The confusion at UC Davis concerning tannins



I recently attended a seminar by UC Davis professor Dr. Andrew Waterhouse who was giving updates of various research projects at the university. Are you aware of the Tannin Assay that has been recently developed? I had posed a question to him about the assay that wasn't answered to my satisfaction so I thought I would direct it to you. (Let me preface by stating that I am just a lowly MW student, so my understanding of the subject is surely more limited than that of a trained winemaker)

I had asked him if the assay, which is protein based, would quantify the total tannins or just the astringency of tannins. He said there was no distinction between the two, tannins = astringency. Based on much of what I've read about bookending tannins, micro-ox, etc. this seemed hard for me to believe.

To add to the debate, just this morning on Jancis Robinson's member forum I was reading an account that Dr. Boulton was lecturing in Spain and made the statement that there is not such a thing as “ripe or green tannin”; research shows that what we call ripe or unripe tannin is just amount of tannins.

Can you weigh in from your perspective?

Dear Adam:

Well, I certainly disagree with both of these learned men. Jeez, don't get me started.

To me the differences in tannin which I was taught by Patrick Ducournau, and regularly observe and teach are as obvious as the color of the sky. You don't need an academic argument. Just taste some wines. I'm not sure what is meant by "ripe" tannins, but the term "tannin verte" has been used for at least fifty years at the University of Bordeaux and is discussed in Peynaud's book, The Taste of Wine. It's in everyday use with any decent French oenologiste, along with "dur," "fondue," and "sec." It doesn't mean unripe -- all young wine has green tannin, which then either converts to hard via oxygenation or to dry if it is deprived of it or if too much hangtime has occurred.

There is absolutely no evidence to support these men's point of view on tannins, and Dr. Adams' results make it plain that the position is indefensible. In addition, every other winemaking university says otherwise. Chief among them is the Tannin Wheel developed by Richard Gawel at the University of Adelaide, which specifies forty different tannin descriptive terms, and was developed in collaboration with the University of Stellenbach and Prof. Michel Moutounnet at Montpelier.

Dr. Boulton, under whom I served as teaching assistant, is a chemical engineer, a damned good one, but unqualified in this area, although he did some brilliant work elucidating the phenomenon of copigmentation, which paradoxically is probably the source of what we call green tannins. If you put the question to Roger, are copigmentation colloids, which he demonstrated are 100% monomer, trisant in wine, he'd say yes. But he's never considered that this structural change in the wine might cause a difference in astringency from green to hard. It's not his area.

Dr. Waterhouse is an analytical chemist. Neither of them have sensory training nor do they do research in this area. They are both spouting the gospel according to Ann Noble, the queen of Davis astringency research. Ann did excellent work on time-intensity of astringency in the 80's, and took the position they espouse early on, seeming to want her work, which had nothing to do with distinguishing types of astrigency, to be the final word on the subject of astringency.

I once tried to teach Ann the differences in tanins, bringing examples for her to taste. She insisted on "cleansing her palate" before each taste by rinsing her mouth with some disgusting protein gunk, after which she couldn't discern anything. She makes her sensory subject do the same thing. Consequently, she has failed to reproduce the tannin types in her experiments. Two habits pervade Davis. One is that they think that failing to show a difference is the same thing as proving there isn't one. Secondly, they are not really a member of the international community, and don't trust any work that isn't done at the Wickson Hall ivory tower, feeling that the French and Australians aren't real scientists.

Waterhouse seems to be unaware that Doug Adams' work with microoxygenated wines in our collaborative study in 1999, which showed an increase in SPP (stands for "short polymeric pigment") as well as LPP in young wines. Doug (it's his assay) and his assistant Jim Harbertson were very surprised, as they had never before observed SPP increases. These are the polymerized protiens that aren't precipitated by the test protein. Since the oxygenation softens the tannins, it showed that in reality, Doug's assay tests for astringency rather than polymer, as you were asking. But I guess Andy wasn't in that loop.

So you have the answer to your question. In general, LPP is over tetramer (4 units daisy-chained together or more), but in wines where oxygenation, lees stirring or other enrobage strategies soften the tannins (without changing the total amount) SPP grows during the treatment. In young reds, the LPP grows too, because monomer is being polymerized, but once the monomer is used up after a year or so, it shrinks, indicating a softening (loss of afinity for binding to protein, be it salivary mucins or the bovine serum albumin in the assay).

Well structured wines go from hard grippy tannins to soft, enrobed tannins as they mature, and then dry out when the wine goes over the hill. At that point, astringency (and LPP) increases again, despite that the total tannin is remaining constant or even decreasing through precipitation.

I recommend you talk to Doug Adams or to Jim Harbertson about the assay. Andy and Roger aren't really qualified to discuss it. You can also read about the assay on my website, where we recommend it for grape maturity determination. If you want to know if tannins have different characteristics, taste some wines. It's perfectly obvious.

But there are none so blind as those who will not see. Frankly, I gave up on these guys a while ago, and recommend you do the same.




larry schaffer:


Interesting info on this site, as usual. And no surprise that there are some who agree with what one professor 'professes' and others who disagree . . .

Could you please elaborate on your 1999 study? Curious to hear more details if possible.

Working in Dr. Adam's lab for 2 years and conducting hundreds and hundreds and hundreds of these assays, I've got a pretty good feel for the assay, its results, and how they correlate with 'astringency' in general.

I conducted 25 tastings where 'trained' tasters tried syrahs from around the world and then rated each on a scale of 1-10 for bitterness, sourness, and astringency. We had one wine used in all tastings and this became our 'anchor' wine. The concept was to see what correlation if any there was between the objective data obtained in the Assay and subjective data culled from tasters' perceptions of astringency. Though the correlation was not perfect, it was statistically signifant enough to show that the objective astringency obtained in the assay and subjective rankings of astringency were linked - and as such, Dr. Waterhouse is not quite 'wrong' in his assesssment.

Now onto Boulton . . . I am happy to hear him discuss this topic because it is one that does not get much play - everyone simply assumes that there are such things are 'harsh', 'soft', 'velvety', 'green' tannins, etc.

While in this same lab, we asked winemakers to send samples of wines they felt were incredibly 'harsh', 'hard', 'green', or 'soft'. . . . In pretty much every case, the differences can down to numeric ones - 'harsh' and 'grippy' wines simply had higher levels of tannins than 'softer' ones . . .

Now onto micro-ox - there is no doubt that this process helps 'soften' the wine and speed up the aging process by forcing monomeric anthocyanins to combine with tannins creating both LPP and SPP. From what I saw and others did as well, the ones more important for softening were and are LPP - SPP is really not that important in this equation - SPP is more bitter than astringent in nature.

That said, as a wine ages, I have not seen any increase in LPP numbers whatsoeve3r - curious where you got this data?

Let's continue this discussion - I'm curious to hear more from your end . . .


Dear Larry:

Thanks for chiming in. It is great to have someone familiar with the assay in on the discussion.

We need to be careful that our nomenclature doesn't chase its tail. Pluto was ruled a non-planet because it didn't conform to our theories of planetary formation. But that means the theory is wrong, as Pluto is clearly one of the "wanderers" about which the term was coined.

Similarly, SPP and LPP are simply the numbers we get when we do the assay. SPP isn't necessarily shorter than LPP, and in the study we showed the opposite in some cases. I do believe that LPP correlates more with astringency than with polyphenolic content, but in truth there is no assay for polymer except by subtraction from monomer. We are able to do this physically with ultrafiltratoon, and there is a body of work using this approach which was done at AWRI. HPLC similarly can differentiate the small stuff from the big stuff, but iss unable to differentiate different tyupes of big stuff -- you just get one big peak.

Through these means we get the picture that polymerization to tetramer and above is complete within a year or two; faster if MOx is employed. This is what I said. LPP doesn't increase after that, but astringency does, once the wine fallls apart into grainy, dirty tannin. Prior to that in sound wines we get a softening of astrigency without much change (if anything a decline) in LPP.

If, and I say if, Andy Waterhouse actually claimed that astringency correlates with tannin content, he is wrong. However, if his way of measuring tannin content is LPP, then he'd be right. However LPP doesn't measure tannin polymer, so the argument is circular. When you say "higher levels of tannins," what assay were you using?

To my mind, there is nothing to assume about different types of tannin. The distinctions are perfectly obvious to the trained palate. Any collector knows that in general as sound wines age, they get less astringent for a while, and as they fall apart, they get very dry.

The standard wine you used may have shown bitterness for SPP, but this isn't necessarily the case for all such tannins, and those achieved through elevage techniques such as lees stirring or MOx probably are quite different in sensory properties from your model system. Again, SPP is simply the number you get from the assay -- the tannin that isn't bleachable and doesn't adhere to BSA. There is a whole world of different polyphenolics that fit into this category.

I would caution that your assertion that MOx speeds up ageing only encompasses the rather trivial Phase Two and Phase Three work. In early, pre-ML work, as predicted by SIngleton's explanation of oxidative phenolic polymerization (the vicinyl diphenol cascade), anti-oxidative rates actually increase. Thus, done well, MOx extends longevity and retards the ageing process. In the hands of the unskilled, yes, it can certainly do the opposite.

Concerning the 1999 study, it was written up by Patrick Sullivan. I have lots of results from this ten-winery collaboration, which I should post in a blog specific to the subject. From the point of view of this discussion, I'll center on one diagram, a factor analysis of sensory properties from four treated wines. It shows quite clearly that the treated wines had less green tannins, lower grit size (using standard sandpaper grades from 80 to 1500 compared to palate impression), lower vegetal aromas, LESS oxidized aromas, and more fruity aroma.

More on this when I get a chance. Meantime, I certainly recommend the Adams assay for getting a bead on astringency, and it is also a wonderful way to look at monomeric anthocyanin during grape riening, a key tool in preventing field oxidation through excessive hangtime.

Larry Schaffer:



Thanks for the reply. One thing we should discuss is the true importance of LPP or SPP vs total tannins as determined by the assay itself. It is this number that is what was correlated in my sensory study, NOT LPP or SPP. I would tend to disagree about the importance of SPP in general - we found its significance much less important to looking into more information about the wine vs. LPP or total tannins.

Also, the ascertation that LPP numbers decrease and then suddenly increase as a wine gets older is not something I witnessed at all - not to say it doesn't happen, but I am not aware of this.

I do think it's important to differenatiate nomenclature here a bit so as not to confuse others . . . just as it's important to let viewers of this blog know about your take in general about M0X - I think you might to yourself a disservice to not state your stance on the subject matter - and your financial stake as well . . .



The whole point of the Adams assay was a more reliable substitute for the Yves Glories gelatin index (used throughout Europe but rejected by Davis as unreproducible, and probably rightly so) as a predictor for binding to salivary protein. Gelatin is hard to standardize. BSA is THE standard protain, and I think Doug was fortunate that a compact spherical protein that looks nothing like a mucin seemed to work anyway.

LPP has always been intended as an astringency determinant, and I think successfullly so, to the extent that astringency is caused by salivary protein binding to polyphenolics. SPP isn't supposed to be important to astringency -- it's the non-adherent stuff. I do suspect it's very important for macromolecular effects like the aromatic integration Sue Eberler has talked to me about.

Yes, I need to write an article about MOx. There is a great deal of material available for technical people on the subject, but I need to boil down a complex topic into something digestible for the interested lay readership.

By the way, I have no financial interest in MOx. While I used to sell Oenodev's gear, we sold that company five years ago. I decided to maintain the technical material on because I think no winemaker can properly do his job without a fundamental understanding of how wine behaves around oxygen, which is very couter-intuitive.

You and I at the moment have somewhat different points of view, and I can sympathisize that as hard as you have worked as a researcher, you will be looking through that lens as long as it serves you as a winemaker. For me, Davis training was very useful when I was making whites at R.H.Phillips, but the doctrine broke down for me when I satrted working with big reds. Red wine chemistry is so complex that it's very difficult to talk about from a skeptical scientific perspective, and I recommend developing technique instead of perfecting theory. That's what Mark Bunter is after, and that's what pays the bills.

Winemaking is not fundamentally about advancing the depths of Truth; it's about putting something soulfully delicious on the table. Quite simply, it's a branch of cuisine. When I talk to Davis trained, research-oriented winemakers, I feel like a chocolatier who is trying to teach how to make truffles, bittersweet, or milk chocolate, and I'm getting this blank stare from people who have studied cocoa powder all their lives, worked very hard at it, but can't cook. They're saying to me, "Our results show that these distinctions don't exist; our numbers indicate that it's all just cocoa powder." So I'm saying this is obviously nonsense to anyone who knows what chocolate is.

My interest is very practical -- to pass along chocolate-making methods (in Chocolate MOx is called "conching") to red wine makers who are currently bottling the equivalent of cocoa powder -- harsh, unintegrated, and dry. This doesn't mean everybody has to make wine the same way, but the acumen should be transmitted more freely. What's missing is a feel for wine itself.

I have no dog in this fight. The main financial vested interest in this discussion is the Davis folks we have named, who continue to attract grant money to pursue blind studies rather than to tap into the experienced professionals such as Patrick Ducournau, who have been working with the technique daily for decades. Why the two aren't working together is appalling to me. Part of the Davis disease is to consider people in the trade as hopelessly tainted by vested interest, while failing to acknowledge their own axes to grind.

mark bunter:


Clark et al
This is all quite interesting, but possibly verges on the technical for technicality's sake. I followed the link to the web page describing the two assays, one for comparative qualitative sensory analysis of grapeskins, one for measuring astringency using a spec. Both seem a little tedious for practical use, although well-considered and rather ingenious. I especially like that this skin assay removes the obvious interference of sweetness, which has inexplicably been heretofore overlooked. Winemakers say "I pick when the grapes taste good"- of course they taste best at 28 Brix, you idiot! 
What I take away is that you, Clark, say the Adams assay for astringency can determine when grape tannins have attained their optimal quality/quantity (there seems to be argument about which), which could inform picking decisions. Two questions arise. The procedure specifies "wine sample". Is this assay useful on field juice samples, if so, how are they prepared? What are the ideal numeric absorbance parameters that coincide with ripe tannins, in juice?
My general stance on these issues is that, thank God, wine is just too complex to be understood or predicted using methods such as these. However, if there is a useful (and practical) tool, I want to know about it. Regarding astringency and tannins, it's difficult or impossible to understand all the factors that influence our sensory experience, and resulting intellectual judgements. I want to make the best wine I can, and I want as much knowledge as I can get to help me do that. I seem to have little control over the ultimate quality or nature of a wine's astringency- it always becomes obvious after a while, and I may change vinification procedures to reflect past outcomes, but I could never predict at harvest whether, upon maturity, a wine's tannins will be rich, velvety and mouthfilling, or harsh, dry and unpleasant. So many other factors come into play. I am a little skeptical that an absorbance number will reliably indicate how the wine will taste years down the road. Does anyone have any practical, empirical knowledge to share regarding the usefulness of this assay, or is this all just a really fun academic intellectual jousting tournament? I would buy a spec if this stuff really worked. I don't wish to use micro-ox, so no mox testimonials need apply, thank you. Great monograph, as usual, Clark, and interesting reply posts from people with valid opinions and real experience. Thanks to all. I am especially grateful for the vignette of Dr. Noble with a mouthful of protein goo.


OK, practical takehome message is this. Firstly, a cheap spec has a thousand uses, and you should save up for one, if only to track you ML's. You don't need the UV bulb -- save yourself two grand that way and read at 340nm.

To do the assay on grapes, you take a 200 berry sample, weigh it, squeeze out the pulp and seeds with thumb and forefinger, and make back up to the original wieght with an 8% alcohol citric buffer in a blender. You can look at and taste this just like wine, and save it in your fridge, observing changes during maturity.

The critical values from the Adams assay for grape ripeness are Bleachable Pigment (unpolymerized color) and A420, the browning index. During ripening, BP rises (as it's created in the berry) and then falls as oxidation sets in and polymerizes it. You divide A420 by BP and get a horizontal line which suddenly goes verticle. That's your smoke alarm.

Corey Morris:


We have a mobile cross flow filtration business in the Willamette Valley in Oregon. We use a 2008 Koch machine with hollow fiber cartrdiges which will remove particles to .2 microns. I recently had a winemaker note that he thought the Pinot we filtered took out some of the "greeness" of the tannin profile in his 07 wine. (A vintage known for unripe fruit in some cases.) Is there any science to support this claim, and how does the filtration process affect structure in our Pinots?

I have given my reply a separate post as "Crossflow Pros and Cons." Thanks for your note.

Matthew Simson:


Have you found that the tannin assay works better on some reds than others? I am getting good results on Cabernet Sauvignon and Chambourcin using Adam's modified assay, but have not obtained reasonable anthocyanin or SPP levels for Norton grape wine. Any suggestions for this wine which is an ultra-dark wine?

Norton is a very unusual wine because unlike most wines which finish fermentation at around -1.5 brix by hydrometer (due to the lower density of alcohol than water), Norton finishes at around 0 brix, meaning it contains about 1.5% dissolved solids. I believe these to be polysaccharide in nature, although mannoproteins and pectins could play a role. These seem to coat the tannins, giving Norton the pleasant richness and body which is so highly prized, and like many macromolecular colloids also bring about aromatic integration and the resulting soulfulness. (These also cause Norton to make fabulous sparkling wines.) These affects are sometimes offset in Norton by high TAs which mar the smoothness of the tannins.

I haven't tried Adams on Norton, but I would imagine that it rates very high in SPP and low in LPP. I don't know how this would affect the BP.

I'm not sure what you mean by unreasonable results. Can you be more specific?

Mason Kleifgen:

To marketplace in terms of quality rather than cost, and in order to differentiate consequently, you require to play along the canonical format of the 4 Ps marketing plan. That is, Price, Product, Place and Promotion obviously you recognise the cardinal attributes of the merchandise, and the cost, but for place you should think about the type of mass who are willing to give over 4x price of competing merchandise whereas the low-cost option may be sold where accent is on cost, your ware will be suited to places/distributors where the clients will be willing to pay for quality. Thank you for this article! I've just learned a really perfect news site about seo advertising Attempt it!