In house research from Château Margaux
A first growth studies organics/biodynamics, destemming/stems and corks/screwcaps

London, Febraury 21, 2012

Paul Pontallier (left) with Margaux commercial manager Aurelien Valance

In what was probably a first, Paul Pontallier of Château Margaux came to London to present a tasting based on some of the in-house research that has taken place at this first-growth Bordeaux property.

The event was hosted by Yvon Mau, an important Bordeaux negociant, and compered by Richard Bampfield MW. Gathered were many of the great and good of the London wine scene, and a few extras (like me).

‘Experimenting is not news in Bordeaux,’ said Pontallier as he introduced the session. ‘If we are where we are it is because our predecessors were great experimenters. They were first great observers: they observed the terroir and fine-tuned how to grow the vines.’

Pontallier keenly believes in the need for proper experimentation rather than just random implementation of the latest ideas. ‘So much knowledge is now available; so many trends go by—but few remain,’ he emphasized. ‘In order to be able to discriminate the knowledge or technology that will be able to make a positive step, we need to experiment. At the level we are it is much easier to let quality go down rather than up.’

He added that he’s impressed by the research that Bordeaux University does, but that they don’t always study subjects relevant to the particular interests of Château Margaux. So just over 10 years ago, Margaux started their own R&D department, which now numbers two employees. That’s an impressive commitment for just one producer.

‘Lots of people follow trends or their own beliefs, but this doesn’t belong to our culture,’ emphasized Pontallier. ‘Our culture is based on experimental science. We need to experiment very seriously and not draw conclusions too fast. We repeat the experiments. We practice doubt.’

‘It is important not to take our own a prioris for granted,’ he said. ‘Wines are probably as disappointing as people. We must be ready for big surprises.’

Pontallier was quick to point out the frustrations of wine research. One of these is that there are just so many variables involved in growing vines and making wine. ‘The problem here is isolating just one factor. You are playing with several factors.’ 

Another is that in trials you end up having to do microvinifications, making very small quantities of wine. This is quite difficult, and sometimes the results can be counfounded by the fact that many microvinified wines are poor, making it hard to see the differences caused by the experimental variables. Fortunately, Pontallier’s team are now very good at doing these microvinifications.

‘As soon as you experiment in detail you realise that it is not clear,’ warns Pontallier. ‘Things are always more complicated than one would think.’

Trial 1: comparing organics, biodynamics and conventional viticulture

‘We want to get closer and closer to organics,’ says Pontallier. ‘This is something I started 25 years ago. We are very close to organics, and haven’t used any pesticides or insecticides for more than 10 years. We still spray chemicals against mildew, powdery mildew and botrytis though. Could we do this differently?’

‘I hope that in two to three years we might be 100% organic for the grand vin, but not for the second, third or fourth wine. The very best plots are the ones with the best drainage and the least vigour, and which are much less sensitive to mildew and powdery mildew. We will start there.’   

 Pontallier is a little sceptical about the claims of biodynamics, but he has trialled it with the help of a consultant, alongside organics and conventional viticulture. ‘Most of the difference between organic farming and biodynamic farming looks religious from an outsider’s point of view,’ he says.

The plot chosen for the trial was a slightly second-rate plot, with Cabernet Sauvignon that has never made it into the first wine. ‘It is a little bit vigorous. We wanted to put ourselves not in the easiest situation, but a more difficult one.’

The trial started in 2008, so this wine, a 2010, was in the third year of management. There are replicates of each treatment. The trial is now in its fifth year, and you’d expect the differences, if any, to be more pronounced with each vintage. ‘Farmers say that the soil has to recover from its chemical history,’ says Pontallier. ‘This probably makes sense.’

He can’t see any differences between the actual vines, though. ‘Our biodynamic consultant can see differences, but I think he looks with religious eyes. Our soils have not been totally mistreated, so maybe the differences aren’t huge.’

One of the problems with organic viticulture is the reliance on copper-based treatments for dealing with powdery mildew. ‘We use much less copper than in the past,’ says Pontallier. ‘In the past we used 8–12 kg per hectare and sometimes more. 1 kg/hectare is now the maximum. There has been an improvement in spraying technology and we have also realised that more is not useful. One reason that switching to organics is now feasible is because we can now use less copper. Without such a reduction it would have been unthinkable.’

The three trial wines were tasted blind. I much preferred wine number 1, which was a bit more elegant and less green than the other two. Then of the other two, which were more supple, I preferred number 2, which was a bit less green. Pontallier prefers wine 2. Around the room, there was a spread of opinions.

It turned out that wine 1 was the biodynamic wine, and wine 2 the organic wine. Alcohol of the biodynamic wine was 13.4, the organic was 13.6 and conventional 13.8. No indication of statistical significance was given.   


Trial 2 Stems or destemming?

This was involving a better wine: a 2009 Cabernet Sauvignon from a plot which, in good years, makes it into the first wine.

‘We wanted to see how important it is to destem,’ explained Pontallier. ‘Our tradition has been to almost totally destem. From the early 20th Century at Margaux destemming was a standard procedure.’

He points out that some are now suggesting that using some stems could be a good thing. And on the other side, some estates have become more fastidious about removing even the tiniest bits of stem. Margaux destemming leaves some tiny pieces of stems in the ferment, such that 0.03–0.05% of the ferment is stems.

In this trial, the standard Margaux destemming was compared with 1% stem additions, and 1% stem additions but with the stems cut into tiny pieces. The wines were once more tasted blind.

To Pontallier, the results from this trial are obvious. The current approach produces the best wines, and the 1% stems in pieces the worst. But he is still cautious about generalizing the result. ‘We shouldn’t draw too general conclusions. For this wine I think destemming works, but for other plots, such as a rich wine with soft tannins, it might be different.’


Trial 3 Cork versus screwcap

Pontallier has approached the issue of alternative closures with appropriate caution. ‘We might be exchanging one problem for another: this happens many times in life.’Margaux have been trialling alternative closures since 2002, but the trials with synthetic corks didn’t last long as the wines quickly oxidised. ‘After two or three years it was a disaster.’

‘We all know the problems of cork,’ he says. ‘We have all been so frustrated and disappointed by corked bottles. The proportion of corked bottles has certainly gone down over the last 20 years. We have the best corks but it is still unacceptable that there is a single bottle that is corked.’

‘We thought that any other closure would be welcome as long as it is for the better. We are not afraid of changing as long as we are sure it is for the better and not for the worse.’

Trials began in 2002 with Pavillon Rouge (second wine) and have taken in other wines over time, including the Grand Vin. But when news got out that Margaux was experimenting with screwcaps, it all got a bit crazy. ‘I talked to some people about our experiments with screwcaps, and once this news had been published a lot of people thought that we had decided to bottle the whole or half the crop with screwcaps,’ recalls Pontallier. ‘I received plenty of mails. Half of them said “you are crazy” and the other half congratulated us that we had moved to screwcaps and that this was the modern way.’

‘Using a screwcap for white or red wines when you are sure they are all drunk after two or three years makes a lot of sense,’ says Pontallier. ‘But for wines that we expect to go through a very special evolution for 10, 15, 20 or 30 years, we don’t exactly know. We are not brave (or stupid) enough to use screwcaps without this knowledge.’

This trial involved two wines—a white and a red—and three different closures. Good quality natural cork of the type used for all top wines, and two screwcaps—one of which Pontallier describes as gas permeable (presumably with the saranex liner), and one of which is impermeable (the tin/saran liner). The latter is the type used almost exclusively for fine wines in Australia and New Zealand.

The red wine trial involved the third wine from 2003. This is a wine that at that time would have been sold off in bulk. Now it is bottled as a third wine. Unfortunately this was a poor wine indeed: quite oxidative and developed and with a rustic, almost medicinal edge that suggested some Brettanomyces. There is also the problem that natural cork is quite variable, and so we were comparing the screwcapped wines with just one expression of natural cork. Was this a good cork with very low oxygen transmission, or one with a bit more?

Of the three wines, I preferred number 2, and then number 3. Number 1 was my least favourite, as it was the most evolved. It turned out that 2 was natural cork, 3 was tin/saran screwcap and 1 was saranex screwcap. But none of the wines were particularly good, and there wasn’t a huge amount of difference between 2 and 3. A small majority in the room preferred 3 to 2. Fewer liked number 1. But this is with a wine that was probably a bit tired and somewhat oxidative when it was bottled in 2005, and which isn’t destined for long ageing. There’s a reason Margaux were selling it in bulk.

For the white wines, a 2004 second wine, the differences between the bottles was more marked. I liked number 1, which was the outlier. It was aromatic with some breadth to the palate, combining grapefruit and honey notes. I thought it was the most developed, and the best for current drinking. Numbers 2 and 3 were closer together, and really quite different wines to number 1. Wine 2 had a slight chemical-tasting edge to it, and showed fresh, pithy grapefruit characters. I felt it the freshest, but not the nicest. Wine 3 was probably my favourite, and was lively with fresh grapefruit character.

It turned out that 1 was cork, 2 was saranex screwcap and 3 was tin/saran screwcap. Pontallier, rather perplexingly, suggested that 2 and 3 were showing traces of oxidation and that 1 was the freshest, and his preferred. This was not the view held by the majority of the room. The key take home message is that the closure changes the wine. So which wine do you prefer?

It is unlikely that Margaux will be bottled under screwcap any time soon. ‘We are also experimenting with the first wine and Pavillon,’ says Pontallier, ‘but it is too early: 10 years is nothing for these wines.’   

‘Our wines have an extraordinary potential to remain fresh for as long as a century or even more. So we have to be pretty sure if we are going to change the closure.’

A video of the event:

Find these wines with


Back to top