Résumé
Predicting microbial safety of fresh products in Modified Atmosphere Packaging (MAP) systems implies to take into account the dynamic of O2 and CO2 exchanges in the system and its effect on microbial growth. In this purpose we coupled mathematical models of gas transfer (permeation through packaging and solubilisation / diffusion within food) with predictive microbiology models that take into account the effect of CO2 and O2 partial pressure in headspace and corresponding dissolved concentrations in the food. This mechanistic model was validated in simplified and in real conditions using dedicated challenge-tests performed on poultry meat, fresh salmon and processed cheese, inoculated with either Listeria monocytogenes or Pseudomonas fluorescens.Once validated, this model could be used as a Decision Support Tool in order to optimize the initial packaging atmosphere (level of O2 and CO2) and / or the geometry (ratio headspace volume to food mass). This tool could also be used to identify the packaging gas permeability the most suitable for maintaining the targeted % of gas initially flushed in the pack within a given tolerance. This approach permits a better dimensioning of MAP of fresh produce by selecting the packaging material fitted to “just necessary” (and not by default the most barrier one). The connexion of this model with dedicated databases gathering gas permeabilities of commonly used packaging materials allows us to obtain as output a ranking of the most suitable materials. This tool would be very useful for all stakeholders of the fresh produce chain. A demonstration of this Decision Support Tool is here proposed with the pipeline between mathematical models and related databases.