A wine is technically considered dry when the sugar level falls below 0.5%. A dry wine that also has a total acidity level above 10 g/L will likely be harsh and astringent to the consumer. Granted, there are exceptions to this generalized idea of wine chemistry. The perception of dryness can vary based on other aspects of the wine, such as acidity, dry extract, and aroma. A low tannin red wine can tolerate higher acid levels than a tannic red, and a white wine high in dry extract may tolerate higher acidity than one with less body. The level of “dryness” will also play factor in the perception of harshness in the wine, as an increase in sweetness can help to balance acidity (think lemonade).
University of Minnesota-developed wine grapes are rarely harvested with a TA under 10 g/L. It is not uncommon to see total acidity at harvest of 15-18 g/L in Frontenac. Even the newest cultivar, Marquette, sees total acidity ranging from 0.9-1.3% – numbers almost unheard of for red V. vinifera cultivars – especially when Brix at harvest on these same grapes falls within normal to high levels for dry wine production (24-27). Thus, if a winemaker desires to produce a dry wine using University of Minnesota grape cultivars, he or she must always be concerned with mitigating the harshness of these high acids.
Wine Grape Acids and their Reduction. There are three general methods one can use to lower high acidity in wine: physical methods (blending and amelioration), chemical methods (bicarbonates), and biological methods (yeast and bacteria). A winemaker may also choose to produce a sweet wine to counter-balance the acidity. This may actually be the best approach for high-vigor vineyards, vineyards on cool sites, or in short growing seasons. Nonetheless, dry table wines are desirable for many winemakers and consumers. For the acid levels seen in Minnesota vineyards, the best approach is most likely a combination of all three methods if the wine is going to end up dry.
Two acids make up over 90% of the acids found in grapes: tartaric acid and malic acid. From a wine stability standpoint, tartaric acid is extremely important. It helps a wine to maintain a low pH, thus inhibiting microbial spoilage. Chemical methods of reducing acidity all act mainly on the reducing the amount of tartaric acid in the wine, while biological methods all reduce the amount of malic acid in your wine. Those familiar with consuming green apples such as ‘Granny Smith’ are familiar with the sour flavor coming from the malic acid. Thus, a wine that contains high quantities of this acid can have a sour character. All living cells consume malic acid (malate) through their respiration cycle either by breaking down another molecule (usually a carbohydrate such as sugar), or by consuming malic acid directly from their environment
Biological Deacidification. The most common method of biological deacidification is the addition of lactic acid bacteria to the wine. The bacteria will directly consume malic acid and convert it to lactic acid, an acid with a softer less sour taste, in a process known as malolactic conversion or malolactic fermentation (MLF). Though not a true fermentation because no sugar is involved, carbon dioxide is a byproduct of the conversion, so the wine looks as though it is fermenting. Nearly all red wines around the world undergo MLF and some white wines also benefit from acid reduction of this practice. Traditionally, wines were left without sulfur protection following alcoholic fermentation to undergo MLF naturally – a process that could take several months. The risks involved with leaving the wine unprotected by sulfur dioxide (spoilage by other organisms and oxidation), have pushed many wineries to use a malolactic bacteria starter culture.
Nonetheless, yeast also have the capability to consume malate during cellular respiration and convert it to ethanol. It has long been known that certain yeast, especially several non-saccharomyces yeasts (Schizosaccharomyces pombe, Hanseniaspora occidentalis, Issatchenkia orientalis) are especially efficient at consuming malate. Because these yeasts have poor alcohol tolerance, they must always be used in conjunction with Saccharomyces species in order to complete fermentation in wine. While S. pombe has been available commercially for some time for use wine production, the development of other non-Saccharomyces yeasts for commercial use is a hot topic at the moment. We will likely see more of these yeasts available in an active-dry form to use in sequential yeast inoculations for wine.
Until then, we decided this winter to look at some of the commercially available yeast strains that have some reported ability to reduce malic acid, and trialed them with University of Minnesota cultivars. After consulting with several enological product suppliers, we came up with a list of several yeast with reported malate-reducing capabilities: Lalvin C (Lallemand), Exotics (Anchor), Lalvin ICV Opale (Lallemand), and Uvaferm VRB (Lallemand). We also trialed a non-Saccharomyces yeast that Lallemand has made available in an active dry form for sequential inoculations: Torulaspora Delbrueckii (sold commercially as Level 2TD). Although its malate-consumption hadn’t been verified, a technician at Lallemand had recommended it because they had observed some softening of the acidity in wines that had been fermented using it.
Yeast deacidification trial. We did a small trial with these yeasts in which we used juice from the 2012 vintage that had been frozen. For each MN cultivar, we trialled 3 different yeasts, and used a fourth yeast as a control. One lot of juice was divided into 20 micro-vinification lots of 500 mL each. Thus each yeast was replicated in 5 fermentation lots. For this initial trial, we were concerned with monitoring mainly the chemistry change using each yeast. For white wines we used Lalvin DV10 as control, and for red wines we used ICV GRE (Lallemand) as a control yeast. Both are considered reliable fermenters with no reported malate degradation. The unusually hot weather in 2012 caused initial brix levels to be extremely elevated, so initial malate numbers reflect juice that had been diluted to bring the sugar concentration down to 25° Brix.
With Frontenac Gris, we started with an ameliorated juice that had a total acidity of 9.92 g/L, a pH of 3.00, and 5.1 g/L of malic acid. Although all the added yeasts showed some reduction from the initial malate levels in the juice, the acid reduction seen in the Lalvin C, Exotics, and the combination of Torulaspora delbrueckii with Exotics all were significantly lower than the control yeast (p <0.05). We used Lalvin C in a larger lot following this trial in order to evaluate the sensory impacts of this yeast. It’s worth noting that in all 10 micro-vinifications in which Exotics was used, the wines exhibited some stuck fermentations. Thus, some care may be needed when using this yeast in order to complete fermentation in low pH juices.
The La Crescent juice that we divided up for the micro-vinification trials was ameliorated to 25 Brix, which left the starting malate levels at 5.3 g/L. The decrease in malic acid during fermentation was less pronounced than what we saw with the Frontenac Gris fermentation. In fact, only the vinification lots in which Exotics was used showed a statistically significant drop in malic acid (p< 0.05). ICV Opale is advertised to lower malate levels by 0.1 to 0.4 g/L. Our trials show that it exceeded this levels in high malate musts, however, this decrease was not significantly lower than our control yeast which has no reported malate reducing properties. We also saw stuck fermentations when using Exotics in the fermentation lots with La Crescent.
Our Frontenac was pressed and fermented as a rosé. Again, it was necessary to ameliorate to reduce the high sugars that we achieved in 2012, however, the initial malate concentration of the juice was still relatively high at 4.6 g/L. Interestingly enough, all yeast used for this trial caused a decrease in the final malic acid concentration of the wine. All observed differences in malate reduction were statistically significant (p<0.05), except for the two lots that were fermented with Lalvin C. There is essentially no difference between the observed malate reduction when using Lalvin C in conjunction with T. delbrueckii yeast. This (along with the other results seen when using T. delbrueckii) suggests that any impact on the perception of acidity due to this yeast is likely not related to malate degradation. All the Frontenac fermentations finished dry with no stuck or sluggish characters.
Marquette was also pressed immediately and fermented as a rosé. The ameliorated juice had an initial malic acid concentration of 4.1 g/L. Exotics and VRB showed identical malate reduction capabilities, and even though the difference between these two yeasts and the control (ICV GRE) was only slight, the difference is statistically significant (p=0.046). Nonetheless, this difference is probably negligible in the sensory profile of the wine. Once again, Lalvin C proved to have the greatest potential for malate reduction, with a 1.10 g decrease in malic acid from the juice.
It is important to keep in mind that there are many different tools available to a winemaker to manage high acidity in their wines. The selection of yeasts that we looked at here are only a small example of what is available on the market. It is important to talk with technicians who supply your winery in order to get a better idea of what products might help with managing your acidity. This coming vintage (2013) will likely promise to be a challenging one, as we have yet to see bud-break in Minnesota this first week of May. Unless we see a warm summer and an extended warm fall, growers and winemakers need to start thinking now about how they might manage the vines and prepare for a high acid harvest.