Baert Katleen, Valero Antonio, De Meulenaer Bruno, Samapundo Simbarashe, Ahmed Monzur Morshed, Bo Li, Debevere Johan, Devlieghere Frank
Department of Food Safety and Food Quality, Ghent University, Coupure Links 653, B-9000, Ghent, Belgium.
Int J Food Microbiol. 2007 Sep 15;118(2):139-50. doi: 10.1016/j.ijfoodmicro.2007.07.006. Epub 2007 Jul 13.
The objective of the present study was to develop validated models that describe the effect of storage temperature on the growth rate and lag phase of six Penicillium expansum strains. The growth of the selected strains was therefore studied on Apple Puree Agar Medium (APAM) at 30, 25, 16, 10, 4 and 2 degrees C. Growth rates and lag phases were estimated using linear regression. Several secondary models were evaluated and for the growth rate, a modification of the extended Ratkowsky model was selected. Regarding the lag phase, the Arrhenius-Davey model provided the best adjustment to the observed data. Model validation was performed in two steps. Firstly, the developed models were validated on APAM. The obtained bias factors (Bf) ranged from 0.91 to 1.14 and the accuracy factors (Af) were <1.2 for the validation performed on APAM, indicating that the models were good predictors of the true mean colony growth rate and lag phase. Afterwards, an external validation was carried out in apples. For the growth rate, Bf ranged from 0.64 to 0.81 and Af<1.39, indicating conservative predictions. On the contrary for the lag phase, a clear deviation was observed between predictions and observed values on apples (0.35<Bf<0.7 and Af>1.6). These results highlight that the use of simulation or synthetic media for the development of predictive models for the lag phase of moulds can lead to inadequate predictions and that a validation on the real food matrix is necessary. Application of the developed models is possible in the framework of Quantitative Risk Assessment to develop control strategies against blue mould rot in apple and enables the inclusion of strain variability. However, possible underestimation of the lag phase should be taken into account.
本研究的目的是建立经过验证的模型，以描述储存温度对六种扩展青霉菌株生长速率和延迟期的影响。因此，在30、25、16、10、4和2摄氏度的苹果泥琼脂培养基（APAM）上研究了所选菌株的生长情况。使用线性回归估计生长速率和延迟期。评估了几个二级模型，对于生长速率，选择了扩展的Ratkowsky模型的一种修正形式。关于延迟期，Arrhenius-Davey模型对观测数据提供了最佳拟合。模型验证分两步进行。首先，在APAM上对所建立的模型进行验证。在APAM上进行验证时，得到的偏差因子（Bf）范围为0.91至1.14，准确因子（Af）<1.2，这表明这些模型是真实平均菌落生长速率和延迟期的良好预测指标。之后，在苹果上进行了外部验证。对于生长速率，Bf范围为0.64至0.81，Af<1.39，表明预测较为保守。相反，对于延迟期，在苹果上的预测值与观测值之间观察到明显偏差（0.35<Bf<0.7且Af>1.6）。这些结果突出表明，使用模拟或合成培养基来建立霉菌延迟期预测模型可能会导致预测不足，并且有必要在实际食品基质上进行验证。所建立的模型可应用于定量风险评估框架，以制定针对苹果青霉腐烂的控制策略，并能够纳入菌株变异性。然而，应考虑延迟期可能被低估的情况。
