased 0.02 3.41, 0.07 - 2.81, and 0.04 2.89 log CFU/g for wheat, almond, and date products, respectively ( Appendix I ) . The great est reduction occurred when the temperature and a w were at their maximum values . 5.2.2 Generalized linear model (GLM) The GLM w as deve loped using the Salmonella inactivation data ( Appendix F ) to determine the impact of time, temperature, product structure, and a w of wheat, almond, and date products on Salmonella survival , including the interaction between these parameters. Because samples were treated in sealed containers , temperature and a w remained static and did not change within any single experiment . Across all products , Salmonella populations decreased more rapidly when the temperature and a w were increased ( Figure 5 . 2 and Appendix F ). In addition, products with similar structure s (large particle kernels and pieces), but d ifference compositions, resulted in significantl y different 68 lethality ( Figure 5 . 2 C) . The Salmonella lethality rate was faster for date pieces than for wheat kernels and almond kernels. In comparison, the lethality of Salmonella was lower in peanut butter than in wheat flour when samples were equilibrated to the same water activity ( 0.45 a w ) and thermal ly treated at 80°C (Syamaladevi et al., 2016a) , which is consistent with the almond - wheat comparison in this study. Within the same product type, product structure influenced the Salmonella inactivation results ( Figure 5 . 3 ) for almond and date products. Salmonella lethality rates in fabricated products (i.e., meal, butter, and paste) were lower compared to whole products (non - fabrication). However, wheat product structure did not impact the lethality of Salmonella at 0.45 a w and T = 80°C. Based on the GLM regression ( Table 5 . 3 - 5.5 ), the interaction s of product structure with time , and the interaction of temperature and product structure with time, both had an effect on Salmonella inactivation in all product s ( P < 0 .05) . For a w , the interaction of a w and time had an effect ( P < 0.05) on Salmonella inactivation in wheat and almond products , but was not significant ( P > 0.05) for the date products. Salmonella inactivation in date products was impacted by the interaction of a w and structure. Overall, all of t hese results indicate that temperature, product structure, and a w impact ed Salmonella inactivation for all of the products. However, to further understand the nature of the relation between Salmonella inactivation and product structure, temperature, and a w for low - moisture foods, primary and secondary kinetic models were needed. 69 Figure 5 . 2 Isothermal (80°C) Salmonella survival curves and log - linear model fit for : (A) almond kernels at 0.45 a w and three different temperatures (80, 85, and 90°C) , (B) almond kernels at three different a w (0.25, 0.45, and 0.65) and at 80°C, and (C) almond kernel s, wheat kernels, and date pieces at 0.45 a w and 80°C. 70 Figure 5 . 3 Isothermal (80°C) Salmonella survival curves and log - linear model fit for : (A) wheat products, (B) almond products, and (C) date products at 0.45 a w . 71 Table 5 . 3 GLM regression for the effect of treatment on Salmonella inactivation (log CFU/g) in Source Estimate SE tStat P value t - 1.22 0.29 - 4.20 0.00 * T - 0.04 0.09 - 0.45 0.65 a w 6.08 15.37 0.40 0.69 Structure 3.86 3.57 1.08 0.28 t x T 0.02 0.00 4.42 0.00 * t x a w 9.32 0.82 11.43 0.00 * t x structure - 0.26 0.07 - 3.60 0.00 * T x a w - 0.08 0.18 - 0.45 0.65 T x structure - 0.05 0.04 - 1.10 0.27 a w x structure 1.40 6.97 0.20 0.84 t x T x a w - 0.12 0.01 - 12.04 0.00 * t x T x structure 0.00 0.00 3.83 0.00 * t x a w x structure 0.01 0.03 0.40 0.69 T x a w x structure - 0.03 0.08 - 0.39 0.70 * Significant term at . Table 5 . 4 GLM regression for the effect of treatment on Salmonella inactivation (log CFU/g) in Source Estimate SE tStat P value t - 0.23 0.13 - 1.76 0.08 T - 0.19 0.10 - 1.94 0.05 a w - 10.55 16.43 - 0.64 0.52 Structure - 3.76 4.09 - 0.92 0.36 t x T 0.00 0.00 1.75 0.08 t x a w 0.64 0.22 2.92 0.00 * t x structure 0.12 0.04 2.80 0.01 * T x a w 0.16 0.19 0.83 0.41 T x structure 0.07 0.05 1.42 0.16 a w x structure 8.70 7.85 1.11 0.27 t x T x a w - 0.01 0.00 - 3.12 0.00 * t x T x structure 0.00 0.00 - 2.85 0.00 * t x a w x structure - 0.01 0.01 - 1.02 0.31 T x a w x structure - 0.14 0.09 - 1.52 0.13 *Significant term at . 72 Table 5 . 5 GLM regression for the effect of treatment on Salmonella inactivation (log CFU/g) in Source Estimate SE tStat P value t 4.79 0.95 5.06 0.00 * T - 0.27 0.13 - 2.04 0.04 * a w - 38.25 20.61 - 1.86 0.06 Structure - 16.11 6.22 - 2.59 0.01 * t x T - 0.07 0.01 - 5.42 0.00 * t x a w 0.72 1.24 0.58 0.56 t x structure - 2.08 0.39 - 5.30 0.00 * T x a w 0.50 0.27 1.86 0.06 T x structure 0.21 0.08 2.64 0.01 * a w x structure 44.81 12.84 3.49 0.00 * t x T x a w - 0.01 0.02 - 0.48 0.63 t x T x structure 0.03 0.01 5.81 0.00 * t x a w x structure - 0.26 0.10 - 2.58 0.01 * T x a w x structure - 0.62 0.17 - 3.70 0.00 * *Significant term at . 5.2.3 Primary models Primary model parameters were estimated for each Salmonella inactivation data set (triplicate) for every combination of temperature, product structure and type, and a w for both the log - linear and Weibull models ( Table 5 . 6 - 5.9 ). Results indicated that the log - linear model was the more - likely - correct model for wheat (17/27), almond (18/27), and date (16/18) products. However, the % likelihood was fairly low (51 80% for wheat products, 52 80% for almond products, an d 61 86% for date products). Therefore, both parameters for both models were presented to compare Salmonella thermal resistance across all products ( Table 5 . 6 - 5.9 ). 73 Few studies have compared the log - linear and the Weibull models for isothermal treatment of low - moisture products . Villa - Rojas et al. (2013) reported that the Weibull model better predict ed Salmonella Enteritidis PT 30 inactivation in almond kernel flour. However, the primary models from Villa - Rojas et al. were analyzed using the R 2 value , wh ich will always favor the Weibull model over the simpler log - linear model. In the present study, the log - linear model was more - likely - correct ( AIC c likelihood ~65%) for almond meal at 0.65 a w and T = 8 0°C . For thermal inactivation of Salmonella in wheat flour , Smith et al. (2016) suggested that the log - linear model was more - likely - correct for 75 85°C and a w values of 0.310 0.700 based on AI C c . T he likelihood was ~66 - 76% for all a w at T = 80°C, which was similar to wheat flour in this study, except for 0.25 a w at 85 and 90°C. Santillana - Farakos et al. (2013) reported that the Weibull model better fit the Salmonella inactiva tion data for whey protein powder (0.19 0.43 a w , 21 80°C); however, their analyses also w ere based on adjust ed R 2 and RMSE values. As noted above, these performance measures will always favor the Weibull over log - linear models, which is why AIC c is an important comparison tool. Based on the AIC c results, the log - linear model generally was the more - likely - correct model, but not for all of the individual products. In develop ing the secondary model, the Weibull shape factor p was considered for each product type. The shape factor was between 0.60 1.69 for wheat products, 0.61 1.2 0 for almond products, and 0.61 1.43 for date products. Although, t he relationship between shape factor and temperature/a w ( Appendix J ) w as tested, no statistically sign ificantly relationship between shape factor and temperature was seen for any of the products. The relationship between a w and shape factor was also random (i.e., no trend) for almond and date products. However, the shape factor did increase with increasing a w in wheat products ( P < 0.05). 74 In addition, the shape factor p was compared to p =1 using a paired t - test for all products. The shape factors from the date products were not significantly different ( P > 0.05). For the wheat and date products, most of the shape factors were also not significantly different from 1 ( P > 0.05; 23/27 correct for wheat products and 25/27 correct for almond products). Therefore, this analysis additionally supports the choice of the log - linear model, given the absence of any systematic trends in the shape factor. Consequently, the log - linear model (p =1) was selected to further develop the secondary model in order to account for the effects of temperature and a w for all products . 75 Table 5 . 6 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c values for wheat kernels, meal, and flour. Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Wheat kernels 0.25 a w 80°C 20.1 ± 2.1 0.66 - 17.8 21.1 ± 8.1 1.03 ± 0.27 0.67 - 15.0 0. 80 85°C 10.3 ± 0.7 0.49 - 33.4 8.2 ± 2.6 0.88 ± 0.14 0.49 - 31.4 0.73 90°C 4.7 ± 0.4 0.64 - 16.9 3.0 ± 1.3 0.79 ± 0.16 0.63 - 16.0 0. 6 2 0.45 a w 80°C 10.2 ± 1.0 0.61 - 21.8 10.7 ± 3.8 1.03 ± 0.25 0.62 - 19.0 0. 80 85°C 3.3 ± 0.2 0.44 - 39.1 5.1 ± 0.8 1.38 ± 0.17 0.40 - 43.0 0. 1 3 90°C 1.2 ± 0.1 0.45 - 36.1 1.5 ± 0.3 1.16 ± 0.17 0.45 - 34.7 0. 6 7 0.65 a w 80°C 3.7 ± 0.2 0.67 - 17.0 6.2 ± 1.4 1.36 ± 0.21 0.63 - 18.3 0.34 85°C 1.4 ± 0.1 0.45 - 36.5 1.7 ± 0.3 1.15 ± 0.14 0.45 - 35.3 0.64 90°C 0.5 ± 0.0 0.56 - 26.4 0.4 ± 0.1 0.93 ± 0.15 0.57 - 24.0 0.77 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 76 Table 5 . 6 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c values for wheat kernels, meal, and flour Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Wheat meal 0.25 a w 80°C 33.5 ± 1.3 0.19 - 83.5 21.9 ± 2.9 0.75 ± 0.05 0.15 - 94.2 ~ 0.0 0 85°C 18.1 ± 1.5 0.46 - 36.7 10.4 ± 3.8 0.72 ± 0.13 0.44 - 37.7 0.3 8 90°C 5.4 ± 0.3 0.36 - 50.3 3.3 ± 0.8 0.76 ± 0.08 0.32 - 54.4 0.11 0.45 a w 80°C 13.5 ± 1.5 0.74 - 10.2 4.5 ± 2.8 0.60 ± 0.13 0.68 - 12.6 0.2 3 85°C 4.3 ± 0.5 0.78 - 8.1 2.0 ± 1.2 0.68 ± 0.16 0.75 - 8.4 0.47 90°C 1.0 ± 0.1 0.68 - 14.1 0.8 ± 0.4 0.89 ± 0.24 0.69 - 11.5 0.78 0.65 a w 80°C 3.8 ± 0.3 0.50 - 32.5 3.6 ± 1.1 0.96 ± 0.18 0.51 - 29.8 0. 80 85°C 1.3 ± 0.1 0.35 - 52.2 1.3 ± 0.3 1.01 ± 0.12 0.35 - 49.4 0. 80 90°C 0.5 ± 0.1 0.78 - 8.2 0.7 ± 0.2 1.69 ± 0.69 0.77 - 7.2 0.6 3 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 77 Table 5 . 6 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c values for wheat kernels, meal, and flour Products Log - l inear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Wheat flour 0.25 a w 80°C 37.0 ± 4.1 0.61 - 21.5 27.7 ± 12.8 0.83 ± 0.22 0.62 - 19.3 0.7 5 85°C 20.7 ± 1.6 0.47 - 31.0 8.4 ± 3.0 0.64 ± 0.09 0.39 - 38.7 0.02 90°C 6.4 ± 0.5 0.64 - 18.6 3.3 ± 1.5 0.72 ± 0.13 0.61 - 19.4 0.40 0.45 a w 80°C 11.6 ± 1.5 0.80 - 6.2 5.0 ± 3.4 0.65 ± 0.18 0.77 - 6.3 0. 50 85°C 3.5 ± 0.5 0.75 - 10.8 2.2 ± 1.4 0.75 ± 0.24 0.75 - 9.1 0. 70 90°C 0.9 ± 0.1 0.67 - 16.0 0.9 ± 0.4 1.03 ± 0.34 0.68 - 13.2 0.80 0.65 a w 80°C 3.3 ± 0.2 0.36 - 50.4 2.7 ± 0.6 0.89 ± 0.10 0.36 - 48.9 0.68 85°C 1.1 ± 0.1 0.37 - 49.3 1.5 ± 0.3 1.28 ± 0.20 0.36 - 49.2 0.51 90°C 0.4 ± 0.1 0.70 - 12.7 0.3 ± 0.2 0.87 ± 0.37 0.72 - 10.0 0. 80 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 78 Table 5 . 7 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c values for almond kernels, meal, and butter. Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Almond kernels 0.25 a w 80°C 17.6 ± 1.7 0.62 - 21.1 18.4 ± 0.3 1.03 ± 0.24 0.63 - 18.3 0. 8 0 85°C 10.1 ± 1.1 0.66 - 17.1 12.2 ± 4.2 1.15 ± 0.30 0.67 - 14.7 0.7 7 90°C 6.1 ± 0.7 0.67 - 16.2 3.2 ± 1.7 0.71 ± 0.17 0.64 - 16.4 0.4 7 0.45 a w 80°C 24.8 ± 3.1 0.89 - 1.3 11.9 ± 8.2 0.70 ± 0.19 0.87 - 0.4 0.6 1 85°C 11.9 ± 1.0 0.51 - 25.9 9.2 ± 3.2 0.86 ± 0.16 0.51 - 23.6 0.7 6 90°C 5.6 ± 0.7 0.71 - 12.2 3.3 ± 2.0 0.73 ± 0.21 0.70 - 10.9 0.65 0.65 a w 80°C 11.5 ± 1.5 0.80 - 6.4 6.0 ± 3.9 0.71 ± 0.20 0.79 - 5.7 0.58 85°C 3.1 ± 0.7 0.84 - 4.1 2.5 ± 1.7 0.82 ± 0.42 0.85 - 1.5 0.782 90°C 1.1 ± 0.1 0.71 - 12.2 1.4 ± 0.5 1.20 ± 0.37 0.72 - 9.8 0.77 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 79 Table 5 . 7 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Almond meal 0.25 a w 80°C 75.2 ± 4.3 0.23 - 74.3 67.6 ± 11.6 0.91 ± 0.12 0.23 - 72.1 0.7 5 85°C 42.3 ± 2.6 0.35 - 48.9 33.5 ± 8.5 0.86 ± 0.13 0.35 - 47.4 0.67 90°C 21.3 ± 1.3 0.46 - 37.3 12.6 ± 3.9 0.76 ± 0.11 0.43 - 38.9 0.3 2 0.45 a w 80°C 48.7 ± 3.7 0.41 - 39.7 35.8 ± 10.9 0.82 ± 0.14 0.40 - 38.5 0.64 85°C 23.4 ± 2.0 0.54 - 26.1 13.1 ± 5.3 0.73 ± 0.13 0.51 - 26.8 0.4 1 90°C 9.9 ± 0.6 0.39 - 46.0 5.9 ± 1.6 0.75 ± 0.10 0.36 - 48.6 0.2 2 0.65 a w 80°C 20.1 ± 0.9 0.29 - 60.0 16.4 ± 3.1 0.88 ± 0.09 0.28 - 58.8 0.6 6 85°C 7.4 ± 0.4 0.35 - 51.3 5.4 ± 1.2 0.83 ± 0.10 0.34 - 51.6 0.46 90°C 2.7 ± 0.1 0.14 - 99.9 2.8 ± 0.3 1.03 ± 0.08 0.15 - 97.4 0.7 8 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 80 Table 5 . 7 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Almond butter 0.25 a w 80°C 61.6 ± 5.2 0.42 - 42.1 62.3 ± 16.9 1.01 ± 0.21 0.43 - 39.3 0.80 85°C 36.0 ± 1.7 0.32 - 56.3 23.7 ± 5.0 0.79 ± 0.08 0.29 - 59.3 0.1 8 90°C 18.4 ± 0.7 0.33 - 54.3 13.3 ± 2.6 0.85 ± 0.08 0.32 - 55.6 0.3 5 0.45 a w 80°C 48.9 ± 7.2 0.81 - 6.0 19.8 ± 15.3 0.61 ± 0.20 0.79 - 5.8 0.5 3 85°C 23.6 ± 2.6 0.73 - 12.2 11.6 ± 6.8 0.69 ± 0.17 0.71 - 12.1 0.5 2 90°C 8.0 ± 0.7 0.59 - 23.2 6.5 ± 2.6 0.88 ± 0.20 0.60 - 20.8 0.77 0.65 a w 80°C 13.7 ± 0.8 0.42 - 41.8 10.1 ± 2.7 0.85 ± 0.11 0.41 - 41.0 0.6 0 85°C 4.7 ± 0.3 0.45 - 36.7 4.2 ± 1.1 0.93 ± 0.13 0.46 - 34.1 0.7 8 90°C 1.7 ± 0.1 0.42 - 39.9 1.1 ± 0.3 0.79 ± 0.10 0.40 - 41.5 0.30 Parameters were estimated only in each a w and temperature condition, and only compared within each row. 81 Table 5 . 8 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c values for date pieces and date paste. Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Date pieces 0.25 a w 80°C 8.2 ± 1.0 0.55 - 27.5 11.5 ± 2.7 1.42 ± 0.41 0.54 - 26.5 0.6 2 85°C 2.9 ± 0.3 0.57 - 23.1 4.5 ± 1.0 1.43 ± 0.29 0.54 - 24.0 0.3 9 90°C 1.1 ± 0.2 0.91 - 0.2 1.9 ± 0.6 1.71 ± 0.66 0.89 0.8 0.61 0.45 a w 80°C 5.4 ± 0.9 0.92 2.0 5.4 ± 3.2 1.00 ± 0.41 0.95 5.2 0.8 3 85°C 3.0 ± 0.7 1.11 9.6 3.5 ± 2.2 1.13 ± 0.57 1.14 12.6 0.8 2 90°C 1.2 ± 0.4 1.64 29.7 0.8 ± 1.1 0.73 ± 0.57 * 1.67 32.3 0.7 9 0.65 a w 80°C 5.9 ± 1.4 1.23 13.7 8.7 ± 5.0 1.41 ± 0.78 * 1.25 16.3 0.78 85°C 2.8 ± 0.4 0.93 1.7 1.8 ± 1.2 0.79 ± 0.26 0.93 3.6 0.7 3 90°C 1.0 ± 0.1 0.98 4.0 0.8 ± 0.5 0.87 ± 0.31 1.00 6.7 0. 80 Parameters were estimated only in each a w and temperature condition, and only compared within each row. * This value is not significantly different ( = 0.05) from zero. 82 Table 5 . 8 Parameter estimates (mean ± standard error) for the log - linear and Weibull models, root mean squared error s (RMSE), and AIC c Products Log - linear model Weibull model D - value (min) RMSE (log CFU/g) AIC c (min) p RMSE (log CFU/g) AIC c Relative likelihood of log - linear over Weibull model (per AIC c ) Date pieces 0.25 a w 80°C 32.6 ± 4.7 0.42 - 39.9 27.4 ± 9.8 0.82 ± 0.27 0.43 - 37.5 0.76 85°C 16.4 ± 2.4 0.44 - 39.8 9.6 ± 4.5 0.61 ± 0.19 0.42 - 39.9 0.48 90°C 5.2 ± 0.7 0.45 - 34.9 5.2 ± 1.6 1.01 ± 0.36 0.46 - 32.1 0.8 1 0.45 a w 80°C 14.5 ± 1.4 0.23 - 68.6 12.9 ± 2.5 0.83 ± 0.18 0.23 - 66.7 0.7 2 85°C 7.5 ± 1.4 0.46 - 36.4 7.8 ± 2.3 1.06 ± 0.48 0.47 - 33.6 0. 80 90°C 3.2 ± 1.5 0.63 - 18.2 3.2 ± 2.1 0.71 ± 0.75 * 0.64 - 15.5 0.79 0.65 a w 80°C 5.2 ± 0.6 0.54 - 27.2 3.8 ± 1.7 0.80 ± 0.21 0.54 - 25.3 0.7 2 85°C 1.8 ± 0.2 0.78 - 7.6 1.0 ± 0.6 0.75 ± 0.19 0.77 - 6.6 0.62 90°C 0.8 ± 0.2 0.90 3.3 0.5 ± 0.5 0.76 ± 0.43 * 0.93 7.0 0.86 Parameters were estimated only in each a w and temperature condition, and only compared within each row. * This value is not significantly different ( = 0.05) from zero. 83 5.2.4 Secondary model The GLM analyses , reported above, indicated that temperature, product structure, and a w had an effect on Salmonella inactivation rates . However, the product structure did not have a consistent effect on inactivation of Salmonella . F or example, product structure did not impact ( P > 0.05) Salmonella thermal resistance ( Table 5 . 6 ) i n wheat products at 0.45 and 0.65 a w at any of the temperatures (D - value s were compared via 95% CI). Salmonella thermal res istance in almond meal and almond butter at 0.25 and 0.45 a w also were equivalent ( P > 0.05) for all of the temperatures. Because the date s were only fabricated into paste, giving a mix of two and three levels of product structure, the product structure could not be calculated as a model parameter in the secondary model ( log - linear/ Bigelow - type model; Equation 2 1 ) . Instead log - linear/ Bigelow - type model parameters were estimate d for each of the product types and structure s . Reduction of Salmonella populati ons during thermal come - up time exceeded 3 log s in some cases ( Appendix J ). To reduce the impact of the varying time 0 populations on model parameters , normalized survivor data (log N/N 0 ) were used to estimate model parameters in all products for the secon dary model . 5.2.4.1 Reference conditions Optimization of the reference conditions (T ref and a w, ref ) for each model was required before fitting the models (Dolan et al., 2013) . Additionally, it is almost impossible to estimate other parameters (i.e., Z T and Z aw ) when the reference conditions were not close to opt imum references, due to the high parameter correlation (Schwaab and Pinto, 2007) . The reference conditions ( Figure 5 . 4 ) for each of the models generally were near the middle of the temperature and a w ranges for all of the products . The modified - Bigelow model of wheat 84 flour from Smith et al. (2016) also supported the mid - range for the reference conditions . However, Datta (1993) concluded that the reference temperature should be very close to the maximum experimental temperature . The difference in the present results ( Table 5 . 9 ) may have been influenced by the static vs. dynamic experimental temperatures. Figure 5 . 4 Example of reference conditions for the log - linear/Bigelow - type model of almond kernels. 5.2.4.2 Model evaluation The log - linear/ Bigelow - type model parameters for each of the products w ere estimated using a fixed reference condition. The RMSE and AIC C values ( Table 5 . 9 ) indicate that the secondary model for almond meal performed the best across all product types . A model for date pieces provided the highest RMSE and AIC C , indicating the uncertainty and pote ntial to overfit a 85 model, respectively. The highest individual parameter relative error for all of the models also was from the date pieces (35% for Z aw ). 86 Table 5 . 9 Parameter estimates (mean ± standard er ror) for the log - linear/ Bigelow - type models (secondary models), relative error (%), root mean squared error (RMSE), and AIC c values. Products Reference Conditions Parameter Estimate Relative Error (%) RMSE AIC c Wheat kernels T ref = 83.9°C Dref (min) 3. 7 8 ± 0. 05 1.4 0.60 - 242.79 a w, ref = 0.493 Z T (°C) 12.0 ± 0.2 1 1.8 Z aw 0. 471 ± 0.0 08 1.7 Wheat meal T ref = 83.1°C Dref (min) 5. 57 ± 0.1 1 2.0 0.75 - 133.17 a w, ref = 0.466 Z T (°C) 11. 2 ± 0. 29 2.6 Z aw 0. 388 ± 0.0 08 2.1 Wheat flour T ref = 82.7°C Dref (min) 6. 19 ± 0.2 2 3.6 1.05 27.24 a w, ref = 0.472 Z T (°C) 9. 88 ± 0. 36 3.6 Z aw 0.3 33 ± 0.0 10 2.9 Almond kernels T ref = 81.4°C Dref (min) 14. 6 ± 0.5 0 3.4 0.97 - 6.62 a w, ref = 0.451 Z T (°C) 16.2 ± 0.8 4 5.2 Z aw 1. 29 ± 0. 181 14.0 Almond meal T ref = 85.2°C Dref (min) 17. 5 ± 0.2 2 1.3 0.47 - 357.41 a w, ref = 0.451 Z T (°C) 13.8 ± 0. 29 2.1 Z aw 0.5 09 ± 0.0 11 2.1 Almond butter T ref = 83.8°C Dref (min) 15. 3 ± 0.3 0 2.0 0.75 - 134.56 a w, ref = 0.483 Z T (°C) 12.8 ± 0. 35 2.7 Z aw 0.4 28 ± 0.0 10 2.4 Date pieces T ref = 76.3°C Dref (min) 2.2 1 ± 0. 08 3.6 1.08 38.01 a w, ref = 0.469 Z T (°C) 11. 7 ± 0. 58 5.0 Z aw 3.6 2 ± 1. 268 35.0 Date paste T ref = 73.6°C Dref (min) 4. 08 ± 0.1 0 2.5 0.51 - 289.45 a w, ref = 0.543 Z T (°C) 12.6 ± 0.5 1 4.0 Z aw 0.4 13 ± 0.0 13 3.1 87 Residual analysis of the fitted model ( Figure 5 . 5 - 5.7 ), show ed different trend s for different products. The best fit ting model for almond meal was the log - linear/Bigelow - type , because RMSE and AIC c were the smallest. Wheat kernels and date p ieces showed a RMSE > 1 log (N/N 0 ), and a high AIC c , corresponding to the distributed data in Figure 5 . 5 A and Figure 5 . 7 A. Figure 5 . 5 Observed and predicted log (N/N 0 ) for the log - linear/ Bigelow - type model for : (A) wheat kernels, (B) wheat meal, and (C) wheat flour. Figure 5 . 6 Observed and predicted log (N/N 0 ) for the log - linear/Bigelow - type model for : (A) almond kernels, (B) almond meal, and (C) almond butter. 88 Figure 5 . 7 Observed and predicted log (N/N 0 ) for the log - linear/Bigelow - type model for : (A) date pieces and (B) date paste. When compared within product type, the bias ( Table 5 . 10 ) was not significantly different ( P > 0.05; confirmed by ANOVA). All mo del s underestimat ed the values for all products , except date pieces , which overestimated . The g reatest model bias was seen for almond products . Table 5 . 10 Model bias for each product from the Bigelow - type models. Negative value s indicate u nderpredict ion of actual lethality. Product Bias (log N/N 0 ) kernels/pieces meal flour/butter/paste Wheat - 0.02 - 0.10 - 0.08 Almond - 0.24 - 0.11 - 0.17 Date 0.03 NA - 0.05 89 The relationship between the D - value, that were calculated from the log - linear and the log - linear/Bigelow - type models, and a w ( Figure 5 . 8 - 5.10 ) showed some systematic errors , especially in almond products at low temperature and a w ( Figure 5 . 9 ). Therefore, the overall model may not be sufficiently accounting f or the interactive effect of temperature and a w on Salmonella thermal resistance. Figure 5 . 8 Relationship of D - value, estimated from log - linear ( symbols ) vs log - linear/Bigelow - type (line) models, and a w for wheat kernels, wheat meal, and wheat flour. 90 Figure 5 . 9 Relationship of D - value, estimated from log - linear ( symbol s ) vs log - linear/Bigelow - type (line) models, and a w for almond kernels, almond meal, and almond butter. 91 Figure 5 . 10 Relationship of D - value, estimated from log - linear ( symbols ) vs log - linear/Bigelow - type (line) models, and a w for date pieces, and date paste. Few studies reported Z T and Z aw based on the Bigelow - type model . Smith et al. (2016) developed secondary model for wheat flour (T ref = 80°C and a w, ref = 0.52), with RMSE of 0.78 log CFU/g. The calcula ted D 80°C, 0.52 aw for wheat flour in the present study was 8.26 min , which was higher than 2.52 min from Smith et al. (2016) . Their Z T (15.2°C) was also higher than the 9.88°C in this study, but the Z aw (0.33) was similar to value in this study . It should be noted that their experiment temperature was 75 - 85°C and a w was 0.310 - 0.700 a w . Product composition may also have caused the differences. The higher fat content of whole wheat flour compared to white wheat flour may also responsible for these differences. In addition, the range of temperatures may have had more of an impact than the range of a w , indicating the differences of Z T and Z aw in both studies. Villa - Rojas et al. (2013) also developed a Bigelow - type model for almond kernel flour. The ir Z T (8.28°C) and Z aw (0.187) w ere considerably lower than in the present study. Their a w, ref was fixed at 1 , and they used 121°C for T ref , which was above the experimental temperature range 92 (56 80°C). Moreover, the a w range (0.65 to 0.95) was almost entirely above that of the present study, which would be expected to affect model parameters. Other recent studies reported the similar multivariable inactivation models for non - isothermal and non - isomoisture treatment s of low - moisture products . Using similar Salmonella inactivation model for pistachios, Casulli (2016) reported that t he ir Z T (37.1°C) was higher than the present 16.2°C for almond kernels, but the Z aw (0.26) was lower than the 1.29 Z a w in this study. While these results suggest that temperature changes may have less of an influence in non - isothermal treatment s , a w changes had more of an influence when the product moisture changed dramatically during thermal processing. Also, the difference in product composition may have had an impact on the esti mated model parameters. Jeong et al. (2009) developed the modified MSU inactivatio n model based on process and dew point temperature for almonds. For the dry treatment (5% MC in oven), the Z T value (14.68 °C ) was close to 16.24 °C. Garcés - Vega (2017) further developed the modified MSU model for low - and high - humidity values. The Z T (69.1°C and 106°C) were much higher than in this study. The humidity of the process conditions may have caused the difference between the model parameters. Unfortunately, another parameter based on water properties cannot be compared due to the difference in the design of the experiments (i.e., closed containers in the present study vs. open - air heating in Jeong et al. 2009 ) The SSC analysis ( Figure 5 . 11 ) show s the correlation between model parameters. The SSC show s a similar result for all of the products ( Appendix K ) ; however, differences in the magnitude of SSC over time can be seen . Results also suggest a correlation between Z T and Z aw . A dditional analyses examined the correlation between Z T and Z aw ( Figure 5 . 12 ). 93 Figure 5 . 11 Example of SSC for the log - linear/ Bigelow - type model of almond kernels. Figure 5 . 12 Example of SSC for the log - linear/ Bigelow - type model for: (A) almond kernels and (B) almond meal . 94 Almond kernels and date pieces exhibited similar correlation trends between Z T and Z aw ( Figure 5 . 12 A and Appendix K ). Wheat products, almond meal, almond butter, and date paste also showed similar trends as in Figure 5 . 12 B (See Appendix K ). Results suggest a significant correlation between Z T and Z aw in almond kernels and date pieces. However, the relative errors of Z aw for almond kernels and date pieces were relatively high (14% and 35%, respectively). Additionally, the Z aw values were larger than the actual a w range (0 - 1). These results suggest that temperature has a greater impact on the D - value in large particle samples, confirming the relationship between Z T and Z aw ( Figure 5 . 13 ). For the small er - particle samples (wheat products, almond meal, almond butter, and date paste), the Z T and Z aw relative error results indic ated that the two factors were correlated at the beginning of the simulated experiment, but the n became uncorrelated when the experiment was completed (Appendix K , Figure K.2 - K .6 and K .8). Additionally , the relationship between Z T and Z aw ( Figure 5 . 13 ) were clump ed together in the figure. Th is result suggest s that particle size (product structure) has an impact on the products corresponding D - value. 95 Figure 5 . 13 . Relationship of Z aw and Z T (°C) for all products. 5.2.5 Water activity effects The thermal resistance of Salmonella is greater at lower a w than at higher a w values. During thermal treatment s, the a w of products was chang ing as the temperature increased. Th is temperature - induced change in a w values during heating may affect the thermal resistance of Salmonella (Syamaladevi et al., 2016a) . Water activity of the wheat and almond products increased ( P < 0.05) after heat ing to 80°C ( Table 5 . 11 ), but the a w, 80°C decreased ( P < 0.05) for the date products. Syamaladevi et al. (2016a) reported that the a w of wheat flour at 0.45 a w, 25°C increased to 0.80 after heating to 80°C , which was higher than the 0.650 in this study. Tadapaneni et al. (2017) reported a a w, 80°C for wheat flour ( initia l ~ 0.45 a w, 25°C ) at 0.73 when using test cells . Differences in measurement methodology may have impact ed the a w, 80°C results. Samples tested by Syamaladevi et al. (2016a) were measured using a vapor sorption analyzer , whereas Tadapaneni et al. (2017) were meas ured by a RH sensor within the test cell , as in the present study . Differences in wheat composition between these studies may have also account for the change in a w . 96 Table 5 . 11 Water activity values ( ± standard deviation) at 25 and 80°C for wheat, almond, and date products Products Measured a w at 25°C at 80°C 0.25 a w Wheat kernels 0.256 ± 0.001 B 0.455 ± 0.000 A Wheat meal 0.256 ± 0.001 B 0.475 ± 0.007 A Wheat flour 0.254 ± 0.001 B 0.465 ± 0.021 A 0.45 a w Wheat kernels 0.456 ± 0.001 C 0.670 ± 0.007 A Wheat meal 0.444 ± 0.001 D 0.670 ± 0.000 A Wheat flour 0.445 ± 0.001 D 0.650 ± 0.007 B 0.65 a w Wheat kernels 0.642 ± 0.004 B 0.780 ± 0.007 A Wheat meal 0.651 ± 0.000 B 0.795 ± 0.000 A Wheat flour 0.651 ± 0.004 B 0.795 ± 0.028 A 0.25 a w Almond kernels 0.249 ± 0.001 C 0.380 ± 0.007 A Almond meal 0.254 ± 0.001 C 0.415 ± 0.007 A Almond butter 0.249 ± 0.001 C 0.325 ± 0.007 B 0.45 a w Almond kernels 0.448 ± 0.002 D 0.525 ± 0.000 B Almond meal 0.442 ± 0.000 D 0.550 ± 0.000 A Almond butter 0.451 ± 0.000 D 0.465 ± 0.000 C 0.65 a w Almond kernels 0.644 ± 0.001 A 0.665 ± 0.000 A Almond meal 0.654 ± 0.000 A 0.690 ± 0.007 B Almond butter 0.662 ± 0.000 A 0.650 ± 0.007 A 0.25 a w Date pieces 0.254 ± 0.005 B , C 0.280 ± 0.014 B Date paste 0.238 ± 0.000 C 0.320 ± 0.000 A 0.45 a w Date pieces 0.456 ± 0.001 A 0.435 ± 0.007 B Date paste 0.459 ± 0.001 A 0.440 ± 0.000 B 0.65 a w Date pieces 0.648 ± 0.001 A 0.600 ± 0.014 B Date paste 0.642 ± 0.001 A 0.610 ± 0.000 B Within the same water activity and product type , means sharing a common superscript letter 97 Syamaladevi et al. (2016a) also reported an a w, 8 0 °C for peanut butter ( initia l ~ 0.45 a w, 25°C ) of 0.04, which was much lower than the almond butter ( a w, 8 0 °C = 0.465) in this study. In addition, Anderson et al . (2017) reported the a w, 80°C of a protein - fat blend (43% protein and 56% fat , and a w, 40 °C of 0.341 ) to be 0.366. Again, these differences may be caused by variation in methodology and product compositions. Moreover, the heating time to 80°C in this study (37 min) was considerably faster than for Syamaladevi et al. (2016a) (samples were reported to equilibrate for 2 weeks before measurement). Oil separation in these almond and peanut butter during longer equilibration process es could also have impact ed the measurement. In this study, a w, 80°C was similar ( P < 0.05) among st all wheat products at 0.25 a w, 25°C . However, Salmonella thermal resis tance in wheat meal and flour at 0.25 a w, 25°C was higher ( P < 0.05 ) than wheat kernels. Corresponding with the almond product results , the a w, 80°C for almond butter was equivalent to a w, 25°C ( P > 0.05) when the initial a w, 25°C was 0.65; however, Salmonella thermal resistance was greater in almond meal ( P < 0.05) than in almond butter and almond kernels. Additionally, Salmonella thermal resistance in date paste (0.25 a w, 25°C ) was greater ( P < 0.05) than on date pieces, but the a w, 80°C was greater ( P < 0.05). In contrast, the a w, 80°C of wheat product (0.65 a w, 25°C ) was not significantly different ( P > 0.05) amongst the product structures , and Salmonella thermal resistance was equivalent ( P > 0.05) for all wheat products. Additionally, Salmonell a thermal resistance among date products (0.45 and 0.65 a w, 25°C ) was equivalent ( P > 0.05) , and the a w, 80°C of date piece and date paste was not different ( P > 0.05). These results indicate that Salmonella thermal resistance may be partially affected by a w at each of the process ing temperature s ; however, some of the inconsistencies would imply that high - temperature a w cannot be the sole explanation for observed differences in Salmonella thermal resistance across product types and different structure s . 98 Additionally, the adsorption isotherm may have an impact on Salmonella thermal resistance. Garcés - Vega (2017) reported that the sensitivity of Salmonella on almonds was more likely influenced by moisture content than by a w . However, his study was non - isothermal and non - isomoisture, which was different from this study. T he measure d a w of the source wheat, almond s, and dates , as originally acquired, in the present study were 0.351, 0.504, and 0.709 a w , respectively. Therefor e, the 0.25 a w wheat products, 0.25 and 0.45 a w almond products, and all date products were in a desorption state, and all other products were in an adsorption state when tested. 5.2.6 Product type and structure effects Salmonella thermal resistance i s influenced by the composition of the food m a tri x . In this study, the impact of product composition ( Table 5 . 12 ) including the sugar profile of date s ( Table 5 . 13 ) on Salmonella thermal resistance was evaluated . W heat and date s were higher in carbohydrate s than almond s , with dates contain 63.3% sugar. Almond s w ere highest in fat among all three products. The se composition al differences (carbohydrate, fat, and sugar) may influence thermal resistance of Salmonella . Additionally, the date moisture content was 232% higher than wheat, and 475% higher than almond s , which could be the most significant factor lead ing to lower thermal resistance. 99 Table 5 . 12 Composit ion results for almond, wheats, and date. Compo nent % Content Refe re nce method Wheat Almond Date Ash 1.7 3.2 1.9 AOAC 920.153 Carbohydrate 81.4 24.0 71.4 Calculation* Protein 7.8 25.7 2.8 AOAC 922.06 Fat 2.0 43.0 0.3 AOAC 950.46 Moisture 7.1 4.1 23.6 AOAC 992.15 * The % content of carbohydrate was the percentage of solids that were not protein or fat. Table 5 . 13 Sugar profile f or date s using high - performance liquid chromatography (HPLC). Sugar type % Content Fructose 14.7 Glucose 17.0 Lactose < 0.04 Maltose < 0.04 Sucrose < 0.04 Total sugar 31.6 % total of sugar profile of date s was base d on 100% total in table 5.12 . DSC result s ( Table 5 . 14 ) showing the thermophysical transition s ( Appendix L ) during thermal processing suggested that a structure transition in almond butter, which may impact Salmonella thermal resistance. In addition, the characteristic peaks observed in date paste resulted from the melting of sugar, which likely could decrease Salmonella thermal resistance. 100 Table 5 . 14 DSC paramet ers ( ± standard deviation) for 0.25, 0.45, and 0.65 a w wheat flour, almond butter, and date paste. Products Glass transition temperature (°C) Characteristic temperature and enthalpy T gi T g T ge T po (°C) T p (°C) Wheat flour 0.25 a w NA 1 NA NA NA NA NA 0.45 a w NA NA NA NA NA NA 0.65 a w NA NA NA NA NA NA Almond butter 0.25 a w 39.3 ± 1.7 B , 2 41.6 ± 0.6 B 44.0 ± 1.2 B NA NA NA 0.45 a w 46.7 ± 0.5 A 48.6 ± 1.0 A 49.6 ± 1.0 A NA NA NA 0.65 a w 44.8 ± 0.2 A 45.5 ± 0.9 A 45.8 ± 0.4 A, B NA NA NA Date paste 0.25 a w NA NA NA 66.4 ± 3.4 A 80.9 ± 1.7 A 11.9 ± 0.7 B 0.45 a w NA NA NA 72.2 ± 0.3 A 83.0 ± 0.0 A 18.0 ± 0.6 A 0.65 a w NA NA NA 51.7 ± 0.2 B 70.9 ± 0.2 B 9.2 ± 0.7 B 1 NA means values could not be estimated due to no thermal transitions . 2 Within a column, values with a common superscript letter were not significantly different ( = 0.05). 101 5.2.6.1 Wheat products Salmonella was most thermal ly resistan t among the wheat products at the D 80°C, 0.25 aw in wheat flour. Grinding the wheat kernels into meal and flour increased (P < 0.05) Salmonella thermal resistance at 0.25 a w . However, Salmonella thermal resistance in the meal and flour were not significantly different from each other regardless of temperature or a w . Moreover, Salmonella exhibited similar thermal resistance P > 0.05) at 0.45 and 0.65 a w , regardless of the product structure. The DSC results ( Table 5 . 14 ) show no thermal transitions for wheat flour regardless of a w or temperature; therefore, it appears unlikely that any such thermophysical changes are affecting Salmonella thermal resistance. However, at 0.25 a w , Salmonella thermal resistance was higher ( P < 0.05 ) in wheat flour compared to wheat kernels for all temperatures. This result indicates that product structure has a greater influence on Salmonella thermal resistance in wheat products at low a w , even though the mechanism for this effe ct is unknown . Unfortunately, previous studies did not report Salmonella thermal resistance on wheat kernels or in wheat meal. Syamaladevi et al. (2016a) reported D 80°C, 0.45 aw of 6.9 min for wheat flour , which was lower than the 11.6 min in this study. Smith et al. (2016) reported a D 80°C, 0.427 aw of 5 . 5 min for wheat flour which also was lower than in this study. The Smith et al. (2016) D - value for wheat flour would be expected to be higher than that of Syamaladevi et al. (2016a) and this study due to the lower a w , but results indicated the opposite . Syamaladevi et al. (2016a) inoculated wheat flour with a Salmonella cocktial, but Smith et al. (2016) and this study used a sing le strain of Salmonella Enteritidis PT 30 as in this study . Another difference was the overall composition of the white wheat fl our used. Syamaldevi et al. (2016a) and Smith et al. (2016) used 102 different commercial brands of wheat flour brands, where in this study whole wheat flour was produced by grinding the wheat kernels in a mill and not re moving the wheat bran. The results reinforce the notion that Salmonella thermal resistance is impacted by produc t composition . 5.2.6.2 Almond products Within a lmond products , Salmonella was most thermal ly resistan t in almond meal regardless of temperature or a w . The fabrication process for turning almond kernels into meal and butter increased Salmonella thermal resistance by 325% at 0.25 a w and T = 80°C ( Table 5 . 7 ) . In comparison, Salmonella thermal resistance in almond meal and butter was equivalent ( P > 0.05) for all temperatures at 0.25 and 0.45 a w . These results suggest that Salmonella thermal resistance o n/in almond products was influenced by more product structure at high a w . At 0.65 a w , Salmonella thermal resistance in almond meal was greater ( P < 0.05) than in/on the butter and kernels for all temperatures. A lmond butter at 0 .65 a w is higher in moisture and therefore lower in fat content. Hence, the reduction in fat content at the higher a w may have contributed to the decreased in thermal resistance (He et al., 2011) . The DSC results ( Table 5 . 14 ) show that the phase transition for almond butter occurred between 42 - 48 °C , regardless of a w . Glass transition temperatures (T g ) for almond butter at 0.45 and 0.65 a w were higher ( P < 0.05 ) than at 0.25 a w , but Salmonella thermal resistance of almond meal and butter at 0.45 a w were not significantly different ( P > 0.05 ) , possibly due to denaturation of almond protein at 80°C (Amirshaghaghi et al., 2017) . Based on these results , and experimental observation during testing, the almond butter changed from a viscous - liquid to a semi - solid product, which behaved similarly to almond meal. 103 Almond kernels have been subjected to various thermal pasteurization processes, using moist - air oven (Jeong et al., 2009) , hot water (Harris et al., 2012) , or hot oil (Abd et al., 2012; Du et al., 2010) . The D 80°C for one hot water treatment was 0.75 min (Harris et al., 2012) and the hot oil treatment reducing Salmonella by 4 logs in 1.2 min at 121°C (Du et al., 2010) . In th e present study, the D 80°C was 23.1 min. Only one study reported the D - value for almond meal . Villa - Rojas et al. (2013) reported a D 80°C, 0.601 aw of 1.63 min, which was much lower than the 20.1 min D - value (0.65 a w ) in the present study. The almond meal of Villa - Rojas et al. (2013) was inoculated after fabrication, whereas the opposite occurred in this study. Therefore, the variable D - values can be partially expla i ned by differences in the inoculation methods (See Chapter 3). In compari ng the almond butter results with prior peanut butter work , Li et al. (2014a) and He et al. (2013) used peanut butter containing 48 and 49% fat , which is close to the almond butter in this study. Li et al. (2014a) 80°C 8 5 °C 9 0°C values of 1.6, 2.3, and 2.6 min, respectively, whereas He et al. (2013) report ed D 90°C values at 0.2, 0.4, and 0.6 a w of 4.8, 3.4, and 2.1 min, respectively. These r esults indicate that Salmonella thermal resistance was greater in almond butter than peanut butter at the temperatures and a w values tested . As described in Chapter 3, the post - fabrication method of inoculat ion was likely responsible for the lower thermal inactivation rate. 5.2.6.3 Date products The fabrication process increased Salmonella thermal resistance in date products. The resistance in date paste was higher ( P < 0.05) than in d ate pieces at 0.25 and 0.45 a w ; h owever , product structure did not impact Salmonella thermal resistance at 0.65 a w . This behavior is in 104 contrast with the fabricated almond product results. Increasing the a w of date products reduced the effect that structure ha d on Salmonella the rmal resistance. The DSC results also suggest that some thermophysical transitions may affect Salmonella thermal resistance in date products. A t 0.65 a w , t he sugars in date paste began melting faster than at 0.25 and 0.45 a w ; however, the enthalpy of transition was higher ( P < 0.05) at 0.45 a w than at 0.25 and 0.65 a w . S ucrose, glucose, and fructose generally do not melt at 80°C (Lee et al., 2011) ; however, the date paste in this study contained more water than their dry system ; therefore, date paste was meltable during thermal treatment. Unfortunately, no prior studies assessed Salmonella thermal resistance on dates. In this study, date products contained 63.3% sugar. Mattick et al. (2001) assesse d Salmonella thermal resistance i n a high sugar content (0.65 a w ) broth at 70 - 80°C , reporting the estimated time for a first log reduction of Salmonella Typhimurium between 0.9 3.6 min, which w ere similar to those seen for date paste in this study , but lower than for the date pieces. 5.2.6.4 Comparison between similar product structure s . P roduct structure was evaluated based o n particle size. Large - particle (wheat kernels, almond kernels, and date pieces), small - particle (almond meal, wheat meal, and wheat flour), and paste (almond butter and date paste) products were compared using D 80°C at different water activit ies . In the large - particle comparison ( Figure 5 . 14 ), a w did not have a large impact on the D - value for almond kernels or date pieces; however, the D - value decreased ( P < 0.05) a s a w increased for wheat kernels. For the small - particle comparis on, all of the products showed a similar trend ; as a w increased , the thermal resistance of Salmonella decreased ( P < 0.05) . 105 Figure 5 . 14 Relationship of D 80°C , estimated from log - linear (dot) vs log - linear/Bigelow - type (line) models, with a w in similar product structure. The D 80°C, 0.45 aw for all of the products ( Table 5 . 15 ) were calculated for all products using the model parameters in Table 5 . 9 . When comparing the D - value s via 95% CI, product structure had an impact on Salmonella thermal resistance across all product types. S tandard errors of the estimated D - value from the primary l og - linear model were 1 .0 1.5 min for wheat products, 3.1 7.2 min for almond products, and 0.4 1.5 min for date products, whereas the standard errors from the log - linear/ Bigelow - type model via global regression ranged from 0.06 to 1.19 min . The 106 range of uncertainty in parameter estimates may be due to product structure effects, causing similar thermal resistance in fabricated products. Table 5 . 15 Calculated D - value s (± standard error) at 0.45 a w and 80°C using log - linear/ Bigelow - type model . Products D 80°C, 0.45aw (min) Wheat kernels 9.80 ± 0.21 A Wheat meal 11.30 ± 0.34 B Wheat flour 13.35 ± 0.68 C Almond kernels 17.83 ± 0.72 A Almond meal 41.79 ± 0.97 B Almond butter 36.15 ± 1.19 C Date pieces 1.08 ± 0.06 A Date paste 2.14 ± 0.11 B = 0.05). Statistical analyses were confirmed via 95% CI. A relationship between Z T or Z aw and product structure w as also seen ( Figure 5 . 15 ). When particle size decreased , Z T increased for almond and wheat products. For Z aw , the Z aw tend ed to decrease to a similar value across all product s as the particle size decreased . These results indicate that a w has a greater influence on the D - value as the particle size decreased . 107 Figure 5 . 15 Relationship between (A) Z T and product structure, and (B) Z aw and product structure. 5.3 Conclusion P rocess t emperature, product structure, and a w all impacted thermal resistance of Salmonella in all three low - moisture product groups (wheat, almond, and dates) . Product fabrication inc reased Salmonella thermal resistance for all products. As a w increased , Salmonella thermal resistance decreased at different rate s due to variation in product composition (% fat, protein, and carbohydrate). The products that yielded the highest thermal resistance response in Salmonella were almond meal and almond butter , likely because the fat content protected Salmonella at the high process ing temperature. Date pieces yielded the lowest thermal resistance for Salmonella , likely due to moisture content of date s being much higher than wheat and almond . The log - linear model was more - likely - correct model to use in this study for predicting Salmonella lethality . The model ( log - linear/ Bigelow - type) that was developed provided a better under standing of the relationship between temperature and a w ; u nfortunately, product structure currently cannot be included as a model term in the secondary model , as it is a discrete state rather than a continuous variable. Regarding the impact of particle size on l ethality of Salmonella , l arge particles result ed in a high relative error in the estimated Z aw . For small particles and paste, both 108 the temperature and a w influence d the Salmonella D - value. The relationship between Z T and Z aw to these par ticle sizes showed a similar impact on Salmonella inactivation. The general conclusion might be that structure effects are large for the step change in structure due to any grinding, but that finer reductions in particle size (e.g., almond meal to butter, or wheat meal to flour) have much less impact on the thermal resistance of Salmonella present in those structure. 109 6 OVERALL CONCLUSIONS AND RECCOMMENDATION S 6.1 Other methodological/preliminary work In addition to the results presented in Chapter s 3 - 5, several other preliminary studies were conducted, with some of their results presented at various conference s . These studies investigated the impact of sample equilibration, kernel surface integrity , and sample containers on Salmonella thermal resist ance , and are summarized in Appendix M to P . 6.2 Overall C onclusions Salmonella thermal resistance in fabricated products (meal, butter, flour, and paste) was higher ( P < 0.05) than in whole products (kernels and pieces) using the pre - and postfabrication protocol s , except for wheat products. Using similar inoculation method s ( pre - fabrication ), Salmonella thermal resistance was lower on wheat meal and flour in the inoculation protocols study (Chapter 3) ( P < 0.05) than in the Chapter 5 study, likely because wheat bran w as partially removed in Chapter 3, resulting in a lower fat content. Fabrication process es also change the microenvironment (e.g., location, attachment) of low - moisture products, which could be the root cause for the obser ved difference s in Salmonella thermal resistance, likely because Salmonella was located in or attached to specific microenvironment s . Contrastingly, Salmonella thermal resistance in fabricated wheat products was not significantly different after fabrication, which may be due to a less discrete microenvironment and lower fat and sugar content compared to other products. Product composition is clearly a very impor tant factor influenc ing Salmonella thermal resistance. P roduct s highest in fat (almonds) yielded the highest thermal resistance among all 110 fabricated products . In contrast, the h igh sugar content and moisture content of date products resulted in much lower thermal resistance compared to wheat and almon d products. During long - term storage of almonds , Salmonella populations decreased; however, Salmonella thermal resistance generally remained unchanged ( P > 0.05) during long term storage. This information is cr itically important, to know that the storage age of almonds does not affect thermal resistance and therefore would not affect the approach to thermal process validations in industry. The relationship between temperature , a w , and Salmonella thermal resistance in low - moisture foods is extremely important . Increasing a w and temperature decreased Salmonella thermal resistance. Additionally, a w did change after heating and was correlated with Salmonella thermal resistance in some of cases; there fore, a w at the process ing temperatures is likely a partial , but incomplete explanation for the observed differences in Salmonella thermal resistance. The resulting primary and secondary models indicated that the log - linear model was the most - likely - correc t for low - moisture foods . A Bigelow - type secondary model was developed based on the log - linear model. Based on model parameters, Salmonella thermal resistance on the large - particle products (kernels and pieces) were influenced by temperature more than a w , but resistance on the small - particle products (meal, butter, flour, and paste) w as affected by both temperature and a w . The models also suggest that Salmonella on/in all the small - particle and paste products has a similar response with temperature and a w . Physical structure also influences Salmonella thermal resistance , which is one of the most novel conclusions of this dissertation . During thermal processing , almond butter changed from a 111 viscous - liquid to a semi - solid, resulting in equivalent D - value s for almond meal and butter. Melting of date paste during thermal treatment may have partially increase d lethality. Unfortunately, the relationships between a w and product structure cannot be directly modeled because product structure is a discrete class variab le . In compari ng D - value s at differen t a w , the moisture content of almond products had a greater influence on Salmonella resistance in very small particles (butter) than on large particles (kernels). Date paste results also support this conclusion. Salmonella in date paste at high a w showed very low thermal resistance compared to the other products at equivalent a w levels . Contrastingly , the structure of medium - sized particles (meal and flour) had an impact at low a w . Also, under very dry conditions, Salmonella thermal resistance in wheat meal and flour was greater than on wheat kernels. Overall, the log - linear/Bigelow - type inactivation models fit well for all products (RMSE from 0.51 to 1.08 log), supporting the robustness of this model form for low - moisture products. All factors (i.e., product type, product structure, temperature, and a w ), except long - term storage, impacted Salmonella thermal resistance. This study also demonstrates that inoculation methods and specific product structures should be c onsidered as critical factors when designing process validation studies. 6.3 Future Work Product structure has been introduc ed here as a new factor that impacted Salmonella thermal resistance in low - moisture foods. However, structure actually encompasses many factors, such as particle size and form (i.e., solid, liquid) . Therefore, it is recommended that future work be designed to test particle size as a continuous variable of multiple additional levels, in order to further evaluate whether particle size can be incorporated as a model term. 112 The physical micro - structure of some low - moisture foods can also chang e during thermal treatment in some of cases. The a w increase s /decrease s when temperature increase s . T he change in a w m ay depend on thermal treatment methods, or also on thermo - physical transition s during heating . The relationship between Salmonella thermal resistance and thermo physical properties and a w at process temperature s is suggested as a topic of further investigation to better understand the impact of product structure during pasteurization. The sorption - isotherm also impacts Salmonella thermal resistance. For example, the moisture content of date s is much higher than that for wheat and almond p roducts at the same a w . This study was based on the iso - a w and a isomoisture treatment ; however, real processes generally are non - isomoisture. Therefore, it is important to determine the impact of dynamic sorption - isotherms on Salmonella thermal resistance , especially in small - particle s, due to their higher impact on the resistance. Lastly, the models presented in this dissertation were based on isothermal/isomoisture data. Therefore, validation of these models in non - iso conditions is essential before the y could be applied for commercial process validation. Future work should encompass pilot - scale thermal treatment (e.g., roasting or toasting) of similar products, to quantitatively validate model performance under dynamic conditions . Overall, Salmonella thermal resistance was impacted by change s in product structure , such as particle size, physical state changes, microenvironment properties, and water relationships (i.e., a w , moisture content, sorption isotherm). Therefore, future research can hopefully advance the science by developing additional novel models that incorporate those fundamental product characteristics, which could lead to an improved general conclusion/theory for product structure effects on Salmonella t hermal inactivation. 113 A PPENDICES 114 Survivor Data for the Inoculation Protocol Experiment (Chapter 3) This appendix shows Salmonella inactivation data during thermal treatment at 80 ° C after samples were inoculated with pre - and post - fabrication protocols and equilibrated to 0.40 0.45 a w . Table A . 1 Salmonella inactivation data during isothermal treatment (80 ° C) for almond meal (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 1 0 7.91 0.00 0 8.94 0.00 16.5 7.51 - 0.40 16.5 9.02 0.08 31.5 6.93 - 0.98 31.5 8.45 - 0.50 46.5 6.68 - 1.23 46.5 7.73 - 1.22 61.5 6.76 - 1.15 61.5 7.71 - 1.23 76.5 6.23 - 1.68 76.5 7.06 - 1.89 91.5 6.25 - 1.66 91.5 6.83 - 2.12 106.5 6.23 - 1.67 106.5 6.82 - 2.12 121.5 6.04 - 1.87 121.5 6.47 - 2.48 136.5 5.85 - 2.06 136.5 5.95 - 3.00 151.5 5.20 - 2.70 151.5 5.94 - 3.01 REP 2 0 7.80 0.00 0 8.81 0.00 16.5 7.24 - 0.56 16.5 8.38 - 0.43 31.5 6.83 - 0.97 31.5 7.39 - 1.42 46.5 6.51 - 1.29 46.5 7.45 - 1.35 61.5 6.02 - 1.78 61.5 6.98 - 1.83 76.5 5.84 - 1.96 76.5 5.51 - 3.30 91.5 5.46 - 2.34 91.5 6.31 - 2.50 106.5 5.00 - 2.80 121.5 3.65 - 5.16 121.5 5.34 - 2.46 136.5 2.90 - 5.91 136.5 4.66 - 3.14 151.5 3.15 - 5.66 151.5 4.98 - 2.81 115 Table A . 1 . Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 3 0 7.75 0.00 0 9.43 0.00 16.5 7.38 - 0.37 16.5 9.14 - 0.29 31.5 6.60 - 1.16 31.5 8.29 - 1.14 46.5 6.34 - 1.41 46.5 7.74 - 1.68 61.5 6.02 - 1.73 61.5 6.89 - 2.54 76.5 5.52 - 2.23 76.5 6.99 - 2.44 91.5 5.45 - 2.30 91.5 6.59 - 2.84 106.5 5.34 - 2.41 106.5 6.01 - 3.42 113.5 5.53 - 2.23 121.5 5.79 - 3.64 136.5 5.45 - 2.30 136.5 5.19 - 4.24 151.5 5.49 - 2.26 151.5 4.91 - 4.52 Table A . 2 Salmonella inactivation data during isothermal treatment (80 ° C) for almond butter (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 1 0 7.69 0.00 0 9.64 0.00 16.1 6.88 - 0.81 15 8.23 - 1.41 31.1 6.37 - 1.32 30 6.55 - 3.09 46.1 5.89 - 1.81 45 5.16 - 4.48 61.1 5.51 - 2.18 60 4.93 - 4.71 76.1 5.22 - 2.48 120 3.22 - 6.43 91.1 4.92 - 2.78 135 3.28 - 6.36 106.1 4.92 - 2.78 150 3.38 - 6.26 121.1 4.67 - 3.02 136.1 4.48 - 3.22 151.1 4.16 - 3.53 116 Table A . 2 . Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 2 0 7.26 0.00 0 9.39 0.00 16.1 6.62 - 0.63 15 8.17 - 1.22 31.1 5.33 - 1.92 30 5.88 - 3.51 46.1 5.36 - 1.89 45 5.23 - 4.16 61.1 5.44 - 1.82 60 4.54 - 4.85 76.1 5.94 - 1.31 75 4.69 - 4.70 86.1 4.94 - 2.32 90 4.06 - 5.33 106.1 4.76 - 2.49 105 3.85 - 5.54 121.1 4.76 - 2.49 120 3.36 - 6.03 136.1 5.02 - 2.24 135 2.18 - 7.21 151.1 4.79 - 2.46 150 2.30 - 7.09 REP 3 0 7.30 0.00 0 8.62 0.00 16.1 5.46 - 1.84 30 6.10 - 2.53 31.1 5.66 - 1.64 45 4.18 - 4.45 46.1 4.74 - 2.56 60 5.04 - 3.58 61.1 5.01 - 2.29 75 3.48 - 5.15 76.1 4.79 - 2.51 120 2.60 - 6.02 91.1 4.92 - 2.38 106.1 5.14 - 2.16 121.1 5.11 - 2.19 136.1 4.94 - 2.36 151.1 4.30 - 3.00 117 Table A . 3 Salmonella inactivation data during isothermal treatment (80 ° C) for wheat meal (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 1 0 8.70 0.00 0 8.59 0.00 6 7.60 - 1.10 15 7.46 - 1.13 12 6.81 - 1.89 30 5.93 - 2.66 18 6.03 - 2.67 45 6.01 - 2.58 24 5.99 - 2.71 60 5.37 - 3.22 30 5.68 - 3.02 75 4.41 - 4.18 36 5.44 - 3.26 90 4.56 - 4.03 42 5.09 - 3.61 105 3.82 - 4.77 48 5.13 - 3.57 REP 2 0 8.28 0.00 0 8.54 0.00 6 7.41 - 0.86 15 7.24 - 1.29 12 6.93 - 1.34 30 5.48 - 3.06 18 6.18 - 2.10 45 5.29 - 3.25 24 5.79 - 2.49 60 4.80 - 3.74 30 5.42 - 2.86 75 4.40 - 4.14 36 4.83 - 3.45 90 3.48 - 5.05 42 4.56 - 3.72 105 3.59 - 4.95 48 3.54 - 4.73 REP 3 0 8.50 0.00 0 8.45 0.00 6 7.41 - 1.08 15 7.03 - 1.42 12 6.93 - 1.56 30 6.16 - 2.29 18 6.18 - 2.32 45 6.33 - 2.11 24 5.79 - 2.71 60 5.35 - 3.09 30 5.42 - 3.08 75 4.24 - 4.20 36 4.83 - 3.67 90 3.84 - 4.61 42 4.56 - 3.94 120 3.69 - 4.76 48 3.54 - 4.95 118 Table A . 4 Salmonella inactivation data during isothermal treatment (80 ° C) for wheat flour (~0.4 a w ) using pre - fabrication and post - fabrication inoculation protocols. Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 1 0 8.71 0.00 0 0 0.00 6 7.61 - 1.10 15 15 - 1.65 12 6.46 - 2.25 30 30 - 2.71 18 6.51 - 2.20 45 45 - 2.90 24 6.07 - 2.64 60 60 - 3.38 30 5.04 - 3.67 75 75 - 3.93 36 4.86 - 3.84 90 90 - 5.33 42 4.28 - 4.43 105 105 - 6.37 48 3.84 - 4.87 REP 2 0 8.53 0.00 0 0 0.00 6 7.20 - 1.33 15 15 - 2.53 12 5.90 - 2.63 30 30 - 2.69 18 5.87 - 2.67 45 45 - 3.03 24 5.48 - 3.05 60 60 - 4.63 30 4.18 - 4.36 75 75 - 5.04 36 3.00 - 5.53 90 90 - 4.71 42 3.63 - 4.90 105 105 - 7.22 120 120 - 7.57 REP 3 0 8.54 0.00 0 0 0.00 6 8.13 - 0.41 15 15 - 2.82 12 6.90 - 1.65 30 30 - 4.19 18 6.92 - 1.62 45 45 - 4.61 24 5.92 - 2.62 60 60 - 5.86 30 5.31 - 3.23 75 75 - 5.26 36 4.82 - 3.72 90 90 - 5.98 42 4.76 - 3.78 105 105 - 6.41 48 4.12 - 4.43 119 Table A . 5 Salmonella inactivation data during isothermal treatment (80 ° C) for date paste (~0.45 a w ) using pre - fabrication and post - fabrication inoculation protocols. Pre - fabrication Post - fabrication Rep Time (min) Log CFU/g Log N/N 0 Time (min) Log CFU/g Log N/N 0 REP 1 0 6.26 0.00 0 5.02 0.00 20 6.70 0.44 20 4.45 - 0.57 40 6.03 - 0.23 40 4.10 - 0.93 60 5.79 - 0.46 60 3.19 - 1.83 80 5.75 - 0.51 100 2.60 - 2.42 100 5.45 - 0.81 120 2.81 - 2.21 120 6.00 - 0.26 140 2.98 - 2.05 140 5.65 - 0.61 180 5.71 - 0.54 REP 2 0 6.80 0.00 0 5.09 0.00 20 7.00 0.21 20 4.09 - 1.00 40 6.95 0.15 40 4.08 - 1.01 60 7.11 0.31 60 5.63 0.54 80 6.07 - 0.73 80 4.89 - 0.20 100 6.23 - 0.57 100 4.34 - 0.75 120 6.45 - 0.34 120 3.53 - 1.56 140 3.54 - 1.55 160 3.03 - 2.06 REP 3 0 7.12 0.00 0 4.90 0.00 20 6.63 - 0.48 20 5.25 0.35 40 7.08 - 0.03 40 3.41 - 1.49 60 6.79 - 0.32 60 3.72 - 1.18 80 6.14 - 0.98 80 4.94 0.04 120 6.24 - 0.88 100 3.35 - 1.55 140 6.38 - 0.73 120 2.86 - 2.04 160 5.84 - 1.28 140 1.90 - 3.00 160 2.23 - 2.67 REP 4 0 6.71 0.00 0 4.69 0.00 30 6.64 - 0.06 20 4.77 0.08 40 6.42 - 0.28 40 4.90 0.21 60 7.27 0.56 60 4.13 - 0.56 100 6.24 - 0.46 80 4.20 - 0.49 120 5.77 - 0.93 100 3.54 - 1.15 140 6.12 - 0.59 120 4.77 0.09 180 6.08 - 0.63 140 2.15 - 2.54 120 Survivor Data for the Long - Term Storage Experiment s (Chapter 4) This appendix presents Salmonella inactivation data for whole almonds (~0.45 a w ) during isothermal treatments (80 ° C) after 0, 7, 15, 27, 68, 70, and 103 weeks of storage at room temperature. Table B . 1 Salmonella inactivation data during isothermal treatment (80°C) for whole almonds after 0, 7, 15, 27, 68, 70, and 103 weeks of storage at room temperature. Storage time (weeks) Time (min) Rep 1 Rep 2 Rep 3 0 0 7.80 7.96 8.06 12 7.23 7.04 7.35 24 5.59 7.29 7.19 36 5.14 4.41 4.67 48 6.01 5.63 - 60 5.94 - 72 - - - 84 - 4.63 3.59 96 4.68 3.98 - 108 - - 3.83 7 0 7.78 7.32 7.62 12 5.69 6.91 6.86 24 4.60 5.45 5.17 36 4.18 5.10 6.11 48 4.16 5.84 4.91 60 3.98 5.16 3.92 72 4.21 - - 15 0 7.02 7.61 7.03 12 - 5.63 5.41 24 6.01 5.11 5.83 36 4.35 4.77 4.12 48 4.37 4.64 - 60 2.60 - - 72 4.18 - - 84 - - 3.85 Data points were excluded due to Salmonella survival counts being lower than 25 colonies/plate. 121 Table B . 1 Storage time (weeks) Time (min) Rep 1 Rep 2 Rep 3 27 0 7.15 - - 10 6.74 6.47 - 20 5.28 4.49 5.64 30 4.02 4.33 4.70 40 5.65 6.55 4.48 50 4.02 3.82 - 70 4.53 - - 68 0 6.05 6.22 5.80 6 5.08 5.56 5.68 12 5.31 - - 18 4.95 5.40 4.00 24 - - 4.97 30 - 4.26 - 36 - - - 42 - 3.87 - 48 - - 3.72 70 0 5.72 7.13 6.22 6 4.94 4.56 6.72 12 4.84 4.46 5.68 18 6.27 3.59 5.87 24 4.08 4.77 3.41 30 - 4.01 3.58 36 - 3.61 - 42 - 2.85 - 48 - 2.60 2.72 103 0 4.75 6.27 4.92 5 3.29 3.62 4.01 10 3.43 6.06 - 15 5.28 3.43 3.94 20 3.43 2.67 4.08 25 4.84 4.91 - 30 - 3.46 2.77 35 - 3.45 2.31 40 - 3.04 3.30 Data points were excluded due to Salmonella survival counts being lower than 25 colonies/plate. 122 Product Properties This appendix includes the size distribution for almond meal, wheat meal, and wheat flour. Almond meal was analyzed using Microtrac Laser light scattering method , and the wheat meal and wheat flour were analyzed using the American Society of Agricultural a nd Biological Engineers (ASABE) standard S319.2 method of determining and expressing fineness of feed materials by sieving. Table C . 1 Size distribution for wheat meal and wheat flour. Wheat meal Wheat flour Size range ( m m) Fraction (%Mass) Size range ( m m) Fraction (%Mass) > 0.814 34.4 > 0.814 6.6 0.814 0.420 23.8 0.814 0.420 26.1 0.420 0.250 10.7 0.420 0.250 14.0 0.250 0.177 7.5 0.250 0.177 13.0 0.177 0.149 6.2 0.177 0.149 24.5 0.149 0.125 1.2 0.149 0.125 5.8 < 0.125 16.2 < 0.125 10.0 Table C . 2 Size distribution for almond meal. Peaks summary Size ( µ m) Fraction (%Vol) 871.9 40.0 227.8 35.2 9.62 24.8 123 Photographs of Experimental Work Figure D . 1 Example of almond kernels conditioning in equilibration chamber. Figure D . 2 Custom - designed stirrer using for equilibrating almond butter. 124 Figure D . 3 Water activity meter for measuring a w at high temperature. Figure D . 4 Example of almond products in plastic bag and aluminum test - cell for thermal inactivation studies . 125 Come - Up Time for Thermal Inactivation This appendix shows the come - up time for thermal inactivation in this study, based on 6 replicates. Table E . 1 Come - up time ( ± standard deviation) for almond products. Sample 80 ° C 85 ° C 90 ° C Almond kernels 2.7 ± 0.4 3.1 ± 0.2 3.1 ± 0.3 Almond meal 2.0 ± 0.1 2.3 ± 0.2 2.4 ± 0.1 Almond butter 2.1 ± 0.1 2.1 ± 0.1 2.0 ± 0.2 Table E . 2 Come - up time ( ± standard deviation) for wheat products. Sample 80 ° C 85 ° C 90 ° C Wheat kernels 0.8 ± 0.2 0.8 ± 0.1 0.8 ± 0.1 Wheat meal 1.6 ± 0.2 1.7 ± 0.3 1.6 ± 0.4 Wheat flour 1.3 ± 0.2 1.6 ± 0.2 2.1 ± 0.1 Table E . 3 Come - up time ( ± standard deviation) for date products. Sample 70 ° C 75 ° C 80 ° C Date pieces 3.4 ± 0.3 3.1 ± 1.3 1.9 ± 0.3 Date paste 1.0 ± 0.1 1.1 ± 0.1 2.5 ± 0.1 126 Survivor Data for Water Activity, Product Structure, and Temperature Experiments (Chapter 5) This appendix presents Salmonella inactivation data after heating in an isothermal water bath at three different temperatures (80, 85, and 90 ° C for wheat and almond products, a nd 70, 75, and 80 ° C for date products). Samples were equilibrated at three different water activities (0.25, 0.45, and 0.65 a w ) before performing the thermal experiments. Table F . 1 Salmonella inactivation data for almond kernels. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 7.70 0.00 0.258 3.02 9 6.36 - 1.35 0.258 3.02 18 6.29 - 1.41 0.258 3.02 27 6.51 - 1.19 0.258 3.02 36 5.16 - 2.54 0.258 3.02 45 4.83 - 2.87 0.258 3.02 54 3.69 - 4.02 0.258 3.02 63 2.30 - 5.40 0.258 3.02 72 2.59 - 5.11 0.258 3.02 REP 2 0 7.55 0.00 0.265 3.25 9 7.16 - 0.39 0.265 3.25 18 5.67 - 1.88 0.265 3.25 27 5.88 - 1.67 0.265 3.25 36 6.18 - 1.37 0.265 3.25 45 5.95 - 1.60 0.265 3.25 54 4.65 - 2.90 0.265 3.25 63 4.31 - 3.24 0.265 3.25 72 3.72 - 3.83 0.265 3.25 REP 3 0 7.09 0.00 0.243 3.03 9 7.21 0.13 0.243 3.03 18 6.18 - 0.90 0.243 3.03 27 5.37 - 1.71 0.243 3.03 36 5.85 - 1.24 0.243 3.03 45 3.97 - 3.12 0.243 3.03 54 4.29 - 2.80 0.243 3.03 63 4.75 - 2.33 0.243 3.03 72 3.62 - 3.47 0.243 3.03 127 Table F . 1 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 80 REP 1 0 7.80 0.00 0.462 3.95 12 7.23 - 0.57 0.462 3.95 24 5.59 - 2.21 0.462 3.95 36 5.14 - 2.66 0.462 3.95 48 6.01 - 1.79 0.462 3.95 60 5.94 - 1.86 0.462 3.95 96 4.68 - 3.12 0.462 3.95 108 2.88 - 4.92 0.462 3.95 REP 2 0 7.96 0.00 0.449 3.58 12 7.04 - 0.92 0.449 3.58 24 7.29 - 0.67 0.449 3.58 36 4.41 - 3.56 0.449 3.58 48 5.63 - 2.34 0.449 3.58 60 4.04 - 3.92 0.449 3.58 84 4.63 - 3.34 0.449 3.58 96 3.98 - 3.99 0.449 3.58 108 2.85 - 5.12 0.449 3.58 REP 3 0 8.06 0.00 0.446 3.96 12 7.35 - 0.71 0.446 3.96 24 7.19 - 0.87 0.446 3.96 36 4.67 - 3.39 0.446 3.96 48 4.28 - 3.78 0.446 3.96 60 6.97 - 1.09 0.446 3.96 84 3.59 - 4.47 0.446 3.96 96 2.60 - 5.46 0.446 3.96 108 3.83 - 4.23 0.446 3.96 0.65 80 REP 1 0 7.81 0.00 0.647 5.06 6 6.82 - 0.99 0.647 5.06 12 6.91 - 0.90 0.647 5.06 18 6.15 - 1.66 0.647 5.06 24 5.95 - 1.86 0.647 5.06 30 4.42 - 3.39 0.647 5.06 36 3.16 - 4.65 0.647 5.06 42 2.98 - 4.83 0.647 5.06 48 4.25 - 3.56 0.647 5.06 REP 2 0 7.20 0.00 0.66 5.60 6 6.97 - 0.22 0.66 5.60 12 5.82 - 1.38 0.66 5.60 18 4.66 - 2.54 0.66 5.60 24 3.46 - 3.73 0.66 5.60 30 4.05 - 3.15 0.66 5.60 36 3.23 - 3.96 0.66 5.60 128 Table F . 1 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.65 80 REP 2 42 2.10 - 5.10 0.66 5.60 48 3.79 - 3.41 0.66 5.60 REP 3 0 6.87 0.00 0.638 5.58 6 5.44 - 1.43 0.638 5.58 12 5.75 - 1.12 0.638 5.58 18 5.34 - 1.53 0.638 5.58 24 3.92 - 2.95 0.638 5.58 30 4.06 - 2.81 0.638 5.58 36 3.64 - 3.23 0.638 5.58 42 4.47 - 2.40 0.638 5.58 0.25 85 REP 1 0 7.61 0.00 0.258 3.02 5 7.30 - 0.31 0.258 3.02 10 6.35 - 1.26 0.258 3.02 15 5.51 - 2.10 0.258 3.02 20 6.19 - 1.42 0.258 3.02 25 4.24 - 3.37 0.258 3.02 30 2.48 - 5.12 0.258 3.02 35 3.26 - 4.35 0.258 3.02 40 2.79 - 4.82 0.258 3.02 REP 2 0 6.82 0.00 0.265 3.25 5 6.12 - 0.71 0.265 3.25 10 6.53 - 0.29 0.265 3.25 15 6.70 - 0.12 0.265 3.25 20 5.92 - 0.90 0.265 3.25 25 5.72 - 1.11 0.265 3.25 30 4.05 - 2.77 0.265 3.25 35 3.91 - 2.92 0.265 3.25 40 3.35 - 3.47 0.265 3.25 REP 3 0 7.17 0.00 0.234 3.03 5 7.31 0.14 0.234 3.03 10 5.87 - 1.30 0.234 3.03 15 5.97 - 1.20 0.234 3.03 20 5.39 - 1.78 0.234 3.03 25 5.37 - 1.80 0.234 3.03 30 4.77 - 2.40 0.234 3.03 35 4.11 - 3.06 0.234 3.03 40 4.67 - 2.50 0.234 3.03 0.45 85 REP 1 0 7.96 0.00 0.428 4.32 7 6.07 - 1.89 0.428 4.32 14 6.62 - 1.34 0.428 4.32 21 6.26 - 1.70 0.428 4.32 28 5.60 - 2.36 0.428 4.32 129 Table F . 1 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 1 35 4.14 - 3.82 0.428 4.32 42 4.45 - 3.50 0.428 4.32 49 3.67 - 4.29 0.428 4.32 REP 2 0 7.80 0.00 0.454 3.65 7 7.00 - 0.79 0.454 3.65 14 5.76 - 2.03 0.454 3.65 21 5.39 - 2.41 0.454 3.65 28 5.62 - 2.17 0.454 3.65 35 4.81 - 2.98 0.454 3.65 42 4.89 - 2.90 0.454 3.65 49 3.40 - 4.40 0.454 3.65 56 2.54 - 5.25 0.454 3.65 REP 3 0 8.09 0.00 0.44 3.92 7 6.60 - 1.50 0.44 3.92 14 6.60 - 1.50 0.44 3.92 21 6.34 - 1.75 0.44 3.92 28 5.18 - 2.92 0.44 3.92 35 3.53 - 4.56 0.44 3.92 0.65 85 REP 1 0 7.44 0.00 0.647 5.06 1 6.97 - 0.47 0.647 5.06 2 6.54 - 0.90 0.647 5.06 3 5.96 - 1.48 0.647 5.06 4 6.30 - 1.14 0.647 5.06 5 6.14 - 1.30 0.647 5.06 6 4.43 - 3.01 0.647 5.06 7 5.66 - 1.78 0.647 5.06 8 6.35 - 1.09 0.647 5.06 REP 2 0 6.99 0.00 0.66 5.60 1 7.06 0.07 0.66 5.60 2 6.27 - 0.71 0.66 5.60 3 6.33 - 0.65 0.66 5.60 4 5.73 - 1.25 0.66 5.60 5 3.74 - 3.25 0.66 5.60 6 5.49 - 1.50 0.66 5.60 7 4.37 - 2.62 0.66 5.60 REP 3 0 6.37 0.00 0.638 5.58 1 6.23 - 0.15 0.638 5.58 2 6.07 - 0.31 0.638 5.58 3 6.04 - 0.34 0.638 5.58 4 5.57 - 0.81 0.638 5.58 5 3.26 - 3.12 0.638 5.58 6 4.14 - 2.24 0.638 5.58 130 Table F . 1 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.65 85 REP 3 7 4.93 - 1.44 0.638 5.58 8 3.19 - 3.18 0.638 5.58 0.25 90 REP 1 0 7.08 0.00 0.258 3.02 3 6.27 - 0.81 0.258 3.02 6 5.42 - 1.66 0.258 3.02 9 4.70 - 2.37 0.258 3.02 12 2.54 - 4.53 0.258 3.02 15 3.20 - 3.87 0.258 3.02 18 2.95 - 4.12 0.258 3.02 21 3.68 - 3.40 0.258 3.02 24 3.18 - 3.90 0.258 3.02 REP 2 0 6.64 0.00 0.265 3.25 3 6.56 - 0.08 0.265 3.25 6 5.35 - 1.29 0.265 3.25 9 3.94 - 2.70 0.265 3.25 12 3.65 - 2.99 0.265 3.25 15 4.37 - 2.28 0.265 3.25 18 3.90 - 2.74 0.265 3.25 24 2.51 - 4.14 0.265 3.25 REP 3 0 6.47 0.00 0.234 3.03 3 6.60 0.13 0.234 3.03 6 5.69 - 0.78 0.234 3.03 9 5.35 - 1.12 0.234 3.03 12 5.12 - 1.35 0.234 3.03 15 4.64 - 1.83 0.234 3.03 18 3.44 - 3.03 0.234 3.03 21 2.70 - 3.77 0.234 3.03 24 3.26 - 3.21 0.234 3.03 0.45 90 REP 1 0 6.72 0.00 0.428 4.32 2.5 7.04 0.31 0.428 4.32 5 6.11 - 0.62 0.428 4.32 7.5 5.96 - 0.77 0.428 4.32 10 5.60 - 1.12 0.428 4.32 12.5 4.97 - 1.75 0.428 4.32 15 4.43 - 2.29 0.428 4.32 17.5 2.81 - 3.91 0.428 4.32 20 3.32 - 3.40 0.428 4.32 REP 2 0 6.85 0.00 0.454 3.65 2.5 6.61 - 0.23 0.454 3.65 5 5.09 - 1.76 0.454 3.65 7.5 5.96 - 0.77 0.428 4.32 10 5.60 - 1.12 0.428 4.32 131 Table F . 1 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 90 REP 2 12.5 3.89 - 2.96 0.454 3.65 15 4.50 - 2.35 0.454 3.65 20 4.73 - 2.11 0.454 3.65 REP 3 0 7.50 0.00 0.44 3.93 2.5 6.58 - 0.92 0.44 3.93 5 5.19 - 2.31 0.44 3.93 7.5 5.03 - 2.47 0.44 3.93 10 3.77 - 3.72 0.44 3.93 15 3.23 - 4.27 0.44 3.93 17.5 3.57 - 3.93 0.44 3.93 20 3.33 - 4.17 0.44 3.93 0.65 90 REP 1 0 6.51 0.00 0.647 5.06 0.5 6.70 0.19 0.647 5.06 1 5.80 - 0.71 0.647 5.06 1.5 5.72 - 0.79 0.647 5.06 2 5.18 - 1.33 0.647 5.06 2.5 4.16 - 2.35 0.647 5.06 3 3.61 - 2.90 0.647 5.06 4 2.37 - 4.14 0.647 5.06 REP 2 0 4.81 0.00 0.66 5.60 0.5 5.31 0.51 0.66 5.60 1 4.71 - 0.09 0.66 5.60 1.5 4.49 - 0.31 0.66 5.60 2 3.48 - 1.33 0.66 5.60 2.5 3.18 - 1.63 0.66 5.60 3 3.70 - 1.11 0.66 5.60 4 2.30 - 2.51 0.66 5.60 REP 3 0 5.89 0.00 0.638 5.58 0.5 6.25 0.36 0.638 5.58 1 3.84 - 2.06 0.638 5.58 1.5 3.67 - 2.22 0.638 5.58 2 4.04 - 1.86 0.638 5.58 2.5 4.04 - 1.86 0.638 5.58 3 3.97 - 1.93 0.638 5.58 3.5 3.12 - 2.78 0.638 5.58 4 1.54 - 4.35 0.638 5.58 132 Table F . 2 Salmonella inactivation data for almond meal. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 8.02 0.00 0.266 2.00 24 7.59 - 0.43 0.266 2.00 48 7.30 - 0.72 0.266 2.00 72 7.08 - 0.93 0.266 2.00 96 6.70 - 1.31 0.266 2.00 120 6.46 - 1.55 0.266 2.00 144 5.95 - 2.07 0.266 2.00 168 5.84 - 2.18 0.266 2.00 192 5.31 - 2.70 0.266 2.00 REP 2 0 8.00 0.00 0.257 1.75 24 7.74 - 0.26 0.257 1.75 48 7.32 - 0.68 0.257 1.75 72 7.06 - 0.94 0.257 1.75 96 6.84 - 1.16 0.257 1.75 120 6.78 - 1.22 0.257 1.75 144 6.35 - 1.65 0.257 1.75 168 6.05 - 1.96 0.257 1.75 192 5.87 - 2.13 0.257 1.75 REP 3 0 8.11 0.00 0.242 2.86 24 7.58 - 0.53 0.242 2.86 48 7.17 - 0.94 0.242 2.86 72 6.74 - 1.37 0.242 2.86 96 6.40 - 1.71 0.242 2.86 120 6.06 - 2.04 0.242 2.86 144 5.81 - 2.30 0.242 2.86 168 5.22 - 2.88 0.242 2.86 192 5.17 - 2.94 0.242 2.86 0.45 80 REP 1 0 8.28 0.00 0.435 3.94 22 7.49 - 0.79 0.435 3.94 44 7.16 - 1.12 0.435 3.94 66 6.76 - 1.52 0.435 3.94 88 6.30 - 1.98 0.435 3.94 110 6.18 - 2.10 0.435 3.94 132 5.70 - 2.58 0.435 3.94 154 5.26 - 3.02 0.435 3.94 176 5.15 - 3.12 0.435 3.94 REP 2 0 7.69 0.00 0.443 3.74 22 7.39 - 0.30 0.443 3.74 44 6.70 - 0.99 0.443 3.74 66 6.31 - 1.37 0.443 3.74 88 5.70 - 1.99 0.443 3.74 110 5.21 - 2.48 0.443 3.74 133 Table F . 2 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 80 REP 2 132 4.89 - 2.79 0.443 3.74 154 4.20 - 3.48 0.443 3.74 176 3.93 - 3.75 0.443 3.74 REP 3 0 8.36 0.00 0.432 4.05 44 7.06 - 1.30 0.432 4.05 66 6.74 - 1.62 0.432 4.05 88 6.00 - 2.36 0.432 4.05 110 5.21 - 3.15 0.432 4.05 132 4.62 - 3.74 0.432 4.05 154 5.20 - 3.16 0.432 4.05 0.65 80 REP 1 0 7.87 0.00 0.640 6.06 10 6.77 - 1.09 0.640 6.06 20 6.85 - 1.02 0.640 6.06 30 6.19 - 1.68 0.640 6.06 40 5.79 - 2.08 0.640 6.06 50 5.21 - 2.66 0.640 6.06 70 4.41 - 3.45 0.640 6.06 80 3.02 - 4.85 0.640 6.06 REP 2 0 7.99 0.00 0.639 6.06 10 7.34 - 0.65 0.639 5.84 20 6.61 - 1.38 0.639 5.84 30 6.04 - 1.95 0.639 5.84 40 5.51 - 2.49 0.639 5.84 50 5.18 - 2.81 0.639 5.84 60 4.44 - 3.55 0.639 5.84 70 4.43 - 3.56 0.639 5.84 80 3.65 - 4.34 0.639 5.84 REP 3 0 7.99 0.00 0.639 5.44 10 7.32 - 0.67 0.639 5.44 20 6.82 - 1.17 0.639 5.44 30 6.29 - 1.71 0.639 5.44 40 5.88 - 2.12 0.639 5.44 50 5.32 - 2.68 0.639 5.44 60 5.01 - 2.99 0.639 5.44 70 4.84 - 3.15 0.639 5.44 80 4.32 - 3.67 0.639 5.44 0.25 85 REP 1 0 7.95 0.00 0.266 2.00 19 7.26 - 0.69 0.266 2.00 38 6.70 - 1.25 0.266 2.00 57 6.35 - 1.60 0.266 2.00 76 5.98 - 1.96 0.266 2.00 95 5.36 - 2.59 0.266 2.00 134 Table F . 2 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 85 REP 1 114 5.13 - 2.82 0.266 2.00 133 4.91 - 3.04 0.266 2.00 152 4.01 - 3.93 0.266 2.00 REP 2 0 7.85 0.00 0.257 1.75 19 7.78 - 0.07 0.257 1.75 38 6.97 - 0.89 0.257 1.75 57 6.60 - 1.25 0.257 1.75 76 6.11 - 1.74 0.257 1.75 95 5.88 - 1.97 0.257 1.75 114 5.38 - 2.47 0.257 1.75 133 5.30 - 2.55 0.257 1.75 152 4.94 - 2.91 0.257 1.75 REP 3 0 7.84 0.00 0.242 2.86 19 7.28 - 0.57 0.242 2.86 57 5.94 - 1.90 0.242 2.86 76 5.48 - 2.36 0.242 2.86 95 5.01 - 2.83 0.242 2.86 114 4.77 - 3.07 0.242 2.86 133 4.11 - 3.73 0.242 2.86 152 3.74 - 4.10 0.242 2.86 0.45 85 REP 1 0 8.15 0.00 0.435 3.94 12 7.61 - 0.53 0.435 3.94 24 7.01 - 1.13 0.435 3.94 36 6.34 - 1.80 0.435 3.94 60 5.16 - 2.98 0.435 3.94 72 5.47 - 2.68 0.435 3.94 84 4.33 - 3.82 0.435 3.94 96 4.24 - 3.91 0.435 3.94 REP 2 0 7.72 0.00 0.443 3.74 12 7.13 - 0.59 0.443 3.74 24 6.59 - 1.13 0.443 3.74 36 6.02 - 1.70 0.443 3.74 48 5.37 - 2.35 0.443 3.74 60 4.84 - 2.88 0.443 3.74 72 4.39 - 3.33 0.443 3.74 84 3.98 - 3.74 0.443 3.74 96 3.45 - 4.27 0.443 3.74 0.45 85 REP 3 0 8.06 0.00 0.432 4.05 12 7.30 - 0.76 0.432 4.05 24 5.30 - 2.76 0.432 4.05 36 5.79 - 2.27 0.432 4.05 48 5.81 - 2.25 0.432 4.05 135 Table F . 2 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 3 60 3.57 - 4.48 0.432 4.05 84 4.11 - 3.94 0.432 4.05 96 3.78 - 4.28 0.432 4.05 0.65 85 REP 1 0 8.00 0.00 0.639 5.84 4 7.24 - 0.75 0.639 5.84 8 6.38 - 1.62 0.639 5.84 12 5.83 - 2.17 0.639 5.84 16 6.13 - 1.86 0.639 5.84 20 4.74 - 3.26 0.639 5.84 24 4.38 - 3.62 0.639 5.84 28 3.56 - 4.44 0.639 5.84 32 3.32 - 4.68 0.639 5.84 REP 2 0 7.74 0.00 0.639 5.44 4 7.00 - 0.74 0.639 5.44 8 6.65 - 1.09 0.639 5.44 12 5.96 - 1.78 0.639 5.44 16 5.35 - 2.39 0.639 5.44 20 3.62 - 4.12 0.639 5.44 24 4.34 - 3.40 0.639 5.44 28 4.09 - 3.65 0.639 5.44 32 3.52 - 4.22 0.639 5.44 REP 3 0 7.51 0.00 0.644 5.72 4 6.83 - 0.68 0.644 5.72 8 6.54 - 0.97 0.644 5.72 12 5.79 - 1.71 0.644 5.72 16 5.05 - 2.46 0.644 5.72 20 4.59 - 2.91 0.644 5.72 24 4.30 - 3.20 0.644 5.72 28 3.93 - 3.58 0.644 5.72 32 3.65 - 3.85 0.644 5.72 0.25 90 REP 1 0 7.92 0.00 0.266 2.00 12 6.98 - 0.94 0.266 2.00 24 6.29 - 1.63 0.266 2.00 36 5.72 - 2.20 0.266 2.00 48 4.97 - 2.95 0.266 2.00 60 4.66 - 3.26 0.266 2.00 72 3.96 - 3.96 0.266 2.00 84 3.49 - 4.43 0.266 2.00 96 3.31 - 4.61 0.266 2.00 136 Table F . 2 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 90 REP 2 0 7.85 0.00 0.257 1.75 12 7.00 - 0.85 0.257 1.75 24 6.36 - 1.50 0.257 1.75 36 5.83 - 2.02 0.257 1.75 48 5.74 - 2.11 0.257 1.75 60 5.24 - 2.61 0.257 1.75 72 4.60 - 3.25 0.257 1.75 84 4.45 - 3.40 0.257 1.75 96 4.16 - 3.69 0.257 1.75 REP 3 0 7.92 0.00 0.242 2.86 12 6.92 - 1.01 0.242 2.86 24 6.32 - 1.60 0.242 2.86 36 5.36 - 2.56 0.242 2.86 48 4.87 - 3.06 0.242 2.86 60 4.17 - 3.75 0.242 2.86 72 3.40 - 4.53 0.242 2.86 84 3.16 - 4.76 0.242 2.86 96 2.40 - 5.53 0.242 2.86 0.45 90 REP 1 0 7.93 0.00 0.435 3.94 5 7.22 - 0.71 0.435 3.94 10 6.92 - 1.01 0.435 3.94 15 5.32 - 2.60 0.435 3.94 20 5.82 - 2.11 0.435 3.94 25 4.50 - 3.43 0.435 3.94 30 4.46 - 3.46 0.435 3.94 35 4.52 - 3.41 0.435 3.94 40 4.48 - 3.44 0.435 3.94 REP 2 0 7.79 0.00 0.442 3.74 5 6.78 - 1.01 0.442 3.74 10 6.10 - 1.69 0.442 3.74 15 5.66 - 2.13 0.442 3.74 20 5.09 - 2.70 0.442 3.74 25 4.73 - 3.06 0.442 3.74 30 4.23 - 3.56 0.442 3.74 35 4.00 - 3.79 0.442 3.74 40 3.37 - 4.42 0.442 3.74 REP 3 0 7.95 0.00 0.432 4.05 5 7.03 - 0.92 0.432 4.05 10 6.71 - 1.25 0.432 4.05 15 6.19 - 1.76 0.432 4.05 20 5.60 - 2.36 0.432 4.05 25 5.53 - 2.42 0.432 4.05 137 Table F . 2 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 90 REP 3 30 4.38 - 3.57 0.432 4.05 35 4.04 - 3.92 0.432 4.05 40 3.28 - 4.68 0.432 4.05 0.65 90 REP 1 0 7.15 0.00 0.639 5.84 1 6.80 - 0.35 0.639 5.84 2 6.45 - 0.70 0.639 5.84 3 6.07 - 1.08 0.639 5.84 4 5.66 - 1.49 0.639 5.84 5 5.51 - 1.64 0.639 5.84 6 4.93 - 2.22 0.639 5.84 7 4.35 - 2.80 0.639 5.84 8 4.33 - 2.82 0.639 5.84 REP 2 0 7.40 0.00 0.639 5.44 1 6.76 - 0.64 0.639 5.44 2 6.68 - 0.72 0.639 5.44 3 6.19 - 1.21 0.639 5.44 4 6.00 - 1.40 0.639 5.44 5 5.49 - 1.91 0.639 5.44 6 4.99 - 2.41 0.639 5.44 7 4.24 - 3.16 0.639 5.44 8 4.21 - 3.19 0.639 5.44 REP 3 0 7.17 0.00 0.644 5.72 1 6.81 - 0.36 0.644 5.72 2 6.45 - 0.72 0.644 5.72 3 6.13 - 1.04 0.644 5.72 4 5.54 - 1.62 0.644 5.72 5 5.42 - 1.75 0.644 5.72 6 5.09 - 2.08 0.644 5.72 7 4.66 - 2.51 0.644 5.72 8 4.39 - 2.78 0.644 5.72 138 Table F . 3 Salmonella inactivation data for almond butter. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 8.49 0.00 0.243 2.98 24 8.00 - 0.49 0.243 2.98 48 7.59 - 0.90 0.243 2.98 72 7.28 - 1.21 0.243 2.98 96 6.92 - 1.57 0.243 2.98 120 5.71 - 2.78 0.243 2.98 144 6.40 - 2.09 0.243 2.98 168 6.05 - 2.44 0.243 2.98 192 5.68 - 2.81 0.243 2.98 REP 2 0 8.26 0.00 0.260 2.90 24 8.91 0.64 0.260 2.90 48 7.81 - 0.45 0.260 2.90 72 7.50 - 0.76 0.260 2.90 96 7.38 - 0.88 0.260 2.90 120 6.94 - 1.32 0.260 2.90 144 5.97 - 2.29 0.260 2.90 168 6.25 - 2.01 0.260 2.90 192 5.61 - 2.65 0.260 2.90 REP 3 0 8.29 0.00 0.250 2.69 24 7.84 - 0.45 0.250 2.69 48 7.09 - 1.20 0.250 2.69 72 7.05 - 1.24 0.250 2.69 96 6.55 - 1.74 0.250 2.69 120 6.14 - 2.15 0.250 2.69 144 5.93 - 2.36 0.250 2.69 168 5.41 - 2.87 0.250 2.69 192 4.34 - 3.95 0.250 2.69 0.45 80 REP 1 0 7.71 0.00 0.428 4.35 22 6.69 - 1.02 0.428 4.35 44 6.02 - 1.69 0.428 4.35 66 5.36 - 2.35 0.428 4.35 88 5.14 - 2.58 0.428 4.35 110 4.27 - 3.44 0.428 4.35 132 4.07 - 3.64 0.428 4.35 154 3.63 - 4.08 0.428 4.35 176 3.28 - 4.43 0.428 4.35 REP 2 0 7.27 0.00 0.443 4.46 22 5.50 - 1.77 0.443 4.46 44 5.62 - 1.65 0.443 4.46 66 4.99 - 2.29 0.443 4.46 88 3.92 - 3.36 0.443 4.46 110 3.77 - 3.50 0.443 4.46 139 Table F . 3 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0. 4 5 80 REP 2 132 3.32 - 3.95 0.443 4.46 154 2.13 - 5.14 0.443 4.46 REP 3 0 7.46 0.00 0.437 4.18 22 6.58 - 0.88 0.437 4.18 44 6.25 - 1.21 0.437 4.18 66 6.20 - 1.26 0.437 4.18 88 5.83 - 1.63 0.437 4.18 110 5.74 - 1.72 0.437 4.18 132 5.11 - 2.35 0.437 4.18 154 5.18 - 2.28 0.437 4.18 176 4.83 - 2.63 0.437 4.18 0.65 80 REP 1 0 7.61 0.00 0.648 5.94 8 6.90 - 0.72 0.648 5.94 16 6.27 - 1.34 0.648 5.94 24 5.78 - 1.83 0.648 5.94 32 5.11 - 2.51 0.648 5.94 40 4.63 - 2.98 0.648 5.94 48 4.33 - 3.28 0.648 5.94 56 3.87 - 3.74 0.648 5.94 64 3.96 - 3.66 0.648 5.94 REP 2 0 7.46 0.00 0.634 5.91 8 6.62 - 0.84 0.634 5.91 16 5.98 - 1.49 0.634 5.91 24 5.45 - 2.01 0.634 5.91 32 4.57 - 2.89 0.634 5.91 40 3.98 - 3.48 0.634 5.91 48 3.36 - 4.10 0.634 5.91 56 2.24 - 5.22 0.634 5.91 64 2.36 - 5.10 0.634 5.91 REP 3 0 7.53 0.00 0.649 5.94 8 6.83 - 0.70 0.649 5.94 16 6.12 - 1.41 0.649 5.94 24 5.50 - 2.03 0.649 5.94 32 4.96 - 2.57 0.649 5.94 40 4.16 - 3.37 0.649 5.94 48 3.90 - 3.63 0.649 5.94 56 3.23 - 4.29 0.649 5.94 64 2.60 - 4.93 0.649 5.94 0.25 85 REP 1 0 8.65 0.00 0.234 3.04 19 7.96 - 0.68 0.234 3.04 38 7.36 - 1.29 0.234 3.04 58 5.72 - 2.92 0.234 3.04 140 Table F . 3 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 85 REP 1 76 6.10 - 2.55 0.234 3.04 95 5.91 - 2.74 0.234 3.04 114 5.20 - 3.45 0.234 3.04 133 4.85 - 3.79 0.234 3.04 152 4.10 - 4.55 0.234 3.04 REP 2 0 8.46 0.00 0.249 2.98 19 7.61 - 0.85 0.249 2.98 38 7.14 - 1.32 0.249 2.98 57 6.48 - 1.99 0.249 2.98 76 6.00 - 2.46 0.249 2.98 95 5.70 - 2.76 0.249 2.98 114 5.12 - 3.34 0.249 2.98 133 4.54 - 3.92 0.249 2.98 152 4.09 - 4.37 0.249 2.98 REP 3 0 8.09 0.00 0.25 2.69 19 7.44 - 0.65 0.25 2.69 38 6.88 - 1.21 0.25 2.69 57 6.24 - 1.84 0.25 2.69 76 5.61 - 2.48 0.25 2.69 95 4.96 - 3.12 0.25 2.69 114 4.52 - 3.57 0.25 2.69 133 4.37 - 3.71 0.25 2.69 152 3.97 - 4.12 0.25 2.69 0.45 85 REP 1 0 7.51 0.00 0.438 4.35 12 6.27 - 1.24 0.438 4.35 24 5.71 - 1.80 0.438 4.35 36 5.02 - 2.49 0.438 4.35 48 3.80 - 3.71 0.438 4.35 60 3.54 - 3.97 0.438 4.35 72 3.44 - 4.07 0.438 4.35 84 2.85 - 4.67 0.438 4.35 96 3.27 - 4.24 0.438 4.35 REP 2 0 7.15 0.00 0.443 4.46 12 6.20 - 0.95 0.443 4.46 24 5.37 - 1.78 0.443 4.46 36 4.84 - 2.32 0.443 4.46 48 4.07 - 3.08 0.443 4.46 60 3.90 - 3.25 0.443 4.46 72 3.04 - 4.11 0.443 4.46 84 2.65 - 4.50 0.443 4.46 96 2.08 - 5.07 0.443 4.46 141 Table F . 3 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 3 0 7.34 0.00 0.437 4.18 12 6.59 - 0.76 0.437 4.18 24 6.27 - 1.08 0.437 4.18 36 5.99 - 1.35 0.437 4.18 48 5.70 - 1.64 0.437 4.18 60 5.31 - 2.03 0.437 4.18 72 4.84 - 2.50 0.437 4.18 84 4.61 - 2.73 0.437 4.18 96 4.29 - 3.05 0.437 4.18 0.65 85 REP 1 0 7.16 0.00 0.648 5.94 3 6.19 - 0.97 0.648 5.94 6 5.61 - 1.55 0.648 5.94 9 5.86 - 1.30 0.648 5.94 12 5.49 - 1.67 0.648 5.94 15 4.29 - 2.87 0.648 5.94 18 3.72 - 3.44 0.648 5.94 21 3.01 - 4.15 0.648 5.94 24 2.45 - 4.71 0.648 5.94 REP 2 0 7.14 0.00 0.634 5.95 3 6.16 - 0.98 0.634 5.95 6 5.48 - 1.65 0.634 5.95 9 4.70 - 2.43 0.634 5.95 12 3.85 - 3.28 0.634 5.95 15 3.28 - 3.86 0.634 5.95 18 3.57 - 3.57 0.634 5.95 21 2.24 - 4.89 0.634 5.95 REP 3 0 7.28 0.00 0.649 5.94 3 6.30 - 0.98 0.649 5.94 6 5.53 - 1.75 0.649 5.94 9 4.98 - 2.30 0.649 5.94 12 4.36 - 2.91 0.649 5.94 15 3.71 - 3.57 0.649 5.94 18 2.86 - 4.41 0.649 5.94 21 1.78 - 5.50 0.649 5.94 24 1.65 - 5.62 0.649 5.94 0.25 90 REP 1 0 8.51 0.00 0.234 3.04 12 7.54 - 0.97 0.234 3.04 24 6.84 - 1.67 0.234 3.04 36 5.86 - 2.65 0.234 3.04 48 5.74 - 2.77 0.234 3.04 60 5.28 - 3.23 0.234 3.04 72 4.65 - 3.86 0.234 3.04 142 Table F . 3 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 90 REP 1 84 3.36 - 5.15 0.234 3.04 96 2.88 - 5.63 0.234 3.04 REP 2 0 8.22 0.00 0.249 2.98 12 7.54 - 0.69 0.249 2.98 24 6.67 - 1.55 0.249 2.98 36 6.02 - 2.20 0.249 2.98 48 4.81 - 3.41 0.249 2.98 60 4.58 - 3.64 0.249 2.98 72 3.54 - 4.68 0.249 2.98 84 3.43 - 4.79 0.249 2.98 96 3.11 - 5.11 0.249 2.98 REP 3 0 7.82 0.00 0.250 2.69 12 7.12 - 0.70 0.250 2.69 24 6.22 - 1.60 0.250 2.69 36 5.76 - 2.06 0.250 2.69 48 4.87 - 2.95 0.250 2.69 60 4.40 - 3.42 0.250 2.69 72 3.86 - 3.96 0.250 2.69 87 3.24 - 4.58 0.250 2.69 96 2.93 - 4.89 0.250 2.69 0.45 90 REP 1 0 7.03 0.00 0.446 4.23 4 6.21 - 0.82 0.446 4.23 8 5.59 - 1.44 0.446 4.23 12 4.92 - 2.11 0.446 4.23 16 4.41 - 2.62 0.446 4.23 20 3.85 - 3.19 0.446 4.23 24 3.24 - 3.79 0.446 4.23 28 2.94 - 4.09 0.446 4.23 32 2.28 - 4.75 0.446 4.23 REP 2 0 7.18 0.00 0.443 4.46 4 6.31 - 0.88 0.443 4.46 8 5.72 - 1.46 0.443 4.46 12 5.58 - 1.60 0.443 4.46 16 4.75 - 2.44 0.443 4.46 20 4.32 - 2.86 0.443 4.46 24 3.71 - 3.47 0.443 4.46 28 3.29 - 3.89 0.443 4.46 32 2.31 - 4.87 0.443 4.46 REP 3 0 7.34 0.00 0.437 4.18 4 6.59 - 0.76 0.437 4.18 8 6.27 - 1.08 0.437 4.18 12 5.99 - 1.35 0.437 4.18 143 Table F . 3 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 90 REP 3 16 5.70 - 1.64 0.437 4.18 20 5.31 - 2.03 0.437 4.18 24 4.84 - 2.50 0.437 4.18 28 4.61 - 2.73 0.437 4.18 32 4.29 - 3.05 0.437 4.18 0.65 90 REP 1 0 6.92 0.00 0.648 5.94 1 6.17 - 0.75 0.648 5.94 2 5.59 - 1.33 0.648 5.94 3 4.90 - 2.02 0.648 5.94 4 4.27 - 2.65 0.648 5.94 5 3.77 - 3.14 0.648 5.94 6 3.10 - 3.82 0.648 5.94 7 2.96 - 3.96 0.648 5.94 8 2.77 - 4.15 0.648 5.94 REP 2 0 6.78 0.00 0.655 5.95 1 6.17 - 0.61 0.655 5.95 2 5.30 - 1.48 0.655 5.95 3 4.68 - 2.10 0.655 5.95 4 4.19 - 2.59 0.655 5.95 5 3.45 - 3.33 0.655 5.95 6 2.93 - 3.85 0.655 5.95 7 2.60 - 4.18 0.655 5.95 8 1.65 - 5.12 0.655 5.95 REP 3 0 6.70 0.00 0.649 5.94 1 5.70 - 1.00 0.649 5.94 2 5.03 - 1.67 0.649 5.94 3 4.53 - 2.17 0.649 5.94 4 3.53 - 3.18 0.649 5.94 5 2.68 - 4.02 0.649 5.94 6 2.00 - 4.70 0.649 5.94 7 2.18 - 4.53 0.649 5.94 144 Table F . 4 Salmonella inactivation data for wheat kernels. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 7.87 0.00 0.256 7.99 10 7.46 - 0.40 0.256 7.99 20 6.62 - 1.25 0.256 7.99 30 6.06 - 1.81 0.256 7.99 40 5.95 - 1.92 0.256 7.99 50 5.07 - 2.80 0.256 7.99 60 5.19 - 2.68 0.256 7.99 70 4.88 - 2.99 0.256 7.99 80 2.93 - 4.93 0.256 7.99 REP 2 0 8.84 0.00 0.253 7.99 10 8.44 - 0.41 0.253 7.99 20 8.10 - 0.75 0.253 7.99 30 7.75 - 1.10 0.253 7.99 40 7.12 - 1.72 0.253 7.99 50 6.47 - 2.37 0.253 7.99 60 6.36 - 2.49 0.253 7.99 70 6.09 - 2.75 0.253 7.99 80 4.67 - 4.18 0.253 7.99 REP 3 0 9.08 0.00 0.253 8.12 10 8.10 - 0.97 0.253 8.12 20 7.61 - 1.47 0.253 8.12 30 7.06 - 2.01 0.253 8.12 40 6.85 - 2.22 0.253 8.12 50 6.31 - 2.77 0.253 8.12 60 5.74 - 3.33 0.253 8.12 70 5.51 - 3.57 0.253 8.12 80 4.71 - 4.37 0.253 8.12 0.45 80 REP 1 0 9.03 0.00 0.451 9.80 5 8.43 - 0.60 0.451 9.80 10 8.31 - 0.73 0.451 9.80 15 8.03 - 1.00 0.451 9.80 20 7.54 - 1.49 0.451 9.80 25 7.25 - 1.78 0.451 9.80 30 6.33 - 2.70 0.451 9.80 35 6.27 - 2.76 0.451 9.80 40 6.40 - 2.63 0.451 9.80 REP 2 0 8.74 0.00 0.440 10.04 5 7.99 - 0.75 0.440 10.04 10 7.31 - 1.43 0.440 10.04 15 7.37 - 1.38 0.440 10.04 20 6.77 - 1.97 0.440 10.04 25 6.12 - 2.62 0.440 10.04 145 Table F . 4 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 80 REP 2 30 5.07 - 3.68 0.440 10.04 35 4.96 - 3.78 0.440 10.04 40 4.18 - 4.57 0.440 10.04 REP 3 0 8.50 0.00 0.450 10.15 5 8.07 - 0.43 0.450 10.15 10 7.57 - 0.93 0.450 10.15 15 6.86 - 1.65 0.450 10.15 20 6.32 - 2.18 0.450 10.15 25 5.82 - 2.69 0.450 10.15 30 5.40 - 3.10 0.450 10.15 35 4.74 - 3.76 0.450 10.15 40 4.10 - 4.40 0.450 10.15 0.65 80 REP 1 0 8.62 0.00 0.651 11.73 3 7.47 - 1.15 0.651 11.73 6 7.02 - 1.60 0.651 11.73 9 5.91 - 2.71 0.651 11.73 12 5.53 - 3.08 0.651 11.73 15 5.25 - 3.37 0.651 11.73 18 3.21 - 5.41 0.651 11.73 21 2.44 - 6.17 0.651 11.73 24 1.31 - 7.31 0.651 11.73 REP 2 0 8.74 0.00 0.655 12.83 3 8.41 - 0.32 0.655 12.83 6 7.96 - 0.77 0.655 12.83 9 7.19 - 1.55 0.655 12.83 12 6.55 - 2.18 0.655 12.83 15 5.97 - 2.76 0.655 12.83 18 4.93 - 3.80 0.655 12.83 21 4.30 - 4.44 0.655 12.83 24 2.69 - 6.04 0.655 12.83 REP 3 0 8.83 0.00 0.648 12.06 3 7.32 - 1.51 0.648 12.06 6 7.58 - 1.26 0.648 12.06 9 7.21 - 1.62 0.648 12.06 12 6.60 - 2.23 0.648 12.06 15 4.86 - 3.97 0.648 12.06 18 3.49 - 5.34 0.648 12.06 21 3.62 - 5.21 0.648 12.06 24 2.32 - 6.51 0.648 12.06 0.25 85 REP 1 0 7.52 0.00 0.256 7.99 6 7.05 - 0.47 0.256 7.99 12 6.19 - 1.33 0.256 7.99 146 Table F . 4 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 85 REP 1 18 5.99 - 1.53 0.256 7.99 24 5.54 - 1.98 0.256 7.99 30 4.46 - 3.07 0.256 7.99 36 4.34 - 3.18 0.256 7.99 42 4.55 - 2.97 0.256 7.99 48 2.71 - 4.81 0.256 7.99 REP 2 0 8.61 0.00 0.253 7.99 6 8.14 - 0.47 0.253 7.99 12 7.12 - 1.48 0.253 7.99 18 6.95 - 1.66 0.253 7.99 24 5.92 - 2.69 0.253 7.99 30 5.54 - 3.07 0.253 7.99 36 4.78 - 3.83 0.253 7.99 42 4.60 - 4.01 0.253 7.99 48 3.61 - 4.99 0.253 7.99 REP 3 0 8.85 0.00 0.253 8.12 6 7.92 - 0.93 0.253 8.12 12 7.23 - 1.62 0.253 8.12 18 6.17 - 2.68 0.253 8.12 24 5.64 - 3.21 0.253 8.12 30 5.51 - 3.34 0.253 8.12 36 4.66 - 4.19 0.253 8.12 42 5.03 - 3.82 0.253 8.12 48 3.53 - 5.32 0.253 8.12 0.45 85 REP 1 0 8.45 0.00 0.451 9.80 2 8.60 0.15 0.451 9.80 4 7.79 - 0.66 0.451 9.80 6 7.38 - 1.07 0.451 9.80 8 7.23 - 1.22 0.451 9.80 10 6.14 - 2.31 0.451 9.80 12 5.94 - 2.52 0.451 9.80 14 5.02 - 3.43 0.451 9.80 16 4.10 - 4.35 0.451 9.80 REP 2 0 8.39 0.00 0.440 10.04 2 8.05 - 0.34 0.440 10.04 4 7.41 - 0.98 0.440 10.04 6 6.97 - 1.42 0.440 10.04 8 6.53 - 1.86 0.440 10.04 10 5.34 - 3.05 0.440 10.04 12 5.00 - 3.39 0.440 10.04 14 4.33 - 4.06 0.440 10.04 16 2.98 - 5.41 0.440 10.04 147 Table F . 4 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 3 0 8.51 0.00 0.450 10.15 2 7.81 - 0.70 0.450 10.15 4 7.21 - 1.31 0.450 10.15 6 7.07 - 1.44 0.450 10.15 8 6.16 - 2.35 0.450 10.15 10 5.61 - 2.90 0.450 10.15 12 5.18 - 3.33 0.450 10.15 14 3.87 - 4.64 0.450 10.15 16 3.29 - 5.22 0.450 10.15 0.65 85 REP 1 0 8.38 0.00 0.651 11.73 1 7.87 - 0.51 0.651 11.73 2 7.28 - 1.10 0.651 11.73 3 5.93 - 2.45 0.651 11.73 4 5.67 - 2.71 0.651 11.73 5 4.78 - 3.60 0.651 11.73 6 4.45 - 3.92 0.651 11.73 7 3.06 - 5.31 0.651 11.73 REP 2 0 8.64 0.00 0.655 12.83 1 8.30 - 0.34 0.655 12.83 2 7.92 - 0.72 0.655 12.83 3 7.31 - 1.33 0.655 12.83 4 6.08 - 2.56 0.655 12.83 5 6.09 - 2.56 0.655 12.83 6 4.72 - 3.93 0.655 12.83 7 4.29 - 4.35 0.655 12.83 8 2.68 - 5.96 0.655 12.83 REP 3 0 8.78 0.00 0.648 12.06 1 8.19 - 0.59 0.648 12.06 2 7.31 - 1.46 0.648 12.06 3 6.83 - 1.95 0.648 12.06 4 6.09 - 2.68 0.648 12.06 5 4.38 - 4.40 0.648 12.06 6 4.04 - 4.73 0.648 12.06 7 4.11 - 4.67 0.648 12.06 8 2.46 - 6.31 0.648 12.06 0.25 90 REP 1 0 7.64 0.00 0.256 7.99 3 7.53 - 0.11 0.256 7.99 6 6.64 - 1.00 0.256 7.99 9 5.59 - 2.05 0.256 7.99 12 4.79 - 2.85 0.256 7.99 15 4.72 - 2.92 0.256 7.99 18 4.04 - 3.60 0.256 7.99 148 Table F . 4 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 90 REP 1 21 4.50 - 3.14 0.256 7.99 REP 2 0 8.85 0.00 0.253 7.99 3 8.14 - 0.71 0.253 7.99 6 7.57 - 1.28 0.253 7.99 9 6.72 - 2.13 0.253 7.99 12 6.04 - 2.81 0.253 7.99 15 5.29 - 3.56 0.253 7.99 18 5.41 - 3.44 0.253 7.99 21 4.18 - 4.67 0.253 7.99 24 4.09 - 4.76 0.253 7.99 REP 3 0 8.70 0.00 0.253 8.12 3 7.24 - 1.47 0.253 8.12 6 6.86 - 1.84 0.253 8.12 9 6.05 - 2.65 0.253 8.12 12 5.03 - 3.67 0.253 8.12 15 4.74 - 3.96 0.253 8.12 18 3.01 - 5.69 0.253 8.12 21 2.98 - 5.73 0.253 8.12 0.45 90 REP 1 0.00 8.61 0.00 0.451 9.80 0.75 8.21 - 0.40 0.451 9.80 1.50 7.79 - 0.83 0.451 9.80 2.25 7.53 - 1.08 0.451 9.80 3.00 6.55 - 2.07 0.451 9.80 3.75 5.94 - 2.67 0.451 9.80 4.50 5.36 - 3.25 0.451 9.80 5.25 4.90 - 3.71 0.451 9.80 6.00 3.59 - 5.02 0.451 9.80 REP 2 0.00 8.33 0.00 0.440 10.04 0.75 7.48 - 0.85 0.440 10.04 1.50 7.31 - 1.01 0.440 10.04 2.25 6.54 - 1.79 0.440 10.04 3.00 6.08 - 2.25 0.440 10.04 3.75 4.89 - 3.43 0.440 10.04 4.50 4.60 - 3.72 0.440 10.04 5.25 3.61 - 4.72 0.440 10.04 6.00 3.53 - 4.80 0.440 10.04 REP 3 0.00 8.10 0.00 0.450 10.15 0.75 7.91 - 0.19 0.450 10.15 1.50 7.13 - 0.97 0.450 10.15 2.25 6.67 - 1.43 0.450 10.15 3.00 6.02 - 2.08 0.450 10.15 3.75 4.99 - 3.10 0.450 10.15 149 Table F . 4 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0. 4 5 90 REP 3 4.50 3.94 - 4.16 0.450 10.15 5.25 3.48 - 4.62 0.450 10.15 0.65 90 REP 1 0.00 8.06 0.00 0.651 11.73 0.33 7.41 - 0.66 0.651 11.73 0.66 6.94 - 1.13 0.651 11.73 0.99 6.48 - 1.58 0.651 11.73 1.32 4.31 - 3.75 0.651 11.73 1.65 3.27 - 4.79 0.651 11.73 1.98 3.04 - 5.02 0.651 11.73 2.31 3.68 - 4.38 0.651 11.73 2.64 2.90 - 5.17 0.651 11.73 REP 2 0.00 7.98 0.00 0.655 12.83 0.33 8.05 0.07 0.655 12.83 0.66 7.45 - 0.53 0.655 12.83 0.99 5.99 - 1.99 0.655 12.83 1.32 5.80 - 2.19 0.655 12.83 1.65 5.51 - 2.47 0.655 12.83 1.98 4.58 - 3.40 0.655 12.83 2.31 3.39 - 4.59 0.655 12.83 2.64 2.97 - 5.02 0.655 12.83 REP 3 0.00 8.25 0.00 0.648 12.06 0.33 7.79 - 0.46 0.648 12.06 0.66 7.23 - 1.02 0.648 12.06 0.99 5.65 - 2.60 0.648 12.06 1.32 5.35 - 2.89 0.648 12.06 1.65 4.98 - 3.27 0.648 12.06 1.98 4.41 - 3.84 0.648 12.06 2.31 3.67 - 4.58 0.648 12.06 2.64 3.47 - 4.77 0.648 12.06 150 Table F . 5 Salmonella inactivation data for wheat meal. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 8.58 0.00 0.254 8.00 13 7.71 - 0.87 0.254 8.00 26 7.06 - 1.52 0.254 8.00 39 6.83 - 1.75 0.254 8.00 52 6.47 - 2.12 0.254 8.00 65 6.04 - 2.55 0.254 8.00 78 5.67 - 2.91 0.254 8.00 91 5.23 - 3.35 0.254 8.00 104 5.14 - 3.44 0.254 8.00 REP 2 0 8.21 0.00 0.253 8.45 13 7.61 - 0.60 0.253 8.45 26 7.12 - 1.09 0.253 8.45 39 6.97 - 1.24 0.253 8.45 52 6.32 - 1.89 0.253 8.45 65 6.33 - 1.88 0.253 8.45 78 5.70 - 2.51 0.253 8.45 91 5.32 - 2.89 0.253 8.45 104 5.34 - 2.87 0.253 8.45 REP 3 0 8.44 0.00 0.253 8.38 13 7.74 - 0.71 0.253 8.38 26 7.22 - 1.22 0.253 8.38 39 7.13 - 1.31 0.253 8.38 52 6.51 - 1.93 0.253 8.38 65 6.43 - 2.01 0.253 8.38 78 5.85 - 2.59 0.253 8.38 91 5.62 - 2.83 0.253 8.38 104 5.07 - 3.37 0.253 8.38 0.45 80 REP 1 0 8.35 0.00 0.457 10.00 8 7.47 - 0.88 0.457 10.00 16 6.84 - 1.51 0.457 10.00 24 6.65 - 1.70 0.457 10.00 32 6.12 - 2.22 0.457 10.00 40 5.87 - 2.48 0.457 10.00 48 4.77 - 3.58 0.457 10.00 56 4.49 - 3.85 0.457 10.00 64 3.98 - 4.37 0.457 10.00 REP 2 0 8.14 0.00 0.460 9.81 8 6.62 - 1.52 0.460 9.81 16 6.12 - 2.02 0.460 9.81 24 6.22 - 1.92 0.460 9.81 32 4.29 - 3.85 0.460 9.81 40 4.07 - 4.07 0.460 9.81 151 Table F . 5 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 80 REP 2 48 3.94 - 4.21 0.460 9.81 56 3.26 - 4.89 0.460 9.81 REP 3 0 8.83 0.00 0.450 10.26 8 6.03 - 2.80 0.450 10.26 16 6.15 - 2.68 0.450 10.26 24 5.16 - 3.68 0.450 10.26 32 4.85 - 3.98 0.450 10.26 40 4.17 - 4.66 0.450 10.26 48 3.10 - 5.73 0.450 10.26 56 3.46 - 5.37 0.450 10.26 0.65 80 REP 1 0 8.06 0.00 0.650 11.94 2 7.12 - 0.95 0.650 11.94 4 6.64 - 1.42 0.650 11.94 6 5.53 - 2.53 0.650 11.94 8 5.29 - 2.78 0.650 11.94 10 5.00 - 3.07 0.650 11.94 12 4.60 - 3.47 0.650 11.94 14 3.71 - 4.36 0.650 11.94 16 2.38 - 5.68 0.650 11.94 REP 2 0 8.12 0.00 0.650 12.17 2 7.35 - 0.77 0.650 12.17 4 7.06 - 1.05 0.650 12.17 6 6.41 - 1.70 0.650 12.17 8 6.17 - 1.95 0.650 12.17 10 6.16 - 1.96 0.650 12.17 12 5.51 - 2.61 0.650 12.17 14 4.61 - 3.51 0.650 12.17 16 4.37 - 3.74 0.650 12.17 REP 3 0 8.17 0.00 0.653 12.56 2 7.53 - 0.64 0.653 12.56 4 7.15 - 1.02 0.653 12.56 6 6.68 - 1.49 0.653 12.56 8 6.12 - 2.05 0.653 12.56 10 5.44 - 2.73 0.653 12.56 12 5.28 - 2.89 0.653 12.56 14 4.53 - 3.64 0.653 12.56 16 4.30 - 3.87 0.653 12.56 0.25 85 REP 1 0 8.24 0.00 0.254 8.00 8 7.20 - 1.04 0.254 8.00 16 6.78 - 1.47 0.254 8.00 24 6.06 - 2.19 0.254 8.00 32 6.12 - 2.12 0.254 8.00 152 Table F . 5 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 85 REP 1 40 5.01 - 3.23 0.254 8.00 48 5.03 - 3.21 0.254 8.00 56 5.39 - 2.86 0.254 8.00 64 3.71 - 4.54 0.254 8.00 REP 2 0 8.17 0.00 0.253 8.45 8 7.39 - 0.78 0.253 8.45 16 7.07 - 1.10 0.253 8.45 24 6.71 - 1.46 0.253 8.45 32 6.19 - 1.98 0.253 8.45 40 6.02 - 2.15 0.253 8.45 48 5.73 - 2.44 0.253 8.45 56 5.37 - 2.80 0.253 8.45 64 5.44 - 2.73 0.253 8.45 REP 3 0 8.14 0.00 0.253 8.38 8 7.67 - 0.47 0.253 8.38 16 6.85 - 1.29 0.253 8.38 24 5.68 - 2.46 0.253 8.38 32 5.85 - 2.29 0.253 8.38 40 5.42 - 2.72 0.253 8.56 48 4.84 - 3.30 0.253 8.58 56 4.25 - 3.88 0.253 8.61 64 3.98 - 4.15 0.253 8.63 0.45 85 REP 1 0 8.29 0.00 0.457 10.00 2.5 7.22 - 1.07 0.457 10.00 5 6.51 - 1.78 0.457 10.00 7.5 6.52 - 1.77 0.457 10.00 10 6.41 - 1.88 0.457 10.00 12.5 5.78 - 2.51 0.457 10.00 15 4.86 - 3.43 0.457 10.00 17.5 4.53 - 3.76 0.457 10.00 20 4.27 - 4.02 0.457 10.00 REP 2 0 8.19 0.00 0.450 9.83 2.5 6.99 - 1.20 0.450 9.83 5 6.01 - 2.18 0.450 9.83 7.5 5.67 - 2.52 0.450 9.83 10 4.89 - 3.30 0.450 9.83 12.5 4.31 - 3.88 0.450 9.83 15 3.04 - 5.15 0.450 9.83 17.5 3.32 - 4.87 0.450 9.83 0.45 85 REP 3 0 7.68 0.00 0.458 9.87 2.5 6.65 - 1.04 0.458 9.87 5 6.07 - 1.62 0.458 9.87 153 Table F . 5 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 3 7.5 4.16 - 3.53 0.458 9.87 10 4.60 - 3.09 0.458 9.87 12.5 4.08 - 3.61 0.458 9.87 15 3.09 - 4.60 0.458 9.87 17.5 2.74 - 4.95 0.458 9.87 20 3.04 - 4.65 0.458 9.87 0.65 85 REP 1 0.00 7.54 0.00 0.634 10.97 0.75 6.94 - 0.60 0.634 10.97 1.50 6.52 - 1.02 0.634 10.97 2.25 6.11 - 1.43 0.634 10.97 3.00 5.38 - 2.16 0.634 10.97 3.75 4.49 - 3.05 0.634 10.97 4.50 3.59 - 3.95 0.634 10.97 5.25 2.96 - 4.58 0.634 10.97 6.00 2.47 - 5.07 0.634 10.97 REP 2 0.00 7.78 0.00 0.650 12.17 0.75 7.26 - 0.52 0.650 12.17 1.50 6.80 - 0.98 0.650 12.17 2.25 5.88 - 1.91 0.650 12.17 3.00 5.08 - 2.70 0.650 12.17 3.75 4.38 - 3.40 0.650 12.17 4.50 4.47 - 3.32 0.650 12.17 5.25 4.01 - 3.78 0.650 12.17 6.00 3.41 - 4.38 0.650 12.17 REP 3 0.00 7.64 0.00 0.653 12.56 0.75 7.03 - 0.61 0.653 12.56 1.50 6.40 - 1.24 0.653 12.56 2.25 6.01 - 1.63 0.653 12.56 3.00 5.11 - 2.53 0.653 12.56 3.75 5.48 - 2.16 0.653 12.56 4.50 4.80 - 2.84 0.653 12.56 5.25 3.75 - 3.89 0.653 12.56 6.00 3.24 - 4.40 0.653 12.56 0.25 90 REP 1 0 8.07 0.00 0.254 8.00 3 6.85 - 1.21 0.254 8.00 6 6.24 - 1.83 0.254 8.00 9 5.99 - 2.08 0.254 8.00 12 5.12 - 2.95 0.254 8.00 15 4.48 - 3.58 0.254 8.00 18 3.76 - 4.31 0.254 8.00 21 3.78 - 4.29 0.254 8.00 24 3.33 - 4.73 0.254 8.00 154 Table F . 5 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 90 REP 2 0 8.09 0.00 0.253 8.45 3 7.18 - 0.91 0.253 8.45 6 6.69 - 1.40 0.253 8.45 9 5.44 - 2.65 0.253 8.45 12 5.89 - 2.19 0.253 8.45 15 5.31 - 2.77 0.253 8.45 18 4.81 - 3.28 0.253 8.45 21 4.60 - 3.48 0.253 8.45 24 3.40 - 4.68 0.253 8.45 REP 3 0 8.06 0.00 0.253 8.38 3 7.56 - 0.50 0.253 8.38 6 6.57 - 1.49 0.253 8.38 9 6.07 - 1.99 0.253 8.38 12 5.36 - 2.70 0.253 8.38 15 4.83 - 3.23 0.253 8.38 18 4.48 - 3.58 0.253 8.38 21 3.72 - 4.34 0.253 8.38 24 3.75 - 4.31 0.253 8.38 0.45 90 REP 1 0 7.48 0.00 0.457 10.00 0.5 7.04 - 0.44 0.457 10.00 1 6.48 - 1.00 0.457 10.00 1.5 6.47 - 1.01 0.457 10.00 2 5.80 - 1.68 0.457 10.00 2.5 5.51 - 1.97 0.457 10.00 3 5.12 - 2.36 0.457 10.00 3.5 4.37 - 3.11 0.457 10.00 4 4.06 - 3.42 0.457 10.00 REP 2 0 7.45 0.00 0.460 9.81 0.5 7.96 0.51 0.460 9.81 1 6.05 - 1.40 0.460 9.81 1.5 5.57 - 1.88 0.460 9.81 2 4.88 - 2.57 0.460 9.81 2.5 4.05 - 3.40 0.460 9.81 3 3.18 - 4.27 0.460 9.81 3.5 3.21 - 4.24 0.460 9.81 4 3.23 - 4.22 0.460 9.81 REP 3 0 6.71 0.00 0.458 9.87 0.5 5.93 - 0.78 0.458 9.87 1 5.76 - 0.94 0.458 9.87 1.5 5.27 - 1.43 0.458 9.87 2 4.73 - 1.97 0.458 9.87 2.5 4.17 - 2.54 0.458 9.87 155 Table F . 5 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 90 REP 3 3 2.96 - 3.75 0.458 9.87 0.65 90 REP 1 0.00 6.47 0.00 0.643 11.94 0.17 5.82 - 0.64 0.643 11.94 0.33 5.10 - 1.36 0.643 11.94 0.50 3.96 - 2.51 0.643 11.94 0.67 5.09 - 1.38 0.643 11.94 0.84 3.39 - 3.07 0.643 11.94 1.00 4.05 - 2.42 0.643 11.94 1.17 2.97 - 3.49 0.643 11.94 1.34 2.18 - 4.29 0.643 11.94 REP 2 0.00 6.88 0.00 0.650 12.17 0.17 6.36 - 0.52 0.650 12.17 0.33 5.72 - 1.16 0.650 12.17 0.50 5.78 - 1.10 0.650 12.17 0.67 4.87 - 2.01 0.650 12.17 0.84 5.27 - 1.62 0.650 12.17 1.00 5.12 - 1.76 0.650 12.17 1.17 5.14 - 1.75 0.650 12.17 1.34 3.54 - 3.34 0.650 12.17 REP 3 0.00 4.72 0.00 0.653 12.56 0.17 6.34 1.61 0.653 12.56 0.33 5.93 1.21 0.653 12.56 0.50 6.26 1.54 0.653 12.56 0.67 5.70 0.98 0.653 12.56 0.84 5.04 0.32 0.653 12.56 1.17 3.92 - 0.80 0.653 12.56 1.34 3.17 - 1.55 0.653 12.56 156 Table F . 6 Salmonella inactivation data for wheat flour . Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 80 REP 1 0 7.94 0.00 0.252 8.09 16 7.51 - 0.43 0.252 8.09 32 7.07 - 0.87 0.252 8.09 48 6.62 - 1.33 0.252 8.09 64 6.64 - 1.30 0.252 8.09 80 6.05 - 1.90 0.252 8.09 96 5.33 - 2.61 0.252 8.09 112 5.48 - 2.46 0.252 8.09 128 5.06 - 2.89 0.252 8.09 REP 2 0 8.34 0.00 0.244 7.27 16 7.49 - 0.85 0.244 7.27 32 7.04 - 1.30 0.244 7.27 48 6.58 - 1.76 0.244 7.27 64 6.47 - 1.88 0.244 7.27 80 6.03 - 2.32 0.244 7.27 96 5.98 - 2.36 0.244 7.27 112 5.51 - 2.83 0.244 7.27 128 5.07 - 3.28 0.244 7.27 REP 3 0 8.13 0.00 0.252 8.30 16 7.29 - 0.84 0.252 8.30 32 7.05 - 1.08 0.252 8.30 48 5.68 - 2.45 0.252 8.30 64 5.64 - 2.49 0.252 8.30 80 5.43 - 2.70 0.252 8.30 96 5.31 - 2.82 0.252 8.30 112 2.60 - 5.53 0.252 8.30 128 3.96 - 4.17 0.252 8.30 0.45 80 REP 1 0 8.44 0.00 0.450 10.00 6 7.75 - 0.69 0.450 10.00 12 7.30 - 1.14 0.450 10.00 18 6.94 - 1.50 0.450 10.00 24 6.41 - 2.03 0.450 10.00 30 6.35 - 2.08 0.450 10.00 36 5.49 - 2.95 0.450 10.00 42 5.97 - 2.46 0.450 10.00 48 5.27 - 3.17 0.450 10.00 REP 2 0 8.27 0.00 0.459 10.31 6 6.84 - 1.43 0.459 10.31 12 6.39 - 1.88 0.459 10.31 18 5.59 - 2.68 0.459 10.31 30 4.47 - 3.81 0.459 10.31 36 3.77 - 4.50 0.459 10.31 157 Table F . 6 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 80 REP 2 42 3.52 - 4.75 0.459 10.31 REP 3 0 8.51 0.00 0.465 10.22 6 7.04 4.99 0.465 10.22 12 6.86 3.52 0.465 10.22 18 6.03 3.34 0.465 10.22 24 5.35 2.52 0.465 10.22 30 4.36 1.83 0.465 10.22 36 3.78 0.84 0.465 10.22 42 3.73 0.26 0.465 10.22 48 3.89 0.21 0.465 10.22 0.65 80 REP 1 0 7.97 0.00 0.646 12.01 2 7.02 - 0.95 0.646 12.01 4 6.40 - 1.57 0.646 12.01 6 6.39 - 1.58 0.646 12.01 8 5.04 - 2.93 0.646 12.01 10 4.30 - 3.67 0.646 12.01 12 3.46 - 4.50 0.646 12.01 14 3.21 - 4.75 0.646 12.01 16 2.62 - 5.35 0.646 12.01 REP 2 0 8.22 0.00 0.637 12.63 2 7.45 - 0.77 0.637 12.63 4 6.54 - 1.68 0.637 12.63 6 6.13 - 2.09 0.637 12.63 8 5.73 - 2.49 0.637 12.63 10 4.88 - 3.34 0.637 12.63 12 4.59 - 3.64 0.637 12.63 14 4.16 - 4.06 0.637 12.63 16 3.24 - 4.98 0.637 12.63 REP 3 0 8.22 0.00 0.652 12.70 2 7.37 - 0.85 0.652 12.70 4 6.70 - 1.52 0.652 12.70 6 6.24 - 1.98 0.652 12.70 8 5.44 - 2.78 0.652 12.70 10 5.41 - 2.81 0.652 12.70 12 4.35 - 3.87 0.652 12.70 14 4.20 - 4.03 0.652 12.70 16 3.39 - 4.83 0.652 12.70 0.25 85 REP 1 0 8.17 0.00 0.245 7.73 11 6.96 - 1.22 0.245 7.73 22 6.08 - 2.09 0.245 7.73 33 5.90 - 2.28 0.245 7.73 44 5.36 - 2.81 0.245 7.73 158 Table F . 6 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 85 REP 1 55 4.41 - 3.76 0.245 7.73 66 4.20 - 3.97 0.245 7.73 77 4.23 - 3.94 0.245 7.73 REP 2 0 8.25 0.00 0.244 7.27 11 7.24 - 1.01 0.244 7.27 22 6.38 - 1.87 0.244 7.27 33 6.02 - 2.23 0.244 7.27 44 5.18 - 3.07 0.244 7.27 55 5.08 - 3.17 0.244 7.27 66 5.28 - 2.97 0.244 7.27 77 5.01 - 3.24 0.244 7.27 88 3.71 - 4.54 0.244 7.27 REP 3 0 7.95 0.00 0.252 8.30 11 7.22 - 0.73 0.252 8.30 22 6.40 - 1.55 0.252 8.30 33 5.69 - 2.26 0.252 8.30 44 5.06 - 2.89 0.252 8.30 55 4.26 - 3.69 0.252 8.30 77 3.37 - 4.58 0.252 8.30 0.45 85 REP 1 0 8.43 0.00 0.450 10.00 1.5 7.97 - 0.46 0.450 10.00 3 7.36 - 1.06 0.450 10.00 4.5 7.03 - 1.40 0.450 10.00 6 7.24 - 1.19 0.450 10.00 7.5 6.24 - 2.19 0.450 10.00 9 6.31 - 2.12 0.450 10.00 10.5 5.93 - 2.50 0.450 10.00 12 6.14 - 2.29 0.450 10.00 REP 2 0 8.09 0.00 0.459 10.31 1.5 8.09 0.00 0.459 10.31 3 6.66 - 1.43 0.459 10.31 4.5 5.76 - 2.33 0.459 10.31 6 5.66 - 2.43 0.459 10.31 7.5 5.30 - 2.79 0.459 10.31 9 4.96 - 3.13 0.459 10.31 10.5 4.26 - 3.83 0.459 10.22 12 4.30 - 3.79 0.459 10.22 REP 3 0 7.56 0.00 0.465 10.22 1.5 6.85 - 0.72 0.465 10.22 3 6.16 - 1.40 0.465 10.22 4.5 5.92 - 1.64 0.465 10.22 6 5.28 - 2.28 0.465 10.22 159 Table F . 6 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 85 REP 3 7.5 5.41 - 2.16 0.465 10.22 9 4.31 - 3.25 0.465 10.22 10.5 4.22 - 3.34 0.465 10.22 12 3.69 - 3.87 0.465 10.22 0.65 85 REP 1 0 7.23 0.00 0.646 12.01 0.5 6.92 - 0.31 0.646 12.01 1 6.72 - 0.52 0.646 12.01 1.5 5.97 - 1.26 0.646 12.01 2 5.38 - 1.85 0.646 12.01 2.5 4.84 - 2.40 0.646 12.01 3 4.75 - 2.48 0.646 12.01 3.5 4.07 - 3.16 0.646 12.01 4 3.37 - 3.87 0.646 12.01 REP 2 0 7.37 0.00 0.637 12.63 0.5 7.07 - 0.30 0.637 12.63 1 6.95 - 0.42 0.637 12.63 1.5 6.33 - 1.04 0.637 12.63 2 5.82 - 1.55 0.637 12.63 2.5 5.05 - 2.32 0.637 12.63 3 5.48 - 1.89 0.637 12.63 3.5 5.02 - 2.35 0.637 12.63 4 3.37 - 4.00 0.637 12.63 REP 3 0 7.36 0.00 0.652 12.70 0.5 7.07 - 0.29 0.652 12.70 1 6.58 - 0.78 0.652 12.70 1.5 6.30 - 1.06 0.652 12.70 2 6.29 - 1.07 0.652 12.70 2.5 5.74 - 1.62 0.652 12.70 3 5.48 - 1.88 0.652 12.70 3.5 4.29 - 3.07 0.652 12.70 4 4.29 - 3.07 0.652 12.70 0.25 90 REP 1 0 7.48 0.00 0.252 8.09 4 6.75 - 0.73 0.252 8.09 8 5.83 - 1.65 0.252 8.09 12 5.01 - 2.47 0.252 8.09 16 4.67 - 2.81 0.252 8.09 20 4.54 - 2.94 0.252 8.09 24 3.31 - 4.16 0.252 8.09 28 4.14 - 3.34 0.252 8.09 REP 2 0 7.96 0.00 0.244 7.27 4 7.42 - 0.55 0.244 7.27 8 6.50 - 1.46 0.244 7.27 160 Table F . 6 Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.25 90 REP 2 12 6.05 - 1.92 0.244 7.27 16 5.52 - 2.44 0.244 7.27 20 4.35 - 3.61 0.244 7.27 24 3.98 - 3.98 0.244 7.27 28 4.47 - 3.49 0.244 7.27 32 2.85 - 5.11 0.244 7.27 REP 3 0 8.08 0.00 0.250 7.33 4 6.38 - 1.69 0.250 7.33 8 5.50 - 2.57 0.250 7.33 12 4.61 - 3.46 0.250 7.33 16 4.09 - 3.99 0.250 7.33 20 3.60 - 4.48 0.250 7.33 24 3.03 - 5.05 0.250 7.33 28 2.55 - 5.53 0.250 7.33 32 1.81 - 6.27 0.250 7.33 0.45 90 REP 1 0.00 6.90 0.00 0.451 10.00 0.37 6.41 - 0.49 0.451 10.00 0.73 6.08 - 0.82 0.451 10.00 1.10 5.99 - 0.91 0.451 10.00 1.47 4.74 - 2.17 0.451 10.00 1.84 5.18 - 1.72 0.451 10.00 2.20 4.71 - 2.19 0.451 10.00 2.57 5.01 - 1.89 0.451 10.00 2.94 4.17 - 2.73 0.451 10.00 REP 2 0.00 6.55 0.00 0.459 10.31 0.37 7.26 0.71 0.459 10.31 0.73 5.59 - 0.96 0.459 10.31 1.10 6.07 - 0.48 0.459 10.31 1.47 5.96 - 0.59 0.459 10.31 1.84 4.43 - 2.12 0.459 10.31 2.20 4.20 - 2.35 0.459 10.31 2.57 3.58 - 2.97 0.459 10.31 REP 3 0.00 6.23 0.00 0.454 10.46 0.37 5.67 - 0.56 0.454 10.46 0.73 5.07 - 1.16 0.454 10.46 1.10 5.18 - 1.05 0.454 10.46 1.47 3.94 - 2.29 0.454 10.46 1.84 4.23 - 2.00 0.454 10.46 2.20 3.29 - 2.94 0.454 10.46 2.57 2.66 - 3.57 0.454 10.46 161 Table F . 6 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w % MC, db 0.45 90 REP 3 2.94 2.75 - 3.49 0.454 10.46 0.65 90 REP 1 0.00 4.07 0.00 0.649 12.82 0.12 3.39 - 0.69 0.649 12.82 0.23 2.73 - 1.35 0.649 12.82 0.35 3.45 - 0.63 0.649 12.82 0.47 2.02 - 2.05 0.649 12.82 0.58 3.40 - 0.67 0.649 12.82 0.70 1.84 - 2.23 0.649 12.82 0.82 1.40 - 2.68 0.649 12.82 REP 2 0.00 5.12 0.00 0.637 12.63 0.12 4.92 - 0.19 0.637 12.63 0.23 4.23 - 0.88 0.637 12.63 0.35 4.25 - 0.87 0.637 12.63 0.47 3.53 - 1.59 0.637 12.63 0.58 3.35 - 1.77 0.637 12.63 0.70 3.38 - 1.73 0.637 12.63 0.93 2.36 - 2.76 0.637 12.63 REP 3 0.00 5.30 0.00 0.652 12.70 0.12 5.13 - 0.17 0.652 12.70 0.23 4.80 - 0.50 0.652 12.70 0.35 4.36 - 0.95 0.652 12.70 0.47 3.49 - 1.81 0.652 12.70 0.58 3.73 - 1.58 0.652 12.70 0.70 2.42 - 2.88 0.652 12.70 0.82 2.78 - 2.52 0.652 12.70 0.93 2.70 - 2.60 0.652 12.70 162 Table F . 7 Salmonella inactivation data for date pieces. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.25 70 REP 1 0 9.19 0.00 0.280 3 9.41 0.22 0.280 6 7.94 - 1.25 0.280 9 7.74 - 1.45 0.280 12 7.57 - 1.62 0.280 15 6.28 - 2.90 0.280 18 7.02 - 2.17 0.280 21 5.66 - 3.53 0.280 24 5.57 - 3.62 0.280 REP 2 0 8.53 0.00 0.265 3 8.98 0.45 0.265 6 7.95 - 0.59 0.265 9 7.94 - 0.59 0.265 12 7.89 - 0.64 0.265 15 6.83 - 1.71 0.265 18 7.46 - 1.08 0.265 21 6.99 - 1.54 0.265 24 4.89 - 3.65 0.265 REP 3 0 8.21 0.00 0.234 3 8.74 0.54 0.234 6 8.37 0.16 0.234 9 7.70 - 0.51 0.234 12 8.08 - 0.12 0.234 15 7.81 - 0.39 0.234 18 7.26 - 0.94 0.234 21 7.14 - 1.07 0.234 24 6.41 - 1.79 0.234 0.45 70 REP 1 0 9.09 0.00 0.442 3 8.26 - 0.84 0.442 6 6.01 - 3.08 0.442 9 5.22 - 3.87 0.442 12 5.19 - 3.90 0.442 18 5.48 - 3.61 0.442 21 3.00 - 6.09 0.442 REP 2 0 8.05 0.00 0.453 3 7.78 - 0.27 0.453 6 7.29 - 0.77 0.453 9 7.05 - 1.00 0.453 12 5.59 - 2.46 0.453 15 7.24 - 0.81 0.453 21 4.99 - 3.06 0.453 163 Table F . 7 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.45 70 REP 3 0 8.02 0.00 0.442 3 6.95 - 1.06 0.442 6 7.14 - 0.88 0.442 9 7.42 - 0.60 0.442 12 7.44 - 0.58 0.442 18 3.97 - 4.04 0.442 0.65 70 REP 1 0 8.31 0.00 0.623 3 7.13 - 1.17 0.623 6 7.60 - 0.71 0.623 9 7.40 - 0.90 0.623 12 6.83 - 1.47 0.623 15 6.98 - 1.33 0.623 21 3.13 - 5.18 0.623 REP 2 0 6.76 0.00 0.633 3 8.64 1.88 0.633 6 8.52 1.76 0.633 12 7.40 0.64 0.633 18 7.34 0.58 0.633 21 3.82 - 2.94 0.633 24 4.91 - 1.85 0.633 REP 3 0 8.41 0.00 0.647 3 7.30 - 1.11 0.647 6 6.73 - 1.68 0.647 9 6.26 - 2.15 0.647 12 3.91 - 4.49 0.647 18 3.22 - 5.18 0.647 0.25 75 REP 1 0 7.64 0.00 0.280 1.5 8.65 1.00 0.280 4.5 6.79 - 0.85 0.280 6 6.89 - 0.75 0.280 7.5 6.43 - 1.21 0.280 9 6.21 - 1.43 0.280 12 5.34 - 2.30 0.280 REP 2 0 9.03 0.00 0.265 1.5 8.49 - 0.54 0.265 3 8.17 - 0.86 0.265 4.5 7.66 - 1.37 0.265 6 7.39 - 1.63 0.265 7.5 6.36 - 2.66 0.265 9 5.75 - 3.28 0.265 10.5 4.38 - 4.65 0.265 12 3.80 - 5.22 0.265 164 Table F . 7 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.25 75 REP 3 0 8.76 0.00 0.234 1.5 8.39 - 0.37 0.234 3 8.16 - 0.60 0.234 4.5 7.80 - 0.96 0.234 6 6.73 - 2.03 0.234 7.5 6.33 - 2.43 0.234 9 7.29 - 1.47 0.234 10.5 4.85 - 3.91 0.234 12 4.76 - 4.01 0.234 0.45 75 REP 1 0 6.28 0.00 0.442 1.5 6.57 0.29 0.442 3 7.34 1.06 0.442 4.5 6.00 - 0.28 0.442 6 6.00 - 0.28 0.442 9 2.58 - 3.70 0.442 REP 2 0 7.23 0.00 0.453 1.5 6.66 - 0.57 0.453 3 5.99 - 1.25 0.453 4.5 6.90 - 0.33 0.453 6 4.57 - 2.66 0.453 7.5 6.13 - 1.10 0.453 9 4.68 - 2.55 0.453 10.5 6.37 - 0.86 0.453 REP 3 0 8.87 0.00 0.442 1.5 7.69 - 1.18 0.442 3 8.08 - 0.79 0.442 4.5 4.60 - 4.27 0.442 6 5.51 - 3.36 0.442 9 4.88 - 3.99 0.442 12 2.42 - 6.45 0.442 0.65 75 REP 1 0 8.29 0.00 0.635 1.5 8.21 - 0.08 0.635 3 7.64 - 0.65 0.635 4.5 6.72 - 1.57 0.635 6 5.81 - 2.48 0.635 7.5 5.83 - 2.46 0.635 10.5 4.50 - 3.78 0.635 12 5.42 - 2.87 0.635 0.65 75 REP 2 0 7.42 0.00 0.644 1.5 8.10 0.68 0.644 3 6.27 - 1.15 0.644 4.5 4.55 - 2.87 0.644 165 Table F . 7 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.65 75 REP 2 6 4.80 - 2.62 0.644 7.5 4.33 - 3.09 0.644 9 4.21 - 3.21 0.644 10.5 3.69 - 3.73 0.644 12 3.66 - 3.76 0.644 REP 3 0 7.32 0.00 0.656 1.5 7.63 0.30 0.656 3 6.37 - 0.96 0.656 4.5 6.85 - 0.48 0.656 6 4.58 - 2.74 0.656 7.5 2.99 - 4.34 0.656 9 3.86 - 3.46 0.656 0 9.05 0.00 0.280 0.5 8.90 - 0.14 0.280 0.25 80 REP 1 1 8.75 - 0.30 0.280 1.5 8.03 - 1.02 0.280 2 7.75 - 1.30 0.280 3 5.88 - 3.17 0.280 3.5 4.51 - 4.53 0.280 4 3.98 - 5.06 0.280 0 7.44 0.00 0.265 0.5 6.51 - 0.94 0.265 REP 2 1 6.99 - 0.45 0.265 1.5 7.03 - 0.41 0.265 2 5.29 - 2.16 0.265 2.5 6.65 - 0.79 0.265 3 7.08 - 0.36 0.265 3.5 6.01 - 1.44 0.265 4 5.50 - 1.95 0.265 0 9.26 0.00 0.234 0.5 9.03 - 0.24 0.234 REP 3 1 8.09 - 1.17 0.234 1.5 8.23 - 1.03 0.234 2 7.79 - 1.47 0.234 2.5 7.82 - 1.44 0.234 3 6.75 - 2.51 0.234 3.5 4.93 - 4.33 0.234 4 5.13 - 4.14 0.234 0.45 80 REP 1 0 6.72 0.00 0.461 0.5 5.74 - 0.98 0.461 1 5.06 - 1.66 0.461 1.5 4.74 - 1.98 0.461 166 Table F . 7 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.45 80 REP 1 2 4.48 - 2.24 0.461 2.5 2.03 - 4.69 0.461 3 1.88 - 4.85 0.461 REP 2 0 8.61 0.00 0.453 0.5 8.34 - 0.27 0.453 1 8.08 - 0.52 0.453 1.5 7.52 - 1.09 0.453 2 5.86 - 2.74 0.453 2.5 5.92 - 2.68 0.453 3 5.69 - 2.91 0.453 3.5 6.10 - 2.51 0.453 4 3.59 - 5.02 0.453 REP 3 0 8.58 0.00 0.455 0.5 8.36 - 0.22 0.455 1 8.53 - 0.04 0.455 1.5 7.60 - 0.98 0.455 2 7.39 - 1.19 0.455 2.5 7.52 - 1.06 0.455 3 6.58 - 1.99 0.455 3.5 6.28 - 2.29 0.455 4 6.23 - 2.35 0.455 0.65 80 REP 1 0 8.53 0.00 0.634 0.5 6.96 - 1.56 0.634 1 7.09 - 1.44 0.634 1.5 7.08 - 1.44 0.634 2 5.22 - 3.31 0.634 2.5 4.60 - 3.93 0.634 3 4.12 - 4.41 0.634 3.5 3.43 - 5.10 0.634 4 3.52 - 5.00 0.634 REP 2 0 7.14 0.00 0.644 0.5 7.09 - 0.05 0.644 1 6.91 - 0.23 0.644 1.5 5.30 - 1.84 0.644 2 3.16 - 3.98 0.644 2.5 3.57 - 3.57 0.644 4 3.95 - 3.19 0.644 167 Table F . 7 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.65 80 REP 3 0 7.43 0.00 0.656 0.5 7.15 - 0.28 0.656 1 6.71 - 0.72 0.656 1.5 5.32 - 2.11 0.656 2 5.48 - 1.95 0.656 2.5 6.47 - 0.96 0.656 3 6.60 - 0.83 0.656 3.5 5.14 - 2.29 0.656 4 2.11 - 5.32 0.656 * Moisture content of date pieces was measured in the different batches. ** Moisture content of 0.25, 0.45, and 0.65 a w date pieces was 10.17, 13.31, and 19.21 %MC, respectively. 168 Table F . 8 Salmonella inactivation data for date paste. Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.25 70 REP 1 0 8.27 0.00 0.244 4 8.09 - 0.18 0.244 8 8.00 - 0.27 0.244 12 7.80 - 0.47 0.244 16 7.72 - 0.55 0.244 20 7.72 - 0.55 0.244 28 7.39 - 0.87 0.244 32 7.38 - 0.88 0.244 REP 2 0 8.05 0.00 0.263 8 7.67 - 0.38 0.263 16 7.51 - 0.54 0.263 24 7.15 - 0.90 0.263 32 7.14 - 0.90 0.263 40 6.80 - 1.25 0.263 48 6.73 - 1.31 0.263 56 6.37 - 1.68 0.263 64 5.88 - 2.17 0.263 REP 3 0 7.69 0.00 0.255 8 6.93 - 0.76 0.255 16 6.69 - 1.00 0.255 24 5.93 - 1.77 0.255 32 6.79 - 0.91 0.255 40 6.19 - 1.50 0.255 48 6.28 - 1.41 0.255 56 5.95 - 1.74 0.255 64 6.21 - 1.49 0.255 0.45 70 REP 1 0 7.81 0.00 0.434 3 7.95 0.14 0.434 6 7.76 - 0.05 0.434 9 7.27 - 0.55 0.434 12 6.96 - 0.85 0.434 15 7.33 - 0.48 0.434 18 6.91 - 0.90 0.434 21 6.43 - 1.39 0.434 REP 2 0 7.95 0.00 0.449 3 7.88 - 0.07 0.449 6 7.26 - 0.69 0.449 9 7.02 - 0.93 0.449 12 6.92 - 1.03 0.449 15 7.13 - 0.82 0.449 18 6.37 - 1.58 0.449 21 6.28 - 1.67 0.449 169 Table F . 8 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.45 70 REP 3 0 8.15 0.00 0.445 4 7.66 - 0.49 0.445 6 7.20 - 0.95 0.445 9 7.31 - 0.84 0.445 12 6.94 - 1.21 0.445 15 6.60 - 1.55 0.445 18 6.84 - 1.31 0.445 21 6.85 - 1.31 0.445 24 6.27 - 1.89 0.445 0.65 70 REP 1 0 7.72 0.00 0.634 2 7.14 - 0.57 0.634 4 6.97 - 0.75 0.634 6 6.53 - 1.18 0.634 8 6.48 - 1.23 0.634 10 6.02 - 1.70 0.634 12 4.59 - 3.12 0.634 14 5.95 - 1.76 0.634 16 4.73 - 2.99 0.634 REP 2 0 8.24 0.00 0.649 2 7.78 - 0.45 0.649 4 7.57 - 0.66 0.649 6 6.34 - 1.90 0.649 8 6.07 - 2.16 0.649 10 6.80 - 1.44 0.649 12 5.66 - 2.57 0.649 14 5.75 - 2.48 0.649 16 5.27 - 2.96 0.649 REP 3 0 7.91 0.00 0.649 2 7.22 - 0.69 0.649 4 7.37 - 0.54 0.649 6 6.16 - 1.75 0.649 8 6.30 - 1.62 0.649 10 4.48 - 3.43 0.649 12 5.07 - 2.84 0.649 14 4.48 - 3.43 0.649 0.25 75 REP 1 0 8.31 0.00 0.254 2 8.07 - 0.24 0.254 4 7.40 - 0.91 0.254 6 7.69 - 0.63 0.254 8 7.01 - 1.30 0.254 10 8.15 - 0.16 0.254 12 7.18 - 1.13 0.254 170 Table F . 8 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.25 75 REP 1 14 6.65 - 1.66 0.254 16 6.39 - 1.92 0.254 REP 2 0 8.09 0.00 0.263 4 7.52 - 0.57 0.263 8 7.22 - 0.88 0.263 12 7.00 - 1.09 0.263 16 6.76 - 1.33 0.263 20 6.51 - 1.58 0.263 24 6.85 - 1.24 0.263 28 6.88 - 1.21 0.263 32 5.64 - 2.45 0.263 REP 3 0 7.75 0.00 0.255 4 6.89 - 0.86 0.255 8 6.68 - 1.07 0.255 12 6.48 - 1.27 0.255 16 6.02 - 1.72 0.255 20 6.63 - 1.11 0.255 24 6.09 - 1.65 0.255 28 6.18 - 1.56 0.255 32 5.67 - 2.08 0.255 0.45 75 REP 1 0 7.73 0.00 0.447 1.5 7.56 - 0.18 0.447 3 7.62 - 0.11 0.447 4.5 7.12 - 0.61 0.447 6 7.01 - 0.72 0.447 7.5 7.18 - 0.56 0.447 9 6.62 - 1.11 0.447 10.5 6.44 - 1.29 0.447 12 6.49 - 1.24 0.447 REP 2 0 7.45 0.00 0.461 1.5 7.04 - 0.42 0.461 3 6.48 - 0.98 0.461 4.5 6.25 - 1.20 0.461 6 7.21 - 0.24 0.461 7.5 6.40 - 1.06 0.461 9 5.90 - 1.55 0.461 10.5 5.34 - 2.12 0.461 12 4.73 - 2.72 0.461 REP 3 0 7.72 0.00 0.449 1.5 7.17 - 0.56 0.449 3 7.04 - 0.68 0.449 4.5 7.01 - 0.72 0.449 171 Table F . 8 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.45 75 REP 3 6 6.56 - 1.16 0.449 7.5 6.75 - 0.97 0.449 9 6.74 - 0.98 0.449 10.5 6.32 - 1.41 0.449 12 6.39 - 1.33 0.449 0.65 75 REP 1 0 8.03 0.00 0.641 0.5 7.73 - 0.30 0.641 1 6.88 - 1.15 0.641 1.5 6.82 - 1.21 0.641 2 6.75 - 1.28 0.641 2.5 6.08 - 1.95 0.641 3 5.92 - 2.11 0.641 3.5 5.43 - 2.60 0.641 4 6.38 - 1.65 0.641 REP 2 0 7.57 0.00 0.634 1 6.97 - 0.61 0.634 2 5.88 - 1.69 0.634 3 6.29 - 1.29 0.634 4 5.16 - 2.41 0.634 5 4.60 - 2.98 0.634 6 3.85 - 3.73 0.634 8 4.51 - 3.07 0.634 REP 3 0 8.11 0.00 0.649 1 7.24 - 0.87 0.649 2 6.89 - 1.22 0.649 3 6.21 - 1.90 0.649 4 4.84 - 3.27 0.649 5 1.72 - 6.39 0.649 6 4.67 - 3.44 0.649 7 3.94 - 4.17 0.649 8 3.32 - 4.78 0.649 0.25 80 REP 1 0.00 8.40 0.00 0.244 0.67 8.11 - 0.29 0.244 1.33 8.02 - 0.39 0.244 2.00 8.15 - 0.25 0.244 2.67 8.13 - 0.28 0.244 3.33 7.89 - 0.52 0.244 4.67 7.56 - 0.84 0.244 REP 2 0.00 7.96 0.00 0.263 1.33 7.68 - 0.28 0.263 2.67 7.25 - 0.71 0.263 4.00 6.95 - 1.01 0.263 172 Table F . 8 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.25 80 REP 2 5.33 6.89 - 1.06 0.263 6.67 6.82 - 1.14 0.263 8.00 6.58 - 1.38 0.263 9.33 6.56 - 1.40 0.263 10.67 5.81 - 2.15 0.263 REP 3 0.00 7.41 0.00 0.255 1.33 6.88 - 0.52 0.255 2.67 6.53 - 0.88 0.255 4.00 6.51 - 0.90 0.255 6.08 6.59 - 0.82 0.255 6.67 6.71 - 0.70 0.255 8.00 6.18 - 1.23 0.255 9.33 5.71 - 1.69 0.255 10.67 5.97 - 1.43 0.255 0.45 80 REP 1 0.00 6.26 0.00 0.469 0.33 6.70 0.44 0.469 0.67 6.03 - 0.23 0.469 1.00 5.79 - 0.46 0.469 1.33 5.75 - 0.51 0.469 1.67 5.45 - 0.81 0.469 2.00 6.00 - 0.26 0.469 2.33 5.65 - 0.61 0.469 3.00 5.71 - 0.54 0.469 REP 2 0.00 7.12 0.00 0.450 0.33 6.63 - 0.48 0.450 0.67 7.08 - 0.03 0.450 1.00 6.79 - 0.32 0.450 1.33 6.14 - 0.98 0.450 2.00 6.24 - 0.88 0.450 2.33 6.38 - 0.73 0.450 2.67 5.84 - 1.28 0.450 REP 3 0.00 7.71 0.00 0.432 0.50 7.64 - 0.06 0.432 0.67 7.42 - 0.28 0.432 1.00 7.27 - 0.44 0.432 1.67 7.24 - 0.46 0.432 2.00 6.77 - 0.93 0.432 2.33 7.12 - 0.59 0.432 2.67 7.08 - 0.63 0.432 0.65 80 REP 1 0.00 4.97 0.00 0.641 0.33 4.11 - 0.86 0.641 0.67 4.56 - 0.41 0.641 173 Table F . 8 (cont d). Target a w Temp (°C) Rep Time (min) Log CFU/g Log N/N 0 Actual a w 0.65 80 REP 1 1.00 4.25 - 0.72 0.641 1.67 3.71 - 1.26 0.641 2.67 1.95 - 3.02 0.641 REP 2 0.00 5.58 0.00 0.634 0.33 4.42 - 1.16 0.634 1.00 4.75 - 0.83 0.634 1.67 1.00 - 4.58 0.634 REP 3 0.00 5.17 0.00 0.649 0.33 4.97 - 0.20 0.649 0.67 2.40 - 2.77 0.649 1.67 3.15 - 2.02 0.649 * Moisture content of date paste was measured in the different batches. ** Moisture content of 0.25, 0.45, and 0.65 a w date paste was 12.02, 13.48, and 21.70 %MC, respectively. 174 Figure F . 1 Isothermal Salmonella survival curves and log - linear model fit for almond products at (A) constant a w (0.45 a w ) with three different temperatures (80, 85, and 90 ° C), and (B) constant temperature (80 ° C) with three different a w (0.25, 0.45, and 0.65 a w ). 175 Figure F . 2 Isothermal Salmonella survival curves and log - linear model fit for wheat products at (A) constant a w (0.45 a w ) with three different temperatures (80, 85, and 90 ° C), and (B) constant temperature (80 ° C) with three different a w (0.25, 0.45, and 0.65 a w ). 176 Figure F . 3 Isothermal Salmonella survival curves and log - linear model fit for date products at (A) constant a w (0.45 a w ) with three different temperatures (70, 75, and 80 ° C), and (B) constant temperature (80 ° C) with three different a w (0.25, 0.45, and 0.65 a w ). 177 Figure F . 4 Isothermal Salmonella survival curves and log - linear model fit of (A) almond kernels, wheat kernels, and date pieces, (B) almond meal, wheat meal, and wheat flour, and (C) almond butter and date paste at constant a w (0.45 a w ) and temperature (80 ° C). 178 Matlab Codes for the GLM Re gression This appendix shows the example of MATLAB code used to fit the generalized linear model for almond products. %% Import data, format for GLM data=xlsread('data.xlsx'); y=data(:,2); %Log N is Response time=data(:,1); %X1 x1=time; temp=data(:,3); %X2 x2=temp; aw=data(:,4); %X3 x3=aw; structure=data(:,5);%X4 x4=structure; %Interaction effects x5=x1.*x2; %time*temp x6=x1.*x3; %time*aw x7=x1.*x4; %time*structure x8=x2.*x3; %temp*aw x9=x2.*x4; %temp*structure x10=x3.*x4; %aw*st ructure x11=x1.*x2.*x3; %time*temp*aw x12=x1.*x2.*x4; %time*temp*structure x13=x1.*x3.*x4; %time*aw*structure x14=x2.*x3.*x4; %temp*aw*structure X=[x1 x2 x3 x4 x5 x6 x7 x8 x9 x10 x11 x12 x13 x14]; %% GLM mdl = GeneralizedLinearModel.fit(X,y) 179 Matlab Codes for Model Fitting This appendix shows the example of MATLAB code used to fit log - linear, Weibull, and secondary models for almond kernels. Example of the log - linear model fitting for almond kernels inactivation (0.25 a w , 80 ° C) % data=xlsread ('input data_arrange.xlsx','almond kernel'); t=data(:,1); %time (min) logn=data(:,2); %log N (CFU/ml) temp=data(:,3); %heating temperature (c) aw=data(:,4); %aw %0.25aw and 80c t_25_80=t(1:27); logn_25_80=logn(1:27); temp_25_80=temp(1:27); aw_25_80=aw(1: 27); %Log - linear Model: log N(t) = - t/D + log N0 %0.25 80C beta0=[5 7]; % beta0= [initial D, initial logN0]; fname=@nonlinearDC; [beta,resids,J,COVB,mse] = nlinfit(t_25_80,logn_25_80,fname,beta0); D_25_80=beta; ci = nlparci(beta,resids,'jacobian',J); rmse=sqrt(mse); % AIC n=length(logn_25_80); K=3; logn_25_80_es = ( - t_25_80/beta(1))+ beta(2); residue = (logn_25_80 - logn_25_80_es); ss = sum(residue.^2); AIC=n*log(ss/n)+2*K+2*K*(K+1)/(n - K - 1); result1 = [beta(1) ci(1) ci(3) n rmse AIC]; result2 = [beta (2) ci(2) ci(4)]; disp('0.25 almond kernels at 80C'); disp(' D - VALUE CIL CIU n RMSE AIC'); disp(result1); disp(' Log N0 CIL CIU'); disp(result2); 180 function y = nonlinearD(beta,t) y= - t/beta(1) + beta(2); en d Example of the Weibull model fitting for almond kernels inactivation (0.25 a w , 80 ° C). % data=xlsread('input data_arrange.xlsx','almond kernel'); t=data(:,1); %time (min) logn=data(:,2); %log N (CFU/ml) temp=data(:,3); %heating temperature (c) aw=data(:,4); %aw %0.25aw and 80c t_25_80=t(1:27); logn_25_80=logn(1:27); temp_25_80=temp(1:27); aw_25_80=aw(1:27); %Weibull Model: log N(t) = - (t/delta)^p(shape parameter) + log N0 %0.25 80C beta0=[10 0.7 7]; % beta0= [initial D, initial logN0]; fna me=@WeibullDC; [beta,resids,J,COVB,mse] = nlinfit(t_25_80,logn_25_80,fname,beta0); D_25_80=beta; ci = nlparci(beta,resids,'jacobian',J); rmse=sqrt(mse); n=length(logn_25_80); K=4; ss = sum(resids.^2,'omitnan'); AIC=n*log(ss/n)+2*K+2*K*(K+1)/(n - K - 1); %E stimated 1 log reduction ts=linspace(min(t_25_80),max(t_25_80),1000); [ypred, delta] = nlpredci(fname,ts,beta,resids,J,0.05,'on','curve'); %confidence band for regression line CBu=ypred+delta; log_require=beta(3) - 1; % - 1 is a log reduction required xinterp=interp1(ypred, ts, log_require); Estimated_log_reduction = xinterp; CIU = interp1(CBu, ts, log_require); % c103 for CI at a log reduction required 181 Upper_CI_for_yobs = CIU; SE= (CIU - xinterp)/1.96; result1 = [beta(1) ci(1) ci(4) n rmse AIC]; result2 = [beta(2) ci(2) ci(5)]; result3 = [beta(3) ci(3) ci(6)]; disp('0.25 almond kernels at 80C'); disp(' Delta CIL CIU n RMSE AIC'); disp(result1); disp(' P CIL CIU'); disp(result2); disp(' Log N0 CIL CIU'); disp(result3); function y = WeibullDC(beta,t) y=( - 1.*(t./beta(1)).^beta(2))+beta(3); end Example of the secondary model fitting for almond kernels inactivation. clear; clc; data=xlsread('Normalization data.xlsx','almond kernel'); x=data(:,1); %time (min) yobs=data(:,2); %log N (CFU/ml) temp=data(:,3); %heating temperature (c) aw=data(:,4); %actual aw A =[x temp aw]; %% Secondary model % log D = log D ref - ((T - Tref)/ZT) - ((aw - awref)/Zaw) % Tref and aw ref were estimated beta0(1)=1; %log D ref beta0(2)=15; %ZT beta0(3)=1; %Zaw %% nlinfit for secondary model fnameINV=@DZ_fix; [beta,resids,J,COVB,mse] = nlinfit(A,yobs,fnameINV,beta0); beta ss=resids'*resids; n=length(x); p=length(beta); 182 rmse=sqrt(mse) condX=cond(J); detX TX=det(J'*J); %% confidence intervals for parameters ci=nlparci(beta, resids,J) %% R is the correlation matrix for the parameters, sigma is the standard error vector [R,sigma]=corrcov(COVB) SS=resids'*resids relstderr=sigma./beta corrdz=R(2,1); %correlation between Dr and zT %% AIC n=length(yobs); K=4; ss = sum(resids.^2,'omitnan'); AICc=n*log(ss/n)+2*K+2*K*(K+1)/(n - K - 1) function logn = DZ_fix(beta,X) % This function represents the secondary model t=X(:,1); temp =X(:,2); aw=X(:,3); %Tref and awref were optimized with smallest correlation between %parameters Tref = 81.4; awref = 0.451; Dvalue = beta(1) - ((temp - Tref )./beta(2)) - ((aw - awref )./beta(3)); D=10.^(Dvalue); logn= - t./D; end 183 Salmonella Population Reductions during Thermal Come - Up Time (Chapter 5) This appendix shows the reduction of Salmonella populations after samples were thermally treated and until the products reached the target temperature. Table I . 1 Salmonella population ( ± standard deviation ) reduction during the thermal come - up time for almond products. Products Salmonella population (log CFU/g) 80 ° C 85 ° C 90 ° C Almond kernels 0.25 a w 0.98 ± 0.26 1.30 ± 0.19 1.74 ± 0.13 0.45 a w 0.60 ± 0.16 0.42 ± 0.26 1.10 ± 0.68 0.65 a w 1.12 ± 0.15 1.47 ± 0.22 2.81 ± 0.77 Almond meal 0.25 a w 0.17 ± 0.08 0.25 ± 0.06 0.21 ± 0.12 0.45 a w 0.37 ± 0.36 0.51 ± 0.30 0.52 ± 0.20 0.65 a w 0.16 ± 0.11 0.34 ± 0.14 0.68 ± 0.20 Almond butter 0.25 a w 0.07 ± 0.37 0.07 ± 0.17 0.32 ± 0.19 0.45 a w 0.89 ± 0.84 1.01 ± 0.84 0.99 ± 0.58 0.65 a w 0.53 ± 0.17 0.86 ± 0.08 1.30 ± 0.22 184 Table I . 2 Salmonella population ( ± standard deviation ) reduction during the thermal come - up time for wheat products . Products Salmonella population (log CFU/g) 80 ° C 85 ° C 90 ° C Wheat kernels 0.25 a w 0.20 ± 0.35 0.21 ± 0.53 0.05 ± 0.54 0.45 a w 0.51 ± 0.61 0.68 ± 0.31 0.92 ± 0.60 0.65 a w 0.30 ± 0.12 0.37 ± 0.20 0.97 ± 0.19 Wheat meal 0.25 a w 0.07 ± 0.19 0.24 ± 0.16 0.32 ± 0.22 0.45 a w 0.02 ± 0.35 0.55 ± 0.27 1.38 ± 0.39 0.65 a w 0.70 ± 0.76 1.13 ± 0.25 3.04 ± 1.01 Wheat flour 0.25 a w 0.14 ± 0.14 0.19 ± 0.10 0.67 ± 0.14 0.45 a w 0.11 ± 0.08 0.68 ± 0.42 2.04 ± 0.23 0.65 a w 0.40 ± 0.12 1.25 ± 0.07 3.41 ± 0.42 Table I . 3 Salmonella population ( ± standard deviation ) reduction during the thermal come - up time for date products. Products Salmonella population (log CFU/g) 70 ° C 75 ° C 80 ° C Date pieces 0.25 a w 0.72 ± 0.12 0.11 ± 0.97 0.89 ± 0.67 0.45 a w 1.18 ± 0.47 0.92 ± 1.12 0.22 ± 0.45 0.65 a w 1.50 ± 1.18 1.59 ± 0.25 1.95 ± 0.62 Date paste 0.25 a w 0.08 ± 0.07 0.04 ± 0.07 0.16 ± 0.04 0.45 a w 0.22 ± 0.14 0.13 ± 0.05 0.75 ± 0.34 0.65 a w 0.29 ± 0.09 0.29 ± 0.12 2.89 ± 0.56 185 Shape Factor in Weibull Model This appendix shows the relationship between shape factor and temperature/a w for all products. Figure J . 1 Relationship of Weibull shape factor with (A) temperature (B) a w for almond kernels, almond meal, and almond butter. 186 Figure J . 2 Relationship of Wei bull shape factor with (A) temperature (B) a w for wheat kernels, wheat meal, and wheat flour. 187 Figure J . 3 Relationship of Weibull shape factor with (A) temperature (B) a w for date pieces and date paste. 188 Scaled Sensitivity Coefficient for the log - linear/Bigelow - Type Model Fi gure K . 1 SSC for the log - linear/ Bigelow - type model of almond kernels. Figure K . 2 SSC for the log - linear/ Bigelow - type model of almond meal. 189 Figure K . 3 SSC for the log - linear/Bigelow - type model of almond butter . Figure K . 4 SSC for the log - linear/Bigelow - type model of wheat kernels. 190 Figure K . 5 SSC for the log - linear/Bigelow - type model of wheat meal. Figure K . 6 SSC for the log - linear/Bigelow - type model of wheat flour. 191 Figure K . 7 SSC for the log - linear/Bigelow - type model of d ate pieces. Figure K . 8 SSC for the log - linear/Bigelow - type model of date paste. 192 Differential Scanning Calorimetry Figure L . 1 DSC thermogram of (A) almond butter (0.25, 0.45, and 0.65 a w ), (B) wheat flour (0.25, 0.45, and 0.65 a w ), and date paste (0.25, 0.45, and 0.65 a w ). A B C 193 Effects of the Fabrication Process on the Water Properties in Almo nd Products not Subjected to Complete Equilibration . To quantify the change in a w during the fabrication process, the almond kernels were equilibrated to 0.25, 0.45, and 0.65 a w and then milled into almond meal and almond butter products as described in Chapter 5. In addition, natural almonds (almonds stored at room temperature) were also fabricated using the same method. The a w and moisture content were measured using three repl icates for all of the products. The moisture content of the almonds in all of the products w ere stable ( P > 0.05) after being mill ed into almond meal and almond butter ( Table M . 1 ), except for the 0.25 a w sample ( P < 0.05), which was due to the sample needing a longer amount of time to come to equilibration. The 0.25 a w samples has the longest equilibration time because of the time required to decrease the water content (desorption) of the sample. The a w of the natural almonds (as received from supplier) was equivalent ( P > 0.05) for all stages of the milling process ; however, the a w of the almonds that were equilibrated to the targets of 0.25, 0.45, and 0.65 a w did change ( P < 0.05) after milling into meal and butter. The changes in a w after going through the milling process was likely due to non - uniform water distribution inside the large particles (i.e., incomplete equilibration) , which cause d the measured change in a w of the almond meal and butter. These results show that the fabricated samples required a re - equilibration process in order to get back to the target a w prior to the milling process. 194 Table M . 1 The a w and moisture content (± standard deviation) of almond kernels, almond meal, and almond butter fabricated from incompletely equilibrated almonds . Initial a w and product structure a w Moisture content (%) 0.25 a w Almond kernel s 0.24 4 ± 0.003 A 2.25 ± 0.07 A, B Almond meal 0.2 27 ± 0.013 B 2.05 ± 0.12 B Almond butter 0.3 50 ± 0.05 8 B 2.38 ± 0.16 A Natural a w Almond kernel s 0.44 0 ± 0.019 A 3.38 ± 0.23 A Almond meal 0.4 09 ± 0.018 A 3.16 ± 0.17 A Almond butter 0.42 0 ± 0.004 A 3.23 ± 0.13 A 0.45 a w Almond kernel s 0.44 2 ± 0.006 A 3.95 ± 0.44 A Almond meal 0.4 48 ± 0.019 A 3.84 ± 0.63 A Almond butter 0.3 62 ± 0.032 B 3.69 ± 0.14 A 0.65 a w Almond kernel s 0.66 5 ± 0.002 A 5.72 ± 0.33 A Almond meal 0.6 49 ± 0.012 A 5.19 ± 0.50 A Almond butter 0.57 5 ± 0.044 B 5.17 ± 0.59 A Within a column and at the same a w , values with a common superscript letter were not significantly different ( = 0.05). 195 Effect of Almond Skin Integrity on Salmonella Thermal Resistance This appendix was partially presented in a poster at the 2015 International Association for Food Protection (IAFP) Annual Meeting (Limcharoenchat et al., 2015) . To evaluate and quantify the effect of skin integrity on Salmonella thermal resistance, almonds were tested with their skin as either whole (fully intact), skin - damaged (partially intact), or blanched (absent). The skin - damaged almonds were produced using a vibratory tumbler (Model 67617, Central Machinery Inc., China) to shake the almond kernels (100 g) for 45 min. Silicon carbide sandpaper (Gri t #36, Rust - Oleum, Illinois, USA) w as glued inside the tumbler to partially remove the almond ski n. The blanched almonds were produced by placing raw almonds (100 g) in to hot water (100°C) for 1 min. The almond skins were then peeled off , excess water removed, and then the almonds were placed in a biosafety cabinet for 1 h ( air speed ~0.33 - 0.38 m/s) to dry . Whole , skin - damaged, and blanched almonds all were surface - inoculated with Sa lmonella Enteritidis PT 30 , equilibrated to ~0.4 0 a w , and thermally treated at 80°C as described in Chapter 4. Salmonella inactivation curves were calculated for each product ( Figure N . 1 ). The D 80°C ( ± standard error ) of the whole, skin - damaged, and blanched almond s were 20 ± 4.5 min, 19.2 ± 1.3 min, and 17.9 ± 3.9 min, respectively. The statistical analysis (ANOVA) result s indicated that the Salmonella thermal resistance on whole , skin - damaged, and blanched almonds were equivalent ( P > 0.05). These results indicate that skin integrity of an almond does not have an impact on Salmonella thermal resistance on almond surface. Si milar product structure may have similar influence on the thermal resistance. 196 Figure N . 1 S urvival (log CFU/g) of Salmonella Enteritidis PT30 during isothermal heating (~80 C) of whole, skin - damaged, and blanched almonds (~0.4 0 a w ). 0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50 60 70 80 Whole almond Skin-damaged almond Blanched almond Log CFU/g Time (min) 197 Effect s of Equilibration Protocol, W ater P roperties , and P roduct S tructure on Salmonella T hermal R esistance on/in A lmond Kernel s, A lmond M eal, and A lmond B utter This appendix was presented in a poster at the 2016 IAFP Annual Meeting (Limcharoenchat et al., 2016) . To quantify the effect s of water properties and product structure on Salmonella Enter itidis PT30 on/in almond kernels, almond meal, and almond butter , almond kernels were inoculated and partially or fully equilibrated to 0.25 a w ( Table O . 1 ) before testing Salmonella thermal resistance at 80°C as described in Chapter 5. Table O . 1 Definition of partial and full equilibration of almond kernels, almond meal, and almond butter. Sample Partial equilibration Full equilibration Almond kernels Surface a w of the almond was measured (~0.25 a w ). Surface a w of almond was measured before splitting the almond in half. Split almond a w (called internal a w ) was measured. Difference between surface and internal a w was < 0.04. Almond meal Equilibrated almond kernels (100 g) were ground (45 s) into meal. Almond meal was re - equilibrated in controlled - environment chambers (~2 days). Almond butter Equilibrated almond kernels (200 g) were milled (16 min) into butter. Dry ice added every 2 min to control Almond butter was re - equilibrated in controlled - environment chambers (~4 - 7 days). 198 The a w and moisture c ontent of the almond products before heating ( Table O . 2 ) indicate that the a w of the samples for partial equilibration was significantly different ( P < 0.05 ), but the moisture content was equivalent ( P > 0.05). These results show that the desorpti on of the whole almonds introduced the variation of the a w after fabrication. Table O . 2 The a w and moisture content (± standard deviation) before heating of almond kernels, almond meal, and almond butter after partial and full equilibration. Product a w Moisture content (%) Partial equilibration Almond kernel s 0.2 45 ± 0.011 B, C 4.07 ± 0.47 A Almond meal 0.28 5 ± 0.005 A 3.57 ± 0.22 A, B Almond butter 0.2 17 ± 0.047 C 3.38 ± 0.23 A, B Full equilibration Almond kernel s 0.25 4 ± 0.011 A, B 3.49 ± 0.05 A, B Almond meal 0.25 1 ± 0.010 A, B 2.20 ± 0.58 C Almond butter 0.25 1 ± 0.008 A, B, C 2.86 ± 0.05 B, C Within a column, values with a common superscript letter were not significantly different ( = 0.05). Salmonella thermal resistance ( Figure O . 1 and Table O . 3 ) of almond meal for partial equilibration was lower ( P < 0.05) than the thermal resistance at full equilibration because of the higher a w . These results suggest that the re - equilibration process was necessary for controlling the effect of a w on Salmonella thermal resistance . 199 Figure O . 1 S urvival ( log CFU/g) of Salmonella Enteritidis PT30 during isothermal heating (~80°C) of the almond kernels, meal, and butter after partial and full equilibration (~0.25 a w ). Table O . 3 D values (± standard deviation) determined by linear regression of the Salmonella survivor curves (Figure O .1) of the almond kernels, almond meal, and almond butter after partial and full equilibration (~0.25 a w ). Product D 80°C (min) Partial equilibration Almond kernel s 18.0 ± 4.2 C Almond meal 51.9 ± 8.7 B Almond butter 48.6 ± 3.5 B Full equilibration Almond kernel s 18.8 ± 2.6 C Almond meal 76.7 ± 13.2 A Almond butter 62.1 ± 6.9 A, B Parameters with the same superscript letter were not significantly different ( = 0.05). 200 Effect of the Type of Inactivation Container Used on Salmonella Thermal Resistance To quantify the effect that inactivation containers used in this study had on Salmonella Enteritidis PT30 (almond kernels, almond butter, wheat kernels, date pieces and date paste) samples were inoculated, equilibrated ( ~ 0.45 a w ), packed into test cells and plastic bags ( Table P . 1 ) before thermally treating the Salmonella at 80°C as described in Chapter 5. Table P . 1 Inactivation container loading for almond kernels, almond kernels, almond butter, wheat kernels, date pieces and date paste. Sample Test cell Plastic bag Almond kernels A single almond was cut into small pieces and loaded for full coverage in the test cell. A single almond was vacuum - packed in a plastic bag (See Chapter 5). Almond butter Almond butter was loaded into a test cell (See Chapter 5). Almond butter was loaded into a plastic bag before sealing (20 x 20 x 1 mm). Wheat kernels Seven wheat kernels were loaded into a test cell. Seven wheat kernels were vacuum - packed into a plastic b ag (See Chapter 5). Date pieces One date piece was loaded into a test cell. One date piece was vacuum - packed into a plastic bag (See Chapter 5). Date paste Date paste was loaded into a test cell (See Chapter 5). Using a test cell, date paste was shaped and sized before vacuum - packing in plastic bag. 201 Salmonella inactivation curves ( Figure P . 1 ) of each product type were compared by using the analysis of covariance (A NCOVA ) in MATLAB. Results suggested that Salmonella thermal resistance of almond butter, date pieces and date paste i n both containers were not significantly different ( P > 0.05); however, Salmonella thermal resistance of almond and wheat kernels in test cells were greater ( P < 0.05) than in plastic bags. However, the whole almond kernels actually had to be broken into pieces to fit into the test cells, and that they were no longer truly whole almonds. These results suggested that the inactivation container did not have an imp act on Salmonella thermal resistance when they have the same surface area of the samples in contact with the container . Figure P . 1 S urvival ( log CFU/g) of Salmonella Enteritidis PT30 during isothermal heating (~80°C) of (A) almond products (~0.25 a w ), (B) wheat kernels (~0.45 a w ), and (C) date products (~0.45 a w ). 202 R EFERENCES 203 R EFERENCES Abd, S. J., K. L. McCarthy, and L. J. Harris, 2012, Impact of storage time and temperature on thermal inactivation of Salmonella Enteritidis PT 30 on oil - roasted almonds . Journal of Food Science , v. 77, p. M42 - M47. Adams, M. R., and M. O. Moss, 2008, Food Microbiology . Cambridge, RSB Publishing. Almond Board of California, 2017, Almond 2017 annual report. Retrieved from http://newsroom.almonds.com/sites/de fault/files/pdf_file/Almond%20Almanac%202017. pdf . Almond Board of California, 2018a, Crunching on the safe side. Retrieved from http://www.almonds.com/consumers/about - almonds/food - safety . Almond Board of California, 2018b, Processing safe products. Retrieved from http://www.almonds.com/processors/processing - safe - product/pasteurization . Amirsha ghaghi, Z., K. Rezaei, and M. H. Rezaei, 2017, Characterization and functional properties of protein isolates from wild almond . Journal of Food Measurement and Characterization , v. 11, p. 1725 - 1733. Anderson, N. M., Y. Luo, and E. M. Grasso - Kelly, 2017, I mpact of temperature dependence of water activity on Salmonella inactivation in a multi - component food system. 7th Annual Food and Veterinary Medicine Science and Research Conference , College Park, MD, October 17 - 18, 2017. Barbosa - C novas, G., Anthony J. Fontana,Jr., Shelly J.Schmidt, Theodore P.Labuza,, 2007, Water activity in foods: fundamentals and applications . Ames, Iowa, Blackwell Pub . Beck, J. V., and K. J. Arnold, 1997, Parameter Estimation in Engineering and Science . Wiley, New York. Beuchat, L. R., and D. A. Mann, 2010, Factors affecting infiltration and survival of Salmonella on in - shell pecans and pecan nutmeats . Journal of Food Protection , v. 73, p. 1257 - 1268. Beuchat, L. R., and D. A. Mann, 2011a, Inactivation of Salmonella on in - shell pecans during conditioning treatments preceding cracking and shelling . Journal of Food Protection , v. 74, p. 588 - 602. Beuchat, L. R., and D. A. Mann, 2011b, Inactivation of Salmonella on pecan nutmeats by hot air treatment and oil roasting . Journal of Food Protection , v. 74, p. 1441 - 1450. 204 Beuchat, L. R., and D. A. Mann, 2014, Survival of Salmonella on dried fruits and in aqueous dried fruit homogenates as affected by temperature . Journal of Food Protection , v. 77, p. 1102 - 1109. Blessington, T., E. J. Mitcham, and L. J. Harris, 2012, Survival of Salmonella enterica , Escherichia coli O157:H7, and Listeria monocytogenes on inoculated walnut kernels during storage . Journal of Food Protection , v. 75, p. 245 - 254. Brar, P. K., L. G. Proano, L. M. Friedri ch, L. J. Harris, and M. D. Danyluk, 2015, Survival of Salmonella , Escherichia coli O157:H7, and Listeria monocytogenes on raw peanut and pecan kernels stored at - 24, 4, and 22 degrees C . Journal of Food Protection , v. 78, p. 323 - 332. Burnett, S. L., E. R. Gehm, W. R. Weissinger, and L. R. Beuchat, 2000, Survival of Salmonella in peanut butter and peanut butter spread . Journal of Applied Microbiology , v. 89, p. 472 - 477. Casulli, K. E., 2016, Improving pathogen - reduction validation meth ods for pistachio processing, Michigan State University. Centers for Disease Control and Prevention, 2004, Outbreak of Salmonella serotype Enteritidis infections associated with raw almonds --- United States and Canada, 2003 -- 2004. Retrieved from http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5322a8.htm . Centers for Disease Control and Prevention, 2007, Multistate outbreak of Salmonella serotype Tennessee infections associated with pe anut butter - United States, 2006 - 2007 (Reprinted from MMWR, vol 56, pg 521, 2007) . Journal of the American Medical Association , v. 298, p. 33 - 35. Centers for Disease Control and Prevention, 2016a, Multistate outbreak of Salmonella paratyphi B variant L( +) tartrate(+) infections linked to JEM raw brand sprouted nut butter spreads (final update). Retrieved from http://www.cdc.gov/salmonella/paratyphi - b - 12 - 15/index.html . Centers for Disease Control and Prevention, 2016b, Multistate outbreak of Salmonella montevideo and Salmonella senftenberg infections linked to wonderful pistachios (final update). Retrieved from http://www.cdc.gov/salmonella/montevideo - 03 - 16/index.html . Chao, C. T., and R. R. Krueger, 2007, The date palm ( Phoenix dactylifera L.): overview of biology, uses, and cultivation . Hortscience , v. 42, p. 1077 - 1082. Chen, Y. H., V. N. Scott, T. A. Frei er, J. Kuehm, M. Moorman, J. Meyer, T. Morille - Hinds, L. Post, L. Smoot, S. Hood, J. Shebuski, and J. Banks, 2009a, Control of Salmonella in low - moisture foods II: hygiene practices to minimize Salmonella contamination and growth . Food Protection Trends , v . 29, p. 435 - 445. 205 Chen, Y. H., V. N. Scott, T. A. Freier, J. Kuehm, M. Moorman, J. Meyer, T. Morille - Hinds, L. Post, L. Smoot, S. Hood, J. Shebuski, and J. Banks, 2009b, Control of Salmonella in low - moisture foods III: process validation and environmental monitoring - part three of a three - part series . Food Protection Trends , v. 29, p. 493 - 508. Chukwu, O., 2010, Moisture - sorption study of dried date fruits . AU Journal of Technology , v. 13, p. 175 - 180. Chung, H. J., S. L. Birla, and J. Tang, 2008, Performance evaluation of aluminum test cell designed for determining the heat resistance of bacterial spores in foods . Lwt - Food Science and Technology , v. 41, p. 1351 - 1359. Crumrine, M. H., and V. D. Foltz, 1969, Survival of Salmonella Montevideo on whea t stored at constant relative humidity . Applied Microbiology , v. 18, p. 911 - 914 . D'Souza, T., M. Karwe, and D. W. Schaffner, 2012, Effect of high hydrostatic pressure and pressure cycling on a pathogenic Salmonella enterica serovar cocktail inoculated int o creamy peanut butter . Journal of Food Protection , v. 75, p. 169 - 173. Danyluk, M. D., A. R. Uesugi, and L. J. Harris, 2005, Survival of Salmonella Enteritidis PT 30 on inoculated almonds after commercial fumigation with propylene oxide . Journal of Food P rotection , v. 68, p. 1613 - 1622. Datta, A. K., 1993, Error estimates for approximate kinetic parameters used in food literature . Journal of Food Engineering , v. 18, p. 181 - 199. Dolan, K. D., and D. K. Mishra, 2013, Parameter estimation in food science . Annual Review of Food Science and Technology , Vol 4, v. 4, p. 401 - 422. Dolan, K. D., V. P. Valdramidis, and D. K. Mishra, 2013, Parameter estimation for dynamic microbial inactivation: which model, which precision? . Food Control , v. 29, p. 401 - 408. Doyle , M. E., and A. S. Mazzotta, 2000, Review of studies on the thermal resistance of salmonellae . Journal of Food Protection , v. 63, p. 779 - 795. Du, W. - X., S. J. Abd, K. L. McCarthy, and L. J. Harris, 2010, Reduction of Salmonella on inoculated almonds expos ed to hot oil . Journal of Food Protection , v. 73, p. 1238 - 1246. Eglezos, 2010, Microbiological quality of wheat grain and flour from two mills in Queensland, Australia . Journal of Food Protection , v. 73, p. 1962 - 1962. Finn, S., O. Condell, P. McClure, A. Amezquita, and S. Fanning, 2013, Mechanisms of survival, responses, and sources of Salmonella in low - moisture environments . Frontiers in Microbiology , v. 4, p. 15. 206 Gaillard, S., I. Leguerinel, and P. Mafart, 1998, Mode l for combined effects of temperature, pH and water activity on thermal inactivation of Bacillus cereus spores . Journal of Food Science , v. 63, p. 887 - 889. Garcés - Vega, F. J., 2017, Quantifying water effects on thermal inactivation of Salmonella in low - mo isture foods, Michigan State University. Gurtler, J. B., M. P. Doyle, and J. L. Kornacki, 2014, The microbiological safety of low water activity foods and spices, New York, Springer. Harris, L. J., A. R. Uesugi, S. J. Abd, and K. L. McCarthy, 2012, Survi val of Salmonella Enteritidis PT 30 on inoculated almond kernels in hot water treatments . Food Research International , v. 45, p. 1093 - 1098. He, Y. S., D. J. Guo, J. Y. Yang, M. L. Tortorello, and W. Zhang, 2011, Survival and heat resistance of Salmonella enterica and Escherichia coli O157:H7 in peanut butter . Applied and Environmental Microbiology , v. 77, p. 8434 - 8438. He, Y. S., Y. Li, J. K. Salazar, J. Y. Yang, M. L. Tortorello, and W. Zhang, 2013, Increased water activity reduces the thermal resistance of Salmonella enterica in peanut butter . Applied and Environmental Microbiology , v. 79, p. 4763 - 4767. Hildebrandt, I. M., B. P. Marks, E. T. Ryser, R. Villa - Rojas, J. Tang, F. J. Garces - Vega, and S. E. Buchholz, 2016, Effects of inoculation procedures on variability and repeatability of Salmonella thermal resistance in wheat flour . Journal of Food Protecti on , v. 79, p. 1833 - 1839. Isaacs, S., J. Aramini, B. Ciebin, J. A. Farrar, R. Ahmed, D. Middleton, A. U. Chandran, L. J. Harris, M. Howes, E. Chan, A. S. Pichette, K. Campbell, A. Gupta, L. Y. Lior, M. Pearce, C. Clark, F. Rodgers, F. Jamieson, I. Brophy, A. Ellis, and P. T. O. Salmonella Enteritidis, 2005, An international outbreak of salmonellosis associated with raw almonds contaminated with a rare phage type of Salmonella Enteritidis . Journal of Food Protection , v. 68, p. 191 - 198. Jankowski, T., and C. K. Rha, 1986, D ifferential scanning calorimetry study of the wheat - grain cooking process. Starch - Starke , v. 38, p. 45 - 48. Jeong, S., B. P. Marks, and A. Orta - Ramirez, 2009, Thermal inactivation kinetics for Salmonella Enteritidis PT30 on almonds subjecte d to moist - air convection heating . Journal of Food Protection , v. 72, p. 1602 - 1609. Kataoka, A., Enache, E., Black, D.G., Podolak, R., and Hayman, M., 2014, Survival of Salmonella Tennessee, Salmonella Typhimurium DT104 , and En terococcus faecium in peanut paste formulations at two different levels of water activity and fat . Journal of Food Protection . , v. 77, p. 1252 - 1259. 207 Keller, S. E., E. M. Grasso, L. A. Halik, G. J. Fleischman, S. J. Chirtel, and S. F. Grove, 2012, Effect of growth on the thermal res istance and survival of Salmonella Tennessee and Oranienburg in peanut butter, measured by a new thin - layer thermal death time device . Journal of Food Protection , v. 75, p. 1125 - 1130. Keller, S. E., J. M. VanDoren, E. M. Grasso, and L. A. Halik, 2013, Gro wth and survival of Salmonella in ground black pepper (Piper nigrum) . Food Microbiology , v. 34, p. 182 - 188. Kimber, M. A., K. Harbir, W. Luxin, M. D. Danyluk, and L. J. Harris, 2012, Survival of Salmonella , Escherichia coli O157:H7, and Listeria monocytog enes on inoculated almonds and pistachios stored at 19, 4, and 24°C . Journal of Food Protection , v. 75, p. 1394 - 1403. Laroche, A., F. Fine, and P. Gervais, 2005, Water activity affects heat resistance of microorganisms in food powders . International Journ al of Food Microbiology , v. 97, p. 307 - 315. Lee, J. W., L. C. Thomas, and S. J. Schmidt, 2011, Can the thermodynamic melting temperature of sucrose, glucose, and fructose be measured using rapid - scanning differential scanning calorimetry (DSC)? . Journal o f Agricultural and Food Chemistry , v. 59, p. 3306 - 3310. Li, C. C., L. H. Huang, and J. Q. Chen, 2014a, Comparative study of thermal inactivation kinetics of Salmonella spp. in peanut butter and peanut butter spread . Food Control , v. 45, p. 143 - 149. Li, H. P., X. W. Fu, Y. G. Bima, J. Koontz, C. Megalis, F. Yang, G. Fleischman, and M. L. Tortorello, 2014b, Effect of the local microenvironment on survival and thermal inactivation of Salmonella in low - and intermediate - moisture multi - ingredient foods . J ournal of Food Protection , v. 77, p. 67 - 74. Limcharoenchat, P., M. James, N. Hall, and B. Marks, 2016, Moisture equilibration and product fabrication methods affect measured thermal resistance of Salmonella Enteritidis PT30 on/in whole almonds, almond mea l, and almond butter . Journal of Food Protection , v. 79, p. 148 - 148. Limcharoenchat, P., B. Marks, E. Ryser, M. James, and N. Hall, 2015, Comparing the effect of product structure on thermal resistance of Salmonella Enteritidis PT30 on/in almond and wheat products: Journal of Food Protection , v. 78, p. 97 - 97. Ma, L., G. D. Zhang, P. Gerner - Smidt, V. Mantripragada, I. Ezeoke, and M. P. Doyle, 2009, Thermal inactivation of Salmonella in peanut butter . Journal of Food Protection , v. 72, p. 1596 - 1601. 208 Mattick, K. L., F. Jorgensen, P. Wang, J. Pound, M. H. Vandeven, L. R. Ward, J. D. Legan, H. M. Lappin - Scott, and T. J. Humphrey, 2001, Effect of challenge temperature and solute type on heat tolerance of Salmonella serovars at low water activity . Applied and Environmental Microbiology , v. 67, p. 4128 - 4136. McCallum, L., S. Paine, K. Sexton, M. Dufour, K. Dyet, M. Wilson, D. Campbell, D. Bandaranayake, and V. Hope, 2013, An outbreak of Salmonella Typhimurium phage type 42 associated with the consumption of raw flour . Foodborne Pathogens and Disease , v. 10, p. 159 - 164. McKellar, R. C., and X. Lu, 2004, Modeling microbial responses in food . Boca Raton, FL, CRC Press. Mogollon, M. A., B. P. Marks, A. M. Booren, A. Orta - Ramirez, and E. T. Ryser, 2009, Effect of beef product physical structure on Salmonella thermal inactivation . Journal of Food Science , v. 74, p. M347 - M351. Moreira, R., F. Chenlo, M. D. Torres, and D. M. Prieto, 2010, Water adsorption and desorption isotherms of chestnut and wheat flours . Indu strial Crops and Products , v. 32, p. 252 - 257. Motulsky, H., and A. Christopoulos, 2004, Fitting models to biological data using linear and nonlinear regression: A practical guide to curve fitting: Oxford, New York, Oxford University Press. News desk, 201 5, USDA: Salmonella tops list of 15 most costly pathogens. January 7, 2015. North American Miller's Association, 2016, Wheat milling process. Retrieved from http://www.namamillers.org/education/wheat - milling - process/ . Okos, M. R., O. Camplanella, G. Narsimhan, R. K. Singh, and A. C. Weitnauer, 2007, Food dehydration . Florida, CRC Press. Pahlevanzadeh, H., and M. Yazdani, 2005, Moisture adsorption isotherms a nd isosteric energy for almond . Journal of Food Process Engineering , v. 28, p. 331 - 345. Peleg, M., 2006, Advanced quantitative microbiology for foods and biosystems: models for predicting growth and inactivation . Boca Raton, Taylor & Francis. Riggs, N., 2015, A date with Medjool. Retrieved from http://www.growingmagazine.com/fruits/a - date - with - medjool . Russo, E. T., G. Biggerstaff, R. M. Hoekstra, S. Meyer, N. Patel, B. Miller, R. Quick, and T. Salmonella Agona Outbreak Invest, 2013, A recurrent, multistate outbreak of Salmonella serotype Agona infections associated with dry, unsweetened cereal consumpti on, United States, 2008 . Journal of Food Protection , v. 76, p. 227 - 230. 209 Sanchez - Zapata, E., J. Fernandez - Lopez, M. Penaranda, E. Fuentes - Zaragoza, E. Sendra, E. Sayas, and J. A. Perez - Alvarez, 2011, Technological properties of date paste obtained from dat e by - products and its effect on the quality of a cooked meat product . Food Research International , v. 44, p. 2401 - 2407. Santillana - Farakos, S. M., J. F. Frank, and D. W. Schaffner, 2013, Modeling the influence of temperature, water activity and water mobi lity on the persistence of Salmonella in low - moisture foods . International Journal of Food Microbiology , v. 166, p. 280 - 293. Santillana - Farakos, S. M., J. W. Hicks, J. G. Frye, and J. F. Frank, 2014a, Relative survival of four serotypes of Salmonella ente rica in low - water activity whey protein powder held at 36 and 70 degrees C at various water activity levels . Journal of Food Protection , v. 77, p. 1198 - 1200. Santillana - Farakos, S. M., R. Pouillot, and S. E. Keller, 2017, Salmonella survival kinetics on pecans, hazelnuts, and pine nuts at various water activities and temperatures . Journal of Food Protection , v. 80, p. 879 - 885. Santillana - Farakos, S. M., D. W. Schaffner, and J. F. Frank, 2014b, Predicting survival of Salmonella in low - water activity foods : an analysis of literature data . Journal of Food Protection , v. 77, p. 1448 - 1461. Scharff, R. L., J. Besser, D. J. Sharp, T. F. Jones, P. Gerner - Smidt, and C. W. Hedberg, 2016, An economic evaluation of PulseNet a network for food borne disease surveilla nce . American Journal of Preventive Medicine , v. 50, p. S66 - S73. Schwaab, M., and J. C. Pinto, 2007, Optimum reference temperature for reparameterization of the Arrhenius equation. part 1: problems involving one kinetic constant . Chemical Engineering Scie nce , v. 62, p. 2750 - 2764. Scott, V. N., Y. U. H. Chen, T. A. Freier, J. Kuehm, M. Moorman, J. Meyer, T. Morille - Hinds, L. Post, L. Smoot, S. Hood, J. Shebuski, and J. Banks, 2009, Control of Salmonella in low - moisture foods I: minimizing entry of Salmonel la into a processing facility . Food Protection Trends , v. 29, p. 342 - 353. Shachar, D., and S. Yaron, 2006, Heat tolerance of Salmonella enterica serovars Agona, Enteritidis, and Typhimurium in peanut butter . Journal of Food Protection , v. 69, p. 2687 - 2691 . Sheth, A. N., M. Hoekstra, N. Patel, G. Ewald, C. Lord, C. Clarke, E. Villamil, K. Niksich, C. Bopp, T. A. Nguyen, D. Zink, and M. Lynch, 2011, A national outbreak of Salmonella serotype Tennessee infections from contaminated peanut butter: a new food v ehicle for salmonellosis in the United States . Clinical Infectious Diseases , v. 53, p. 356 - 362. 210 Shrestha, S., and B. Nummer, 2016, Survival of Salmonella spp. in low water activity chicken base paste and powder formulated at different salt levels . Food Co ntrol , v. 59, p. 663 - 668. Smith, D. F., I. M. Hildebrandt, K. E. Casulli, K. D. Dolan, and B. P. Marks, 2016, Modeling the effect of temperature and water activity on the thermal resistance of Salmonella Enteritidis PT 30 in wheat flour . Journal of Food Protection , v. 79, p. 2058 - 2065. Smith, D. F., and B. P. Marks, 2015, Effect of rapid product desiccation or hydration on thermal resistance of Salmonella enterica Serovar Enteritidis PT 30 in wheat flour . Journal of Food Protection , v. 78, p. 281 - 286. Syamaladevi, R. M., R. K. Tadapaneni, J. Xu, R. Villa - Rojas, J. M. Tang, B. Carter, S. Sablani, and B. Marks, 2016a, Water activity change at elevated temperatures and thermal resistance of Salmonella in all purpose wheat flour and peanut butter . Food Rese arch International , v. 81, p. 163 - 170. Syamaladevi, R. M., J. M. Tang, R. Villa - Rojas, S. Sablani, B. Carter, and G. Campbell, 2016b, Influence of water activity on thermal resistance of microorganisms in low - moisture foods: a review . Comprehensive Review s in Food Science and Food Safety , v. 15, p. 353 - 370. Tadapaneni, R. K., R. M. Syamaladevi, R. Villa - Rojas, and J. Tang, 2017, Design of a novel test cell to study the influence of water activity on the thermal resistance of Salmonella in low - moisture foo ds . Journal of Food Engineering , v. 208, p. 48 - 56. Tan, C. P., and Y. B. C. Man, 2000, Differential scanning calorimetric analysis of edible oils: Comparison of thermal properties and chemical composition . Journal of the American Oil Chemists Society , v. 77, p. 143 - 155. Tomasiewicz, D. M., D. K. Hotchkiss, G. W. Reinbold, R. B. Read, and P. A. Hartman, 1980, The most suitable number of colonies on plates for counting . Journal of Food Protection , v. 43, p. 282 - 286. Tuntivanich, V., A. Orta - Ramirez, B. P. Marks, E. T. Ryser, and A. M. Booren, 2008, Thermal inactivation of Salmonella in whole muscle and ground turkey breast . Journal of Food Protection , v. 71, p. 2548 - 2551. U.S. Department of Agriculture, 2007, Almonds grown in California; outgoing quality c ontrol requirements. Retrieved from https://www.gpo.gov/fdsys/pkg/FR - 2007 - 03 - 30/html/07 - 1557.htm . U.S. Department of Agriculture, 2014, Cost estimates of foodborne illnesses. Re trieved from https://www.ers.usda.gov/data - products/cost - estimates - of - foodborne - illnesses . 211 U.S. Department of Agriculture, 2016, Basic Report:09421, dates, medjool. Retrieved from https://ndb.nal.usda.gov/ndb/foods/show/2424 . U.S. Department of Agriculture, 2017, Noncitrus fruits and nuts. Retrieved from http://usda.mannlib.cornell.edu/usda/current/NoncFruiNu/NoncFruiNu - 06 - 27 - 2017.pdf . U.S. Department of Agriculture, 2018, Data by commodity. Retrieved from https://www.ers.usda.gov/data - products/fruit - and - tree - nut - data/data - by - commodity . U.S. Food and Drug Administration, 2014a, Be Salmonella safe! Retrieved from http://www.fda.gov/animalveterinary/resourcesforyou/animalhealthliteracy/ucm136197.h tm . U.S. Food and Drug Administration, 2014b, Gomacro recalls "almond butter + carob" and Retrieved from http://www.fda.gov/Safety/Recalls/ucm406804.htm . U.S. Food and Drug Administration, 2015, Mahina Mele farm recalls their macadamia nut products due to possible health risk. Retrieved from http://www.fda.gov/Safety/Recalls/ArchiveRecalls/2015/ucm457431.htm . U.S. Food and Drug Administration, 2016, Crescent Specialty Foods Inc. recalls raw pistachios because of possible health risk. Retrieved from http://www.fda.gov/Safety/Recalls/ucm507309.htm . U.S. Food and Drug Administration, 2015, Voluntary recall for Navajo pride bleached flour due to possible health risk. Retrieved from http://www.fda.gov/Safety/Recalls/ArchiveRecalls/2015/ucm445036.htm . U.S. Food and Drug Administration, 2016a, 2015 Recalls, market withdrawals and safety alerts. Retrieved from http://www.fda.gov/Safety/Recalls/ArchiveRecalls/2015/default.htm . U.S. Food and Drug Administration, 2016b, Williams - Sonoma announces the voluntary recall of Meyer lemon poppy quick bread. Retrieved from https://www.fda.gov/Safety/Recalls/ucm533821.htm . U.S. Food and Drug Administration, 2017, Lords Organics recalls ginger powder because of possible health risk. Retrieved from https://www.fda.gov/Safety/Recalls/ucm555723.htm . U.S. Wheat As sociates, 2016, World wheat supply and demand situation. Retrieved from http://www.uswheat.org/supplyDemand/doc/3FD27C5F60601DAB8525 800D0071F983/ $File/S&D%20160812.pdf?OpenElement# . U.S. Wheat Associates, 2018, Wheat letter - January 11, 2018. Retrieved from http://www.uswheat.org/wheatLetter/doc/ 88721B922F90F0F3852582120077496E/ . 212 Uesugi, A. R., M. D. Danyluk, and L. J. Harris, 2006, Survival of Salmonella Enteritidis phage type 30 on inoculated almonds stored at - 20, 49 23, and 35 degrees C . Journal of Food Protection , v. 69, p. 1851 - 1857. van B oekel, M., 2002, On the use of the Weibull model to describe thermal inactivation of microbial vegetative cells . International Journal of Food Microbiology , v. 74, p. 139 - 159. Velasquez, A., T. J. Breslin, B. P. Marks, A. Orta - Ramirez, N. O. Hall, A. M. Booren, and E. T. Ryser, 2010, Enhanced thermal resistance of Salmonella in marinated whole muscle compared with ground pork . Journal of Food Protection , v. 73, p. 372 - 375. Viaz is, S., J. K. Beal, C. Monahan, W. A. Lanier, K. R. Kreil, D. C. Melka, W. D. Boden, J. L. Dion, Z. A. Miller, T. A. Nguyen, L. B. Gieraltowski, and D. L. Zink, 2015, Laboratory, environmental, and epidemiologic investigation and regulatory enforcement act ions in response to an outbreak of Salmonella Bredeney infections linked to peanut butter . Open Forum Infectious Diseases , v. 2, p. 8. Villa - Rojas, R., J. Tang, S. J. Wang, M. X. Gao, D. H. Kang, J. H. Mah, P. Gray, M. E. Sosa - Morales, and A. Lopez - Malo, 2013, Thermal inactivation of Salmonella Enteritidis PT 30 in almond kernels as influenced by water activity . Journal of Food Protection , v. 76, p. 26 - 32. Weise, E., 2009, Salmonella outbreaks lead to food safety changes, US todays. Retrieved from http://www.usatoday.com/money/industries/food/2009 - 04 - 01 - nuts - Salmonella - food - safety_N.htm . Werber, D., J. Dreesman, F. Feil, U. van Treeck, G. Fell, S. Ethelberg, A., P. R. M. Hauri, R. Prager, I. S. T. Fisher, S. C. Behnke, E. Bartelt, A. E. E. Weise, A. Siitonen, Y. Andersson, H. Tscha¨pe, M. H. Kramer, and, and A. Ammon., 2005, International outbreak of Salmonella oranienburg due to German chocolate , BMC Infect. Retrieved from http://www.biomedcentral.com/1471 - 2334/5/7 . Zaitoon, A. M., A. M. Elansari, Y. S. Ahmed, and N. A. Taha, 2016, Phase transitions of samani date palm (Phoenix dactylife ra L.) fruit using differential scanning calorimeter (DSC) . Journal of Food and Agriculture , v. 28, p. 625 - 632. Zhang, H. M., Y. Qi, L. Wang, S. K. Zhang, and X. Y. Deng, 2017, Salmonella survival during thermal dehydration of fresh garlic and storage of dehydrated garlic products . International Journal of Food Microbiology , v. 263, p. 26 - 31.