Predictive modelling and validation of Pseudomonas fluorescens growth at superatmospheric oxygen and carbon dioxide concentrations

The model parameters were estimated in a one-step nonlinear least-squares regression. Two overall models were compared: (1) the Baranyi equation with a first-order linear response surface model to describe μmax and λ as a function of O2 and CO2 and (2) the Baranyi equation with insertion of the first-order equation for μmax and with λ linearly related to μmax-1. Both models described the growth very well. To assess the validity of each overall model, an additional experiment with six gas conditions was carried out. The best correspondence between observed and predicted growth data was obtained with model 1, as based on different validation parameters (MSE, adjusted R2, Bias and Accuracy factors). The 95% confidence limits for a new prediction were estimated for each validation gas condition using a Monte Carlo procedure.