In winemaking, the term "alcoholic fermentation" refers to the transformation of sugars into alcohol. This is a biochemical process affected by yeast. The yeast species which gives the most desirable results is saccharomyces cerevisiae.
Natural yeast flora are always present in vineyards. These in turn are transported to the winery with the grapes. Over 147 species of yeast have been identified as occurring naturally on grapes. However, the methods we use at Château Carsin mean greater control over which yeast actually conducts the fermentation. The early addition of sulphur dioxide to the grapes, together with maintaining low temperatures, decreases the population of native yeast to levels where they would be unlikely to be able to carry out fermentation. Then we can simply choose which type of yeast we want to conduct the fermentation. We inoculate the juice with a pure strain.
There are many advantages to using selected yeast cultures rather than allowing fermentation to be carried out by native (we call them "wild"!) strains. Fermentation commences rapidly and we get a predictable rate of sugar-to-alcohol conversion, and a fermentation that runs to completion.
Besides the main products formed in this complex biochemical reaction, many secondary ones are also produced. Just what, and in what proportions is very dependent on the strain of yeast used. Undesirable by-products include volatile acidity and hydrogen sulphide, or "rotten-egg" gas. Usually their presence, in detectable levels, means that the yeast was not well "looked after" during fermentation. Desirable by-products, on the other hand, are esters and different types of alcohol. Many of these by-products are highly volatile compounds and therefore their presence, even in minute amounts, can have a noticeable effect on aroma and flavour, or can just add a little nuance to a wine.
In 1993 we started looking at how different yeast strains affect the characteristics of the final white wines. Our aim was to see the practical effects of the processes mentioned above. From the outset we were a little sceptical about the importance of the specific yeast strain, and as to whether the secondary compounds formed would be lasting. But our scepticism proved unfounded.
In one case, the same juice was inoculated with 7 different yeast strains. To make the situation even more complex, we carried out fermentation using the same yeast both in new and in one year old oak. The yeast should, according to the manufacturers, "bring out" certain grape variety characteristics. Others were to produce large amounts of ester by-products, or simply be reliable and not produce high levels of undesirable by-products. Some were supposed to be especially good as low-temperature fermenters, etc.
The results were quite amazing. Each yeast certainly has its own attributes. We can generalize from the results, but only where the grape variety Semillon is concerned. When young, Semillon is relatively neutral, but with age it develops toasty and honey characters. A yeast strain called Prise de Mousse, from Champagne, gave intense grapefruit aromas, but on the palate made the wine quite austere. In new oak it allowed the oak to dominate the taste of the wine. Another yeast strain, VL1, from Bordeaux, gave a nice citrus fruit lift to the nose. On the palate the wine is much rounder and fuller.
Another, called VL3, gave a slightly herbal character (some would call it a "cat's piss" character), to the nose and a nice roundness on the palate. Much research is being carried out on yeast strains to see if different ones have different enzymes that release' the flavour compounds of different grapes varieties.
These practical results have been very useful. They help us determine the type of yeast strain with which to ferment our juices as a function of the wine we are making. By this we mean that if a wine is destined to be sold not very long after it is made - a primeur for example - then by using a yeast strain that is more likely to give fresh, fruity aromas that enhance the character of the fruit and a nice weight and softness to the taste, we are making a wine that is easier to market.
Nowadays for a wine that is more serious - l'Etiqette Grise - we use yeast that have a greater effect on the taste of the wine. They help to give it depth and length, and to harmonize the fruit with the oak.
In 1994 we repeated the trials, using different yeast to expand upon our "yeast repertoir". Trials were set up to look at the interaction between oak and yeast. From this trial we didn't discover much about the oak but made a new yeast discovery. Differences between vintages also play a role in the yeast we use for conducting our fermentations. If the wines have higher amounts of acid, then, instead of adjusting the level, which in our opinion takes out other desirable compounds in addition to the acid, we select a yeast that will give richness on the palate and so make the wine more balanced. There is even one strain, which manufacturers claim decreases acidity - it works!
Trials during the 1995 vintage were also carried out. We concentrated on the classic yeast of Bordeaux as well as some old favourites. We even gave some of our "wild yeast" a go! We also looked at the effects grape solids have on the wine.
That is, the juice was not settled and brilliantly clear before the start of fermentation. The results are that "grape solids" do effect the flavours of a wine. They tend to block aromas on the nose of wine, but on the other hand, they do give an extra dimension to the palate - l'Etiqette Grise stuff! We found that the combination of grape solids and native yeast meant that the fermentation was more easily able to come to completion. Our clear and native yeast barriques were left with small amounts of residual sugars because the alcoholic fermentation did not come to completion. This being due to the low nutrient supply combined with low yeast numbers.
In 1996 we have higher levels of acidity in the grapes along with richer fruit flavours - a very good year. Thus we did not use the Prise de Mousse yeast at all instead we used yeast that would bring out the fruit characters and add a little weight to the palate. A greater part of the Cuvée Prestige was fermented with grape solids to increase the depth of flavour on the palate - not all of it though as we want to keep the freshness and fruitiness on the nose.
In 1997 the Sauvignon Blanc was not as aromatic as we would have liked and therefore the use of aromatic yeasts such as VL3 and Prise de Mousse was necessary. The Semillon was at it's best, rich and ripe, therefore whatever yeast we used, it was good.
All of this experimental work is carried out on some of our best-quality juices and therefore our top wines. Perhaps the most important factor we learnt was that by using different strains of yeast to ferment our individual juices we are able, not only to better respect the qualities of each juice, but we are able to give further complexity to our wines. This is important when we are working with a limited number of grape varieties.
Whatever we can achieve by working with yeast, it is still true that a top quality wine can only come from top quality fruit.
And For Red Wines...
The discussion so far has been about fermentation of white wine. The yeast used for the fermentation of red wines has broadly the same significance. It is difficult to conduct trials, because of the greater number of variables involved. Different vineyard sites are perhaps the most important factor. It is not quite the same as a homogenous tank of white juice! The results are more difficult to evaluate because of the already very complex nature of red wines. So at this stage our experimentation has not been more than what is practical based on a "gut-feeling" about the results.
