ABSTRACT
Bacillus weihenstephanensis is a psychrotolerant bacterium belonging to the Bacillus cereus group. Some strains may be cytotoxic although they have not been described as food-poisoning agents so far. The objective of this work is to model the effects of temperature and a(w) downshifts on the lag time of B. weihenstephanensis and the dependence of µ(max) on the growth conditions (temperature and a(w)). Effects of temperature downshifts were studied on 30 experimental conditions (shifts magnitude ranging from 2 to 20 °C, temperature after shift from 10 to 20 °C and a(w) ranging from 0.977 to 0.997). Osmotic shifts were studied for 13 conditions (shift magnitude ranging from 0.008 to 0.020 units of a(w), temperature from 10 to 30 °C, a(w) after shift from 0.977 to 0.997). Experimental results show that temperature downshifts were able to induce considerable lag times (up to 20 days) when occurring near the growth limits. At lower temperatures, osmotic shifts had also a significant effect. Validation experiments in food subjected to changing conditions of temperature showed that the model provided valid predictions in diluted creamed pasta but overestimated bacterial growth in carrot soup (fail safe predictions).
Subject(s)
Bacillus/growth & development , Food Contamination , Temperature , Cold Temperature , Dairy Products/microbiology , Food Microbiology , Forecasting , Kinetics , Models, Theoretical , Osmotic PressureABSTRACT
A dynamic predictive model was developed to describe the effects of temperature, pH and NaCl concentration on the growth of Clostridium perfringens type A. The model for the specific growth rate was based on 81 growth curves generated in our laboratory or obtained from the publicly available ComBase database. Growth curves obtained during cooling were fitted with the dynamic model of Baranyi and Roberts. This made it possible to determine the parameter value reflecting the physiological state of C. perfringens after heating profiles typically applied to bulk meat. The model with the obtained parameters provided a good description of growth of C. perfringens in 24 heating/cooling curves generated specifically for this work (various non-isothermal treatments with a range of combinations of pH and NaCl concentration), and also for existing literature data. The dynamic model was implemented in Perfringens Predictor, a web-based application that can be accessed free of charge via www.combase.cc. It is anticipated that the use of this model and Perfringens Predictor will contribute to a reduction in the food poisoning incidence associated with C. perfringens.
