Tannins from grape seeds have few fans among winemakers, but much of that bad reputation is based on misunderstanding.
Among the contributions of seed tannin are shorter, less astringent polymer chains--a likely boost to wine longevity and a component in building the mid-palate of wines.
Any comparison of grape seed and skin tannins is overshadowed by the changes both go through once they get into wine.
Grape seed tannins get no respect. The way some winemakers talk about them, you’d think they were just another form of MOG, only tucked inside the grapes and impervious to any sorting table. Maybe Thompson Seedless—a vinifera grape, after all—should get another look. We don’t need those low-life seeds to propagate vines anyway, so can’t we just get rid of them altogether?
Alas, a grape seed-free world would be a sad one indeed, and seed tannin-free red wines would be mere shadows of the real thing. The poor little pips get blamed for a host of things they don’t cause, and bring some very useful things to the table—enough so that some of the winemakers most intent on removing seeds end up putting seed tannin back into their wines in the form of powdered additives.
Let’s give some credit where credit is due.
Seeds and skins
There are several well-known structural differences between seed tannins and skin tannins in grapes, a grab bag of advantages, disadvantages and features no one is quite sure how to evaluate.
Skin tannins include one monomer not found in seed tannin: epigallocatechin. Epigallocatechins are distinguished by having three hydroxyl groups hanging off them—in contrast to the two hydroxyls on the other basic monomeric tannin building blocks, catechin and epicathchin. Does this make any difference once we get to wine? Stay tuned.
As Jim Harbertson of Washington State University loves to remind people, seed tannin chains are quite a bit shorter than skin tannin polymers; the ratio is about one to three—10 units on average for seed tannin to around 30 for skin. Therefore, according to the best sensory studies, seed tannins should be, chain for chain, less astringent than their skin-wrapped cousins. Seed tannin often comes with a chunk of gallic acid stuck to the end of the chain, and for some reason, that can make a seed tannin chain of a particular length rougher than a same-sized skin chain—even though gallic acid itself is not astringent. But the chances are good that once that seed tannin gets into the chemical kickboxing of a wine solution, gallic acid will get knocked off, lessening the seed tannin astringency.
In other words, if you have a highly tannic wine, chances are the seeds weren’t responsible for it.
The shortness of seed tannin also has a downside: bitterness. For some reason bitterness, at least in tannin, is only perceived in the presence of large amounts of low-molecular-weight compounds, so the shorter seed chains are more likely to come across as bitter than the longer, higher-weight skin chains. Jim Kennedy at California State University, Fresno, thinks that our bitterness taste receptors may not be big enough to deal with the long chains, so they get a pass and only produce astringency. Harbertson thinks the big ones may get caught up in saliva proteins before they ever get to the taste buds. In any case, seed tannin is one potential source of bitterness—depending, of course, on what happens to those little chains once they get into wine.
Finally, seed tannin is much more plentiful than skin tannin. Harbertson estimates a range of 3.5 to 5 milligrams (mg) of seed tannin and 0.5 to 0.9mg of skin tannin per berry. Kennedy observes that viticultural practices can have a substantial effect on accumulations of both types of tannins: low-vigor vines tend to produce grapes with more skin tannin and more astringency; high-vigor vines produce grapes with more seed tannin and more bitterness.
Then the grapes get picked and the real fun starts.
Grape tannin and wine tannin
Phenolics researchers all agree that we know more about tannins in grapes than tannins in wine—which are, after all, the ones we actually taste. Once the grapes hit the crusher, tannins and most everything else segue into an alternate reality.
Even though seed tannins are more plentiful, water-soluble skin tannins come out earlier and easier than seed tannins, which stubbornly insist on ethanol as a solvent. Except possibly in cases of extreme extended maceration, during which seed tannins continue to be extracted, skin tannins constitute the great majority of inputs. In this light, Jim Kennedy says that the issue for winemakers isn’t so much controlling seed tannin as managing skin tannin.
But far more than the seed-to-skin ratio changes. Australian Wine Research Institute (AWRI) phenolics researcher Paul Smith advises, “Forget everything you know about structure-function relationships for grape tannins. Forget about short or long, seed or skin, galloylated or non-galloylated. Of course all tannin comes from grapes, but once it’s extracted most of it rearranges, breaks apart, recombines, binds more polyphenols, anthocyanins, acetaldehyde, etc., and what is left is ‘wine’ tannin, of which a small amount of material, chemically, still looks like grape tannin. We regularly see that wine tannin can be less than 20% ‘grape’ tannin.”
Sounds almost Biblical, doesn’t it? “The long shall be short, and the short long, and the skins will lie down with the seeds.”
At this point in tannin studies, the 80% of wine tannins that aren’t simple grape extracts are not well understood. “Characterizing the rest of the material is where the frontier is,” says Smith. And the phenolic frontiersfolk have a number of interesting speculations.
Could be that those gallic acid hangers-on get stripped off in the acid bath of wine. Could be that short seed tannins hook up and become more skin-like, and skin tannins break down into seed-like chains. Could be that the greater seed and skin tannin extracted during extended maceration gets masked by the greater amount of polysaccharides extracted at the same time. Could be that, since big molecules like to play with other big molecules, the longer skin tannin chains are more likely to hook up with those polysaccharides, canceling out their previous astringency. Could most likely be all of the above—and more.
Bruce Zoecklein of Virginia Tech points out that astringency can trump bitterness in perception of a wine’s qualities, which can be a good thing. Except in badly unbalanced wines, any bitterness contributed by seed tannins is likely to be covered up, sooner or later, by the astringency from skin tannins. Many winemakers have learned that early bitterness fades with time; less fortunate winemakers have learned that overly aggressive fining for tannin reduction can bring the element of bitterness back into the spotlight.
And now it’s time to return to our friend the epigallocatechin, the one monomeric tannin building block found only in skin tannin. Paul Smith notes that epigallocatechin is the subunit most susceptible to oxidation. Jim Kennedy, who on the whole is not that fond of seed tannin, suggests that its comparative resistance to oxidative reactions (since it has no epigallocatechins) may make it valuable for wine’s longevity, slowing down the pace of oxygen-driven deterioration, Jim Harbertson says we don’t really know the effects of epigallocatechin.
I followed up my initial email exchange with Paul Smith by asking whether he thought the epigallocatechins’ propensity for oxidation was overall a good thing or a bad thing, and what he sent back is a succinct expression of the state of tannin knowledge. Here is his response in its entirety:
“We don’t know if it’s good or bad—if oxidized tannins are soft and mellow, then I guess it’s good. If oxidized tannins are grippy and coarse, I guess it’s bad!” Always diplomatic, those Aussies.
Finally, one important part of wine tannin is the tannin that never gets into the wine at all. Research by Doug Adams at UC Davis, several investigators in France and Keren Bindon at the AWRI has illuminated the ways in which extracted tannins, especially the longer chains, can get trapped back into the skins, a kind of self-fining process. Jim Kennedy says some wines never get half their potential tannins. As an example he uses grapes from California’s toasty Central Valley, where skins toughen up and trap tannins as a form of self-defense against the heat.
In this complex and somewhat fuzzy picture, it’s hard to see seed tannins as the prime bogeyman.
What if we made the seeds go away?
Perhaps the strongest case for the value of seed tannin comes from knowing what happens when they are utterly banished. It’s not good.
Zoecklein and other researchers did some studies a while back in which they removed as many seeds as practical in the early stages of red fermentations, before there was much ethanol to pull out the seed tannin. He describes the resulting wines as having a “donut hole” where the mid-palate should be.
More recently, the same issue has arisen with the hot-button technology of Flash Détente. The rapid heating and vacuum-cooling of grapes with Flash Détente’s version of thermovinification performs almost instant extraction of skin tannins and pigments using only heat and water (juice). Without any ethanol present, seeds remain intact and their tannins under wraps. If winemaking proceeds only with the liquid results, discarding the seeds and shards of skins for convenience, the result is once again a problematic mid-palate.
A final testament to the usefulness of seed tannin is the flourishing market in winemaking additives based on them. AEB offers several products that contain seed tannin, including one based entirely on Malbec seeds—the reason, according to general manager Marco Bertaccini, being that AEB’s supply of Malbec seeds are particularly ripe. He asserts that riper seeds contain more developed (longer-chain) tannins that add astringency, not bitterness, contributing what he describes as an “Old World edge.” He observes that a number of wineries that use AEB’s products do so because their strenuous seed removal has produced wines that need a boost put back in.
Mark Simpson of Artisan Food and Beverage in Vancouver, British Columbia, distributes GrapEX, a seed additive produced by Tarac Technologies in South Australia. Simpson says that seed and skin tannin are both needed as part of a complete wine package, with seed tannin particularly important for the mid-palate. He also thinks that seed tannin can help with co-pigmentation, a process in which pigment is held in solution in stacks of compounds, none of which has any color of its own. And like Bertaccini, he has customers who are too vigilant about removing seeds and their tannins and need help building them back up.
Russ Robbins of Laffort, an international supplier of tannin products worldwide, says some of that company’s tannin additive range includes a portion of seed tannin. He also cautions that the performance of any tannin product in any particular wine is highly variable. “You can take 25 parts per million, 50, 100 of six different tannins, put them into wine and never predict what will do what. Problems in wines aren’t necessarily due to too much tannin, but too many of the wrong kind, all of which hit the mouth in different ways.”
If there is a consensus in this tangle of tannins, it is so obvious as to be almost boring: Quality wine needs balance. Some astringency is a good idea, so are some seed tannins. Too much or too little of anything can get you in trouble.
You can read and reread this column as many times as you want and still not have a road map to great wine structure. But hopefully, you can get past blaming everything on belligerent seed tannins. Let the healing begin.
Tim Patterson is the author of the newly released “Home Winemaking for Dummies.” He writes about wine and makes his own in Berkeley, Calif. Years of experience as a journalist, combined with a contrarian streak, make him interested in getting to the bottom of wine stories, casting a critical eye on conventional wisdom in the process.