Phenols in train-ing?

I recently received a fascinating email that occasioned my learning a few new and interesting things about phenols.

To excerpt from the intial email:

“…(the taster) commented on the consistent style of the wines as having big but very soft tannins. I told him (as I have told so many others but kind of tongue in cheek) that the softness of the tannins was a result “train settling, ” Because our barrel room is located under a railroad overpass, the barrels are gently vibrated with each pasing train and that this gentle vibration helps create longer phenolic chains. The wines are gently “shaken not stirred” 21,000 times/year (that’s how many trains pass overhead each year. [The taster]was quite intrigued by this explanation and suggested that I should see if, in fact, there might be something to that theory. And so… I am asking you, do you think there could be any merit in thinking that the gentle vibration of our barrels some 21,000 times/year could contribute to the softness of the tannins in our wine?”

And from my response:

“…Though I really don’t know much about tannin chemistry, curiously enough, I work and am friends with two experienced grad students who research that very topic. One of these gentlemen has made wine in Argentina for ten or so years, consistently with an interest in polyphenol chemistry, and can probably rightly be considered an expert on the topic. I consulted them on your question and will do my best to summarize what they had to say. Please keep in mind that this is a very complicated question dealing with a very complicated and not fully understood topic on which active research is underway.
First, your assumption that longer phenol chains (i.e. a higher degree of phenol polymerization) are directly associated with “softer” or less-astringent tannins is not true. You are certainly not the only one to anecdotally think that longer tannins are softer tannins, but several published studies conducted over the last ten years have disproven the theory. Longer chains, appear to be more astringent, if slightly less bitter than shorter ones, with the relationship being non-linear and of relatively small magnitude. Stephane Vidal at the INRA Montpellier (an eminent French agricultural and sensory research institution) has done much of this work, along with Ann Noble’s group at UC Davis.
Second, there is no reason to think that vibrations will increase the degree of polymerization of your wine’s phenols. Sorry! To the best of our (my friends’ and my) collective knowledge, there is no data bearing on this question. From simple chemistry and common sense, however, vibration should increase polymerization only if addition of mechanical energy was important in increasing the number of collisions between phenol molecules. Since phenols are already in high concentration in wine, this may not be the case. Moreover, as already mentioned, the “old winemakers’ wisdom” that increased association between phenols and of phenols and acetaldehyde is responsible for the softening of red wines with age has been debunked by combining chemical and sensory analysis.
I’ll admit that the first reaction of my friends to your train-vibration hypothesis was “typical winemaker B-S, looking for a way to distinguish his wines.” Depending on your point of view, I tend to be either a bit more gullible or a bit more creative. Regardless, a bit of musing led us to think of one potential way of relating astringency to your locamotive proximity. As I’m sure you know, yeast lees can have a softening effect on wine. Yeast autolysis, a particular kind of cell death, causes release of polysaccharides and, in particular, mannoproteins into the surrounding wine. Stirring up the lees encourages cell-wall breakdown and the release of these compounds. Yeast-derived polysaccharides associate with phenols (the particulars depend very specifically on the types of polysaccharides involved, but this is a fair generalization) and stabilize the folded molecules, hindering denaturation and the exposure of hydrophobic regions that aggregate to form large and insoluble polymers. Thus, yeast autolysis reduces phenol precipitation. Since the perception of astringency occurs when tannins precipitate, yeast autolysis generally reduces astringency. Returning to your train vibrations, it is possible that the vibrations gently but frequently agitate the lees in your barrels, stirring up dead yeast components, and keeping more of your phenols in solution. This is purely hypothetical, but it’s a thought.
Finally, and perhaps needless to say, all of this depends a great deal on the varieties with which you are concerned, your particular vinification practices, and exactly what Paul (and you, and other tasters) mean by “big but soft tannins.” I can see from the Barrister website that you deal with a lot of varieties and your email implies that Paul saw this softness of tannins throughout the reds which, naturally, suggests association with your winemaking techniques and style. With respect to the yeast-stirring hypothesis, how often you rack and the strain of yeast you use and such will be relevant. “

Now, if I could only induce this winemaking gent — or another with access to a busy bit of railway — to store identical batches of wine under “trained” and “untrained” conditions and perform liquid chromatography analyses for phenols and sensory comparisons between the two conditions. Not precisely a widely-applicable study, but unquestionably interesting. And who knows? If the good-vibrations wines performed significantly better than their restive counterparts, there might be a market niche for gently vibrating wine aging platforms!