Evaluating non-conventional yeast for the production of wines that contain less ethanol


There is growing interest from winemakers in being able to produce wines with lower ethanol content that do not have compromised aroma, flavour, and mouth-feel. There are opportunities across the value chain to implement strategies to achieve this, including viticultural practices, pre-fermentation and winemaking practices, microbiological strategies and post-fermentation practices and processing technologies (Varela et al. 2015).

However, the application of yeast strains that produce less ethanol during fermentation remains a simple and cheap strategy for producers to adopt towards this goal. Unfortunately, all available commercial S. cerevisiae wine yeasts are very similar in terms of ethanol yield; a difference of 0.5% v/v has been observed between ‘high’ and ‘low’ ethanol producers (Palacios et al. 2007; Varela et al. 2008). In order to reduce ethanol concentration, novel yeast need to be isolated or generated.


From the 1980s to the mid-2000s, the average ethanol concentration in Australian wine increased, in part reflecting market trends for wine styles associated with increased grape maturity. Since 2005, ethanol levels have trended downwards but are still significantly above those in the 1980s (Godden et al. 2015). High ethanol concentration can, however, adversely affect wine sensory properties, reducing the perceived complexity of flavours and aromas. In addition, for reasons associated with corporate social responsibility and taxation regimes the wine sector is actively seeking technologies that facilitate the production of wines with lower ethanol content. Non-conventional yeasts have shown potential for production of wines with lower alcohol concentration. These yeast species, largely associated with grapes pre-harvest, are usually present in the early stages of fermentation but in general are not capable of completing alcoholic fermentation unaided.

Recently, the AWRI identified that a Metschnikowia pulcherrima strain was able to produce wine with reduced ethanol concentration when sequentially inoculated with a wine strain of Saccharomyces cerevisiae. Later, the AWRI identified a Saccharomyces uvarum strain which was also able to produce wine with reduced ethanol concentration. This effect was additive when both strains M. pulcherrima and S. uvarum were co-inoculated in sterilised or dimethyl dicarbonate (DMDC) treated grape must at laboratory-scale. In freshly prepared grape musts both yeasts were less effective in reducing alcohol concentration due to competition with indigenous yeast populations. Subsequent pilot-scale winemaking was, therefore, performed using DMDC-treated musts. Several treatments produced wines with lower ethanol concentration than control S. cerevisiae wines. Formal sensory analysis revealed that while wines fermented with S. uvarum, alone or in combination with M. pulcherrima, were lower in alcohol concentration, they were associated with negative sensory attributes. Wines fermented with M. pulcherrima were associated with positive sensory attributes and were slightly lower in alcohol concentration than uninoculated controls. Overall the results suggest that further work is required to render low-alcohol non-Saccharomyces yeasts more robust, and to ameliorate their potential negative sensory impacts.