Probabilistic model for the spoilage wine yeast Dekkera bruxellensis as a function of pH, ethanol and free SO2 using time as a dummy variable

Abstract

The present study uses a probabilistic model to determine the growth/no growth interfaces of the spoilage wine yeast Dekkera bruxellensis CH29 as a function of ethanol (10-15%, v/v), pH (3.4-4.0) and free SO2 (0-50mg/l) using time (7, 14, 21 and 30days) as a dummy variable. The model, built with a total of 756 growth/no growth data obtained in a simile wine medium, could have application in the winery industry to determine the wine conditions needed to inhibit the growth of this species. Thereby, at 12.5% of ethanol and pH3.7 for a growth probability of 0.01, it is necessary to add 30mg/l of free SO2 to inhibit yeast growth for 7days. However, the concentration of free SO2 should be raised to 48mg/l to achieve a probability of no growth of 0.99 for 30days under the same wine conditions. Other combinations of environmental variables can also be determined using the mathematical model depending on the needs of the industry. © 2013 Elsevier B.V.