by Debra Rossouw & Florian Bauer 


A prominent goal in global wine research is the reduction of ethanol levels in wine, without the use of physical alcohol removal methods such as spinning cone columns and reverse osmosis (Howley & Young, 1992; Pickering, 2000). This research is motivated by a confluence of social, economic and environmental factors: On the one hand, many styles of wine that are popular in the global market require grapes to be fully ripe at harvest. Desirable levels of ripeness are frequently linked to high sugar concentrations in the grapes, which invariably, all other factors being equal, lead to higher final ethanol concentration in the wine. On the other hand, while achieving desired stylistic outcome, such high alcohol levels may negatively affect the taste and balance of the wine (Guth & Sies, 2002). From a health perspective, high ethanol levels add to concerns related to alcohol addiction and illnesses. Finally, many countries tax wine based on the percentage of alcohol creating a strong commercial incentive to decrease ethanol levels (Heux et al., 2006).

While pre- and post-fermentation processes have been considered in low ethanol research, a growing body of international research is targeting microbial technology in an attempt to alter fermentative ethanol production. Microbial strategies present an attractive opportunity to decrease ethanol levels while preserving the quality and aromatic integrity of the wine. One aspect of this research relates to yeast strain development through breeding or genetic engineering: The principle behind these approaches is the engineering of yeast strains through altered gene expression to modify carbon fluxes in the cell. One of the key target carbon sinks in these approaches has been glycerol, as several research groups have attempted to re-direct carbon towards glycerol in order to decrease the flow of carbon to ethanol. (Remize et al., 1999; Lopes et al., 2000). These approaches have seen some success in terms of decreasing the ethanol concentration in wine, but off-flavours, such as acetic acid and butanediol, are often produced. Rossouw et al. (2013) demonstrated that an alternative metabolite in central carbon metabolism, trehalose (a disaccharide made from two molecules of glucose), can be targeted as a carbon sink without resulting in the accumulation of undesirable redox-linked metabolites (Rossouw et al., 2013).

A second microbial strategy that has seen growing interest in the last decade involves the use of non-Saccharomyces yeasts in co-fermentation with conventional S. cerevisiae wine yeast strains. S. cerevisiae wine yeast strains have been selected for fermentation efficacy, and as such their primary fermentation kinetics (in particular ethanol yields) fall within a very narrow range. While finding S. cerevisiae strains showing lower than usual ethanol yields is thus not likely, this avenue can be pursued for non-conventional wine-associated yeasts where ethanol yields have been found to vary widely (Contreras et al., 2014; Rossouw et al., 2015). Several studies have reported lower ethanol yields when using non-Saccharomyces yeasts, however, the decreased ethanol production was often linked to high residual sugar levels in these wines (Ciani et al., 2006; Magyar & Toth, 2011). Recent studies have highlighted the diversity of the South African vineyard microbial landscape (Setati et al., 2012), providing the opportunity to access novel species and strains in the context of ethanol research.

Impact of yeast strain and environmental parameters on ethanol yields

Varying key environmental and fermentation parameters, such as must nitrogen content, pH and fermentation temperature, has been shown to significantly impact the production of key yeast metabolites such as glycerol (Rankine & Bridson, 1971; Torija et al., 2003). If glycerol production can be manipulated through environmental factors, and given that glycerol and ethanol production by yeast are often inversely correlated, ethanol production could likewise be manipulated. We therefore undertook to determine which fermentation factors (alone or in combination) could influence the production of ethanol. This systematic approach comprehensively investigated the effects of various combinations of pH, temperature and nitrogen settings on the ethanol yields produced by 15 commercial wine yeast strains. These experimental factors were selected as they can in principle be controlled by the winemaker, thus making implementation practical, cost-effective and user-friendly if successful. However, the data indicate that there are no statistically significant differences in ethanol yields between strains, or between different conditions …