“What Yeast Should I Use?”

The title of this post is one of the most common questions asked by winemakers working with cold-hardy grape cultivars. It is a simple question, but one that doesn’t have an easy answer. I have written on this topic in the past, so let me just throw out something that you probably haven’t heard yet: your yeast choice probably isn’t going to make or break your finished wine. There. I said it. I diminished the importance of yeast choice. To be fair, yeast selection does have an impact on the characteristics of your wine. Poor-quality fruit can be enhanced by choosing the correct yeast, and high-quality fruit can lose some of its potential by choosing the “wrong” yeast. The argument being made here is that your yeast choice isn’t going to make the difference between a wine that is worthy of a gold-medal, and one that is worthy of being poured down the drain.

Frontenac Gris in Wine Preference Study

When yeast choice REALLY matters, it’s when the environment in which the yeast will live (the  juice, and eventually fermenting wine) is inhospitable. Very acidic (pH < 3.2) or very high sugar juice are stressful to yeast, as are very hot or very cold temperatures. Certain strains of yeast are more tolerant than others of these harsh conditions. If for example, you harvest Marquette at 25.5 °Brix and hope to make a dry wine, you’d better make sure that the yeast is tolerant to alcohol levels greater than 15%. Making a late harvest or ice wine? You need a yeast with high osmo-tolerence to handle the high sugar environment.  If you plan on using bacteria to convert the malic acid to lactic acid, you’d better make sure that the yeast is compatible with Malolactic Fermentation (MLF). Do you have a cooling system in your winery? If not, then you probably should pick yeast that can tolerate hotter temperatures. If you plan on cold-fermenting the wine (to guard fruity aromas), the yeast should be tolerant of cold temperatures. All of these planning questions help to eliminate the outright poor yeast choices, then you can get into some of the nitty-gritty details.

Sensory effect of yeast choice. After eliminating yeast strains that won’t work with your juice chemistry and fermentation goals, the main concern is the sensory effect of the yeast strain. In general, cultivated yeast strains will produce low amounts of off-aromas (H2S and VA) when given sufficient nutrients. Some yeast can affect the mouthfeel of a wine by producing higher amounts of glycerol. There are yeast strains that produce high amounts of tutti-frutti ester aromas – great for young wines, but for high-end wines that are going to age a year or more before release, there isn’t much of a point in using these strains. Esters are extremely volatile, and are the first aromas to disappear – sometimes within a few hours of opening the bottle! Other yeasts will enhance the aroma by releasing some of the aroma precursors found in the grapes at harvest. This is all well-and-good, but in the end the yeast can’t do much unless the precursors for these aromas are in the grapes themselves. This is where the big question lies with cold-hardy grapes. For the most part, we know very little about the nature of their inherent aromas. We know that La Crescent is related to Muscat, and has some of the same floral and perfume aromas that are found in all Muscats. We know that it does contain high quantities of monoterpenes, the class of aroma compounds that have these flowery characteristics. However, we also know that Marquette contains significant quantities of monoterpenes, although it is rare to see floral descriptors used when tasting Marquette wines.  Frontenac contains  methoxypyrazines when unripe (similar to the green pepper aroma in Cabernet Sauvignon) and minty aromas (methyl salicylate and menthol).[1] As we learn more about the impact aromas of these grape cultivars, it may affect our decisions for yeast selection. You can read about why these particular yeast strains were chosen for this trial in a previous post.

Yeast trial with cold-hardy grapes. Last year, we decided to ferment the four University of Minnesota grape cultivars with various commercial yeast strains. This was a trial that was sponsored by the Northern Grapes Project, and was replicated at Cornell University with fruit from Vermont and New York. Over the past few weeks, I asked a group of 27 people who all have experience tasting regional wine to participate in a wine sensory panel. The panel consisted of 16 men and 11 women, whose ages ranged from 26 to 74 with a median age of 50. They were served three wines from each of the four grape varieties and asked to rank them from their most preferred to their least preferred. The only difference in the three wines was the type of yeast that was used for fermentation, which is highlighted in the chart below.

The panelists were also asked to write comments on each of the wines. Not surprisingly, many of the tasters noted differences between the wines. On several occasions, it was noted that one of the wines was “far superior” to the two others in the flight, with notes such as “most complex” and “most interesting” written in the comments section. I even had one panelist who stated afterwards (when he found out what the trial had entailed) how he is always surprised by how much yeast choice can “make or break” a wine. In the end, we were testing whether there was a difference in preference for these different wines in order to give recommendations to winemakers. So which of the three yeasts for each grape cultivar were preferred by our tasting panel?

Drum roll please….

For each wine flight, the judges scored the wines in order of preference, with 1=most preferred, and 3=least preferred in the flight. We tallied the total points for each wine and the results are in the charts below. A lower score indicates a higher overall preference (more #1 ranks) by the judges. Statistical analysis was done using the Basker Critical Values for Rank Sum.

The small letter next to the sum indicated whether the difference seen is statistically significant (p < 0.05). If there is the same letter next to the sum, then there is no statistical difference in the observed count. As you can see, for every single yeast trial, no clear difference in preference was shown for one yeast over another yeast in this particular trial.  We may be able to say that for La Crescent, there is a trend towards a preference for yeasts that release monoterpenes (both Cross Evolution® and Elixir enhance floral characters in aromatic whites), but we would need to recruit a larger panel to see if this holds true.  However, at this point, there isn’t a clear preference for those yeasts over a more neutral yeast (DV10).

We chose the yeasts for this trial based on their ability to work well within the chemistry limitations of our varieties.  The subtle differences in these wines that may have been observed by individual panelists didn’t translate into a difference in preference for one wine over another for the group as a whole. This is just to highlight why yeast choice probably isn’t as critical as one might think. In the end, it’s a decision that a winemaker makes based on his or her own personal preference and wine-style goals. This is part of the art of making wine. In the  future, we hope to also do descriptive analysis of these wines, to see if these differences can be appreciated by a panel of consumers. Descriptive analysis will also help guide winemakers towards understanding how yeast choice may affect the sensory characters of their wine.

*For Frontenac we could only used the scores from 26 panelists due to an error on one score card