What do you do when your Brix creeps up to 28¡ã? Do you water down; bleed the juice off first and then water down; or bleed off, water down and extend the contact time all combined?
In the article, ¡°Chemical and sensory effects of saign¨¦e, water addition and extended maceration on high Brix must,¡± James Harbertson and his team took a very ripe (28¡ã Brix) Merlot from Columbia River, Washington and tested all of the above techniques. Then they examined how each impacted the extraction of anthocyanins and tannin, and the sensory profile of the wines. The results were published in the American Journal of Enology and Viticulture, 60(4): 450-460, 2009.
As we know, tannin extraction in small-scale fermentations is not reperesentative of real-world commercial fermentations (but rather as much as 36 percent lower), so the authors made the wise decision of conducting their experiment at the commercial scale (5,000-gallon tanks). I also applaud them for conducting two fermentations per treatment; so often we get away without satisfying this basic research principle. It is likely that the trial took over the commercial winery since there were a total of five treatments as follows:
• Control: Must watered down from 28 to 24¡ã Brix.
• High ethanol: Must watered down from 28 to 26.8¡ã Brix.
• Low saign¨¦e: 16 percent of juice bled off (or saign¨¦e) and replaced by the same volume of water to yield 24¡ã Brix.
• High saign¨¦e: 32 percent of juice bled off and replaced with enough water to yield approximately 24¡ã Brix.
• Low saign¨¦e plus EM: Low saign¨¦e treatment above paired to 20-day extended maceration (instead of seven-day).
It is important that we familiarize ourselves with these treatments before continuing the rest of the trial.
All treatments received three pump-overs per day and were pressed after seven days of contact time, at approximately 6¡ã Brix (except for the low saign¨¦e plus EM, which was pressed after 20 days). Let¡¯s see how the various treatments affected the wine anthocyanins, stable color, tannin and sensory properties, in that order. Because levels of the above components change throughout fermentation, we will focus on the latest measurements that the authors conducted¡ªsix months after press¡ªwhich we will call ¡°final wine.¡±
Anthocyanins are monomeric pigments found in the skins of red grapes and in the pulp of some varieties called ¡°tinteurier.¡± They play an important role in the color of a young red wine, along with polymeric pigments.
During fermentation, anthocyanins increased gradually up to day 7 after which they declined. As a result, they were at their peak around press time, even if this increase was transient for most treatments. However, high saign¨¦e, followed by low saign¨¦e, were the treatments that maintained the highest levels in the final wine. On the opposite extreme, low saign¨¦e plus EM had the lowest anthocyanin concentration after six months. This was due to the fact that this treatment was pressed at 20 days¡ªwell after anthocyanins had peaked and started to decline. The authors note how anthocyanin levels of the high ethanol treatment were not any lower than those of the control. This provides evidence that anthocyanin extraction (solubility) is not compromised by high alcohol concentration. The authors did not look at the effects of the various treatments on visual color.
Stable Color Extraction
Polymeric pigments, also called pigmented polymers, are a mixed bag of pigmented material that result, as wine ages, from the combination of anthocyanins with other wine compounds, such as tannins. Their name was meant to emphasize the fact that, unlike monomeric pigments, which can be easily bleached by adding bisulfite, their polymeric counterparts resisted bisulfate bleaching. However, today we know that not all polymeric pigments resist bisulfite bleaching, not all polymeric pigments are polymeric, and¡ªget this¡ªnot all polymeric pigments are even pigments (some thrown in this category happen to be colorless). What remains true is that this complex mix of mostly colored species is the primary source of stable color in old red wines. To further characterize polymeric pigments based on their size, researchers divided them into those that are able to precipitate protein, which are considered large (Large Polymeric Pigments or LLP) and those unable to do so, which are considered small (Small Polymeric Pigments or SPP). The sum of the two is referred to as TPP or Total Polymeric Pigments.
The amount of polymeric pigments measured in the trial wines (all TPP, SPP and LPP) increased with time, being higher in the final wine than immediately after pressing. The treatments that had the highest polymeric pigments in the final wine were low saign¨¦e plus EM and high saign¨¦e. These treatments also had higher tannin. Since polymeric pigments are formed through the reaction of anthocyanins with tannins, the authors believe that the higher tannin level of the above treatments may have enhanced the formation of polymeric pigments. In addition, the extra maceration time in the low saign¨¦e plus EM (20 days instead of seven) probably contributed to the elevated concentration of polymeric pigments in this treatment.
Tannin Extraction from Skins and Seeds
Tannins are phenolic compounds present in the berry skins and seeds and are the main source of wine astringency. The proportion of tannins in a berry is approximately 20 percent skin tannin and 80 percent seed tannin. Seed tannins are traditionally considered more desirable than skin tannins even though evidence relating tissue origin¡ªtherefore chemical structure¡ªwith the degree of bitterness or astringency is still unsatisfactory.
Tannin extraction in the trial wines increased during fermentation and then leveled off. It is, therefore, no surprise that the authors found one of the treatments to have more than twice the amount of tannin than the rest: low saign¨¦e plus EM, pressed after 20 days of skin contact, well passed the time when tannins peaked. This confirms previous research showing that extended maceration increases tannin concentration. The high saign¨¦e had the next highest level of total tannin.
The authors did some ¡°archeological¡± work next. They went back to the pomace and recovered skins and seeds one by one to be able to ¡°rebuild¡± what had been the solid portion of a whole berry. By comparing the amount of skin tannins in the pomace with those of the berries at harvest they were able to estimate the amount that had been extracted during fermentation. Similarly, by comparing the amount of seed tannins in the pomace with those of the intact berries, they calculated the overall amount of seed tannin extracted.
They found that the control wine and the low saign¨¦e wine had the highest amount of skin tannin. In contrast, low saign¨¦e plus EM, followed by high ethanol and high saign¨¦e, extracted the highest amount of seed tannin. These results show that extended maceration tends to increase seed tannin extraction. In addition, they strongly suggest that increasing the ratio of solids to juice also favors seed extraction. This is in agreement with previous research that showed that extraction of tannin from skins reaches a plateau early, whereas extraction of tannin from seeds keeps increasing as contact time increases.
Where are the Missing Tannins?
Tracing wine tannins to skin and seed tannins removed from the pomace is quite ingenious. But as the authors emphasize, knowing the amount of tannins extracted into the wine is unfortunately not equivalent to knowing the amount of tannins remaining in the wine. Indeed, they observed that, for every treatment, the amount of tannin found in the wine was between 20 to 40 percent less than the amount that had been estimated to have been extracted from the fruit. In other words, the pomace-to-fresh-fruit comparison provides precious information about the amount of tannin extracted from each tissue (assuming tannins from both tissues are equally lost) but not about absolute tannin levels in any given wine.
So where did the rest of the tannins go? The fate of tannins lost during fermentation is quite intriguing and remains a mystery to this day. Some possible causes of tannin loss include: 1) tannin oxidation followed by precipitation and 2) binding of tannins to cell wall components from the skins and pulp.
There is currently a large amount of research exploring the causes of tannin loss during fermentation. The long-term goal is that, one day we will be able to measure, not the levels of tannin in the fruit but the levels of tannin likely to end up in the wine.
Did bleeding off, water addition or extended maceration affect the sensory profile of the final wine? Through duo-trio difference testing the authors found that the answer was yes. The next question was: how so? To answer this, the authors trained a panel to conduct descriptive sensory analysis of the wines. After the panel ¡°graduated¡± for agreement among its members and reproducibility, each member rated the intensity of various aroma, flavor, taste and mouthfeel descriptors. The panelists indicated the intensities on a 15 cm scale anchored at the 1 cm end with the word ¡°low¡± and at the 14 cm end with ¡°high.¡±
Regarding wine aroma, none of the treatments altered the aroma profile of the wines compared to the control. That is, the original aromas in the fruit were not affected by the concentration changes brought about by the treatments. However, fresh fruit flavor was significantly lower in the high ethanol wine. As for taste, low saign¨¦e and high saign¨¦e were judged higher in acidity (sourness) than the control, despite the fact that both had the same titratable acidity and pH. As we know, perception of acidity can be influenced by other parameters in a wine, such as astringency and fullness.
In the mouth, the authors found that low saign¨¦e plus EM was significantly less smooth and more drying than the other treatments. The next most drying treatments were high saign¨¦e and high ethanol. Finally, hotness sensation was, quite predictably, higher in the high ethanol wine than in some of the treatments. Interestingly, low saign¨¦e plus EM was rated similar in hot character to the high ethanol treatment. Once again, hotness sensation is expected to be affected by other parameters in the wine, such as astringency.
Overall, the wines could be separated from the sensory point of view into three contrasting groups:
• Low saign¨¦e, control and high saign¨¦e were all wines with more fresh fruit and a smoother mouthfeel;
• High ethanol stood out for its high perceived alcohol; and
• Low saign¨¦e plus EM stood out for its dryness.
During their commercial-scale experiment, the authors were able to reach the following conclusions:
• Saign¨¦e at the same rate as water addition (18 percent; low saign¨¦e) did not translate into higher concentrations of phenolics or greater aroma or flavor than a standard water addition (19 percent; control);
• Saign¨¦e at a higher rate (32 percent) than water addition (16 percent; high saign¨¦e) resulted in higher levels of tannins, anthocyanins and LPP than the control or the low saign¨¦e;
• Extended maceration combined with low saign¨¦e (low saign¨¦e plus EM) extracted more seed tannin, resulting in a wine less smooth and more drying than the rest of the treatments; and
• A small water addition by itself (4.5 percent; high alcohol)¡ªthat is, much less water addition than in the control (19 percent)¡ªresulted in higher ethanol, higher tannins and anthocyanins, diminished fruit flavors, and a more drying and hot mouthfeel.
So what is a winemaker to do with a high Brix must? Aside from the obvious of trying to pick earlier, the current results point to high saign¨¦e as being the best compromise of those studied (high tannin, high anthocyanin, high LPP, high fresh fruit), always subject to stylistic goals. Even though this was also the only treatment with a net volume loss (16 percent), ending up with less juice may be a small price to pay when the trade-off is a wine with improved chemistry and sensory profile. WBM