Numerical and Statistical Methodology to Analyze Microbial Spoilage of Refrigerated Solid Foods Exposed to Temperature Abuse

Numerical and statistical procedures based on pseudo-zero for the lag and first order reaction kinetics for the exponential growth phase were developed to analyze non-isothermal microbial spoilage. Arrhenius model parameters and their accuracy were estimated for a mixture of Pseudomonas fluorescens, Staphylococcus aureus andAchromobacter lwoffi growing in a seafood model. Linear regressions used with isothermal experiments generated initial values for nonlinear estimations of the frequency (K₀) and activiation energy (E_a) constants. An optimization technique was used to minimize the square difference between experimental and estimated values while parameter accuracy was assessed using a bootstrap method. E_a and In(K₀) were 109±3.4 and 48.3±1.5 kJ/mole for the exponential, and 152±4.0 and 64.4±1.7 kJ/mole for the lag phase, respectively. The Mann-Whitney-Wilcoxon rank sum test showed no significant differences between parameters generated by two different temperature profiles (5% significance level).