Behavior of Listeria monocytogenes in the presence or not of intentionally-added lactic acid bacteria during ripening of artisanal Minas semi-hard cheese

The fate of Listeria monocytogenes during ripening of artisanal Minas semi-hard cheese, as influenced by cheese intrinsic properties and by autochthonous (naturally present) or intentionally-added anti-listerial lactic acid bacteria (LAB) was modeled. Selected LAB strains with anti-listerial capacity were added or not to raw or pasteurized milk to prepare 4 cheese treatments. Counts of LAB and L. monocytogenes, pH, temperature and water activity were determined throughout cheese ripening (22 days, 22±1ᵒC). Different approaches were adopted to model the effect of LAB on L. monocytogenes: an independent approach using the Huang primary model to describe LAB growth and the linear decay model to describe pathogen inactivation; the Huang-Cardinal [pH] model using the effect of pH variation in a dynamic tertiary approach; and the Jameson-effect with Nmax tot model which simultaneously describes L. monocytogenes and LAB fate. L. monocytogenes inactivation occurred in both treatments with added LAB and inactivation was faster in raw milk cheese (−0.0260 h−1) vs. pasteurized milk cheese (−0.0182 h−1), as estimated by the linear decay model. Better goodness-of-fit was achieved for the cheeses without added LAB when the Huang primary model was used. A faster and great pH decline was detected for cheeses with added LAB, and the Huang-Cardinal [pH] model predicted higher pathogen growth rate in cheese produced with raw milk, but greater L. monocytogenes final concentration in pasteurized milk cheese. The Jameson-effect model with Nmax tot predicted that LAB suppressed pathogen growth in all treatments, except in the treatment with pasteurized milk and no LAB addition. The Huang-Cardinal [pH] model was more accurate in modeling L. monocytogenes kinetics as a function of pH changes than was the Jameson-effect model with Nmax tot as a function of LAB inhibitory effect based on the goodness-of-fit measures. The Jameson-effect model may however be a better competition model since it can more easily represent L. monocytogenes growth and death. This study presents crucial kinetic data on L. monocytogenes behavior in the presence of competing microbiota in Minas semi-hard cheese under dynamic conditions. •Three different approaches were evaluated to predict L. monocytogenes behavior in semi-hard cheese.•All proposed models were capable of closely describing L. monocytogenes growth/death potential.•Huang-Cardinal [pH] present better fit quality than the Jameson-effect model.•Th