Lebert I, Dussap C G, Lebert A
Unité de Recherches sur la Viande, Institut National de la Recherche Agronomique, 63122 Saint-Genès Champanelle, France.
Int J Food Microbiol. 2005 Jul 25;102(3):305-22. doi: 10.1016/j.ijfoodmicro.2004.11.021.
The quality and safety of food products depend on the microorganisms, the food characteristics and the process. The prediction of conditions that prevent growth in complex situations due to the characteristics of the process and of the food cannot be obtained by predictive models of bacterial growth only. Thus, a combined modelling approach was developed by integrating three models, which were selected in a first step: (1) a bacterial model that predicts the bacterial growth from the physico-chemical properties of the media; (2) a water transfer model that predicts the effects of the drying process variables on the medium characteristics; and (3) a thermodynamic model that predicts the water activity aw and the pH of the media from its composition. A second step consisted in separately validating each selected model in which all of the physical, chemical or biological parameters appearing in the equations were previously measured. The third step combined the three knowledge models. The global model was validated on the basis of experimental results concerning the growth of Listeria innocua on the surface of a gelatine gel, the surface of which was submitted to a drying process (changes in relative humidity and air velocity). It was shown that bacterial growth models had to be modified: a specific model was set up to predict the maximum growth rate and another for the lag. Additionally, growth models set up in broth could not be applied in gelatine, leading to the development of a specific growth model on a solid surface. The thermodynamic model accurately predicted the pH and aw of bacterial broth in which high concentrations of solutes were added, and those of the solid media, the gelatine. The water transfer model was applied on gelatine data to predict the evolution of its surface aw during the drying process. The three models-bacterial, water transfer and thermodynamic, separately validated-were combined according to an integrated modelling strategy. The water transfer model coupled with the thermodynamic model predicted the aw on the gel surface. The predicted surface aw explained why growth inhibition was observed. Indeed, growth stopped at a predicted surface aw <0.94, corresponding to L. innocua minimum aw during the drying process. The global model satisfactorily predicted L. innocua growth on the surface of the gel. This study proves the validity of the approach and shows that the combination of the water transfer and thermodynamic models compensates for the lack of aw measurement techniques.
食品的质量和安全取决于微生物、食品特性以及加工过程。仅靠细菌生长预测模型无法得出因加工过程和食品特性而在复杂情况下防止微生物生长的条件。因此，通过整合三个模型开发了一种组合建模方法，这三个模型是在第一步中选定的：（1）一个细菌模型，可根据培养基的物理化学性质预测细菌生长；（2）一个水分转移模型，可预测干燥过程变量对培养基特性的影响；（3）一个热力学模型，可根据培养基成分预测水分活度aw和pH值。第二步是分别验证每个选定的模型，其中方程中出现的所有物理、化学或生物学参数都事先进行了测量。第三步是将这三个知识模型结合起来。基于关于无害李斯特菌在明胶凝胶表面生长的实验结果对全局模型进行了验证，该明胶凝胶表面经历了干燥过程（相对湿度和风速的变化）。结果表明，细菌生长模型必须进行修改：建立了一个特定模型来预测最大生长速率，另一个用于预测延迟期。此外，在肉汤中建立的生长模型不能应用于明胶，因此开发了一个在固体表面的特定生长模型。热力学模型准确预测了添加高浓度溶质的细菌肉汤以及固体培养基明胶的pH值和aw值。水分转移模型应用于明胶数据，以预测干燥过程中其表面aw的变化。根据综合建模策略，将分别经过验证的细菌、水分转移和热力学这三个模型结合起来。水分转移模型与热力学模型相结合预测了凝胶表面的aw值。预测的表面aw值解释了为何观察到生长抑制现象。实际上，当预测的表面aw <0.94时生长停止，这对应于干燥过程中无害李斯特菌的最低aw值。全局模型令人满意地预测了无害李斯特菌在凝胶表面的生长情况。这项研究证明了该方法的有效性，并表明水分转移模型和热力学模型的结合弥补了aw测量技术的不足。
