In this study we investigated the effect of the single strain in stabilization of type I sourdough microbial associations by crossing six different Fructilactobacillus sanfranciscensis with five Kazachstania humilis strains. Furthermore, we compared three predictive models, Zwitwering based on Gompertz's equation, Baranyi and Roberts' function and Schiraldi's function to evaluate which one best fitted the experimental data in determining the behaviour of co-cultivated microorganisms. Specific growth rates (μ_m) and lag time (λ) values for each mixed population were assessed. Results showed that the different F. sanfranciscensis strains significantly steer the growth kinetics within the pair and affect the ratio bacterial/yeast cells, as data analysis confirmed, whereas K. humilis accommodates to the bacterial strain. To compare the growth models, Root Mean Square (RMS) values were calculated for each predicted curve by implementing an algorithm based on an iterative process to minimize the deviation among observed and calculated data. Schiraldi's function performed better than the others, revealing, on average, the smallest RMS values and providing the best fitting for over 70% of co-cultivation experiments. Models prove to be consistent in predicting growth kinetics of microbial consortia too.

在这项研究中，我们通过将六种不同的三藩果乳杆菌与五种Kazachstania humilis菌株杂交，研究了单一菌株在稳定 I 型酸面团微生物关联方面的作用。此外，我们比较了三种预测模型、基于 Gompertz 方程的 Zwitwering、Baranyi 和 Roberts 函数以及 Shiraldi 函数，以评估哪一个最适合实验数据来确定共培养微生物的行为。评估了每个混合群体的特定生长率 (μ _m ) 和滞后时间 (λ) 值。结果表明，不同的F. sanfranciscensis正如数据分析所证实的那样，K. humilis菌株可显着控制该对内的生长动力学并影响细菌/酵母细胞的比例，而K. humilis 则适应细菌菌株。为了比较生长模型，通过实施基于迭代过程的算法来计算每条预测曲线的均方根 (RMS) 值，以最大限度地减少观测数据和计算数据之间的偏差。Schiraldi 的函数比其他函数表现更好，平均而言，显示最小的 RMS 值，并为超过 70% 的共培养实验提供最佳拟合。模型在预测微生物群落的生长动力学方面也被证明是一致的。