A logistic model describing lactic bacteria's ability to modify the organic acid concentration during white wine's malolactic fermentation in new and used French oak barrels

Introduction: Variation in wine's organic acids during malolactic fermentation (MLF) affects the wine’s quality. Manipulation of these parameters during MLF represents a challenge for the wine industry. Evaluation of the fermentation kinetics given the abovementioned parameters is important in order to obtain insights into the MLF process, which can help toward optimisation. Thus, MLF kinetics is utilised to evaluate the MLF ability of two bacteria strains in new and used oak barrels.

Methods: In the present study, the variation in different wine parameters such as pH, total and volatile acidity, and organic acid levels (tartaric, malic, lactic and citric acid) was evaluated during MLF in the same initial wine, utilising two commercial lactic bacteria strains. The fermentation was performed in new and used French oak barrels for each strain: citrate-negative Oenococcus oeni (CINEnew, CINEold) and Oenococcus oeni (CH35new, CH35old). The obtained experimental data were fitted to the logistic model for each parameter monitored.

Results: The degree of fitting R² was greater than 0.9891, revealing that the models are very good predictors of the experimental results for substrate consumption (tartaric acid, malic acid, citric acid), as well as product formation (lactic acid). Using the CINE strains in the used barrel produced significantly higher amounts lactic acid and caused the consumption of the highest amount of tartaric and citric acid compared to the rest of the strains. On the other hand, the combination CH35new resulted in the wine with the lowest amount of malic acid.

Conclusions: The logistic model reflected the change in the pH and content of acids during MLF fermentation very well. The results could adequately describe the changes in the levels of organic acids during MLF fermentation in the barrels. This modelling may be particularly useful for forecasting in industrial-scale production.