Vineyard fires

Where there's fire there's smoke

Bacon. Biltong. Salami. Old leather. Stuffiness. Cigars. These flavours might not be unwelcome in a gentlemen's club, but you wouldn't want them in your wine. They're all associated with smoke taint, an issue sure to become ever more relevant as the first wines from vineyards plagued by the fires earlier this year are released. By Maryke Visagie.

 

"Old leather" is a relatively merciful term when it comes to tasting panels' descriptions of smoke tainted wine. "Smoked meat", "disinfectant", "ashtray" or even flat-out "rancid" are all included in the vocabulary used to describe tainted wine.

There is almost always an unpleasant, acrid bitterness on the palate - in fact, it is important to note that even if the taint is not detected on the nose, the unpleasant finish may still appear on the palate.

So even if farm owners and winemakers shared a sigh of relief about their saved vineyards, with the fires in some instances being extinguished mere metres from the vines, their problems were only just beginning.

South Africa isn't alone in this regard. Wild, uncontrolled fires swept through Victoria, Australia, in 2003, 2007 and 2009. Western Australia was a victim in 2007, while Canada's Okanagan Valley had its turn in 2003 and 2009. California suffered in 2008, Greece in 2007 and, of course, the Cape winelands in 2005 and again earlier this year.

Steenberg was one of the properties affected by the 2005 fires. "Smoke taint gets absorbed by the waxy bloom of the berry," says general manager John Loubser. "This definitely imparts a smoky character in the wine. I cringe every time there is a fire because I know just how far reaching its impact is."

Australia is the wine-producing country that can be seen as world leader on the subject, as they regularly experience devastating fires and have done a lot of research on the subject. David Wollan, a director of Memstar, an Australian wine membrane technology specialist with a specific interest in smoke taint, says the problem isn't with vast, uncontrollable veld fires alone: controlled fires for the purpose of lowering further fire risk contribute just as much to the taint problem. (For that matter, even car fumes can play their part.)

CHEMICAL REACTION

It seems a fire can affect even distant vineyards, so the problem might be more widespread than first imagined. Grapes are susceptible to smoke taint any time after flowering, but greatest susceptibility occurs in the period straight after veraison and, when anthocyanins are produced, the grape skins integrate these volatile chemicals into them.

The longer the exposure, the more taint can be expected - but, to further soot the picture, it seems only a relatively short exposure is needed. The only light at the end of this smoky tunnel is that, fortunately, no seasonal carryover is detected.

But first things first. Smoke is essentially the product of thermal degradation of wood components. The breaking down of lignins releases literally dozens of volatile compounds, including phenols, esters, acids, lactones and pyrenes. Depending on the fuel source, these compounds may vary, so a fire in a eucalyptus plantation will release different compounds to a pine forest burning, or a fire in grasslands.

Although smoke releases this myriad compounds, two of these are especially important in studies as they act as indicators for taint problems. They are guaiacol and 4-methylguaiacol, two compounds that aren't necessarily the primary cause of smoke taint but are useful markers as they are easily measured and easily analysed. (Interestingly, 4-ethylguaiacol and 4-ethylphenol are also closely related to the compounds associated with Brettanomyces taint, which may shed some light on the so-called "burnt rubber" issue of recent months, sparked by British wine journalist Jane MacQuitty's controversial comments about South African wine bearing this off-putting characteristic.

Wollan hypothesises that "burnt rubber" might be attributed to smoke taint, rather than the grapes' intrinsic quality, but that is a story for another day and a statement that certainly can't be substantiated without further study.)

CLOUDING THE ISSUE

So, given that smoke contains these volatile components, how are they transferred to the grapes, and does it really have such a huge effect on the grape berries, given that a shirt worn at a braai only needs one night's hanging to get rid of the smoky boerie smell? Could a cloud of smoke that billows over a vineyard for a few hours really render it tainted?

Things are, it seems, infinitely more complicated when it comes to living, "breathing" organisms. The exact mechanism of absorption isn't exactly understood but, according to Wollan, there is a likely explanation. "Smoke particles are thought to be absorbed into the leaves, after which the taint compounds bond with the sugar, transfer to the phloem and then to the skin." This makes smoke taint more problematic in red wine, since it spends more time on the skins.

An Australian study also notes anecdotal evidence that the taint problem intensifies during the winemaking process, specifically during fermentation when the extraction of these compounds from the grape skins is accelerated. Nearly all of the taint compounds are extracted in the first two or three days of fermentation.

TAINTED GRAPES - NOW WHAT?

According to Wollan, numerous trials have been conducted over the years with smokeexposed grapes. Protectants like kaolin clay or acrylic polymers gave mixed results. Vineyards were washed with cold water, warm water, with or without surfactants, ethanol and even milk. There were even experiments with leaf plucking, but none of these really did the trick.

The next step was to intervene in the fining process of the juice or wine. In this instance only very large doses of activated carbon really had any benefit, while other fining agents such as gelatin, PVPP or tannins were unsuccessful or had a marginal effect. Carbon fining of the grape juice, however, often strips it of favourable flavour character.

EASY DOES IT

The only thing Wollan readily recommends is gentle harvest practice. Extraction should be minimised by hand harvesting and gentle pressing. It is also recommended that the grapes be chilled before harvesting, as studies showed that grapes pressed at 10ªC show less taint than grapes pressed at 25ªC.

Another option, not so gentle at all, is to feed the tainted wine through a membrane system. Although some winemakers don't like the idea of reverse osmosis as it can strip wine of other favourable flavours, Wollan uses the simile of a cancerous growth to explain that there really isn't any other choice. "Chemotherapy will inevitably kill healthy tissue as well, but not administering the treatment isn't an option."

The wine is pressurised with a highpressure pump and forced through the membrane system, where the permeate containing the taint compounds filters through a semi-permeable membrane. The permeate then passes through an absorbent treatment (usually activated carbon or resin) where the taint compounds are removed, after which the permeate is returned to the wine.

Wollan is of the opinion that reverse osmosis is by no means the definitive answer, and questions remain. For one, it isn't clearly understood exactly which compounds are removed by this technique and it is suspected that it varies from fire to fire. Furthermore, the monitoring of the treatment relies very much on the palates of the winemaking team. And, last but certainly not least, treatment isn't cheap. Premium brands may be able to afford to treat the wine, but for brands along the lower tier there might be a decision not to bother, depending on the level of taint.

STORIES FROM THE SCARRED VINEYARDS

Helderberg winery Wedderwil was one of the farms hardest hit by the summer fires of 2009, with about five hectares of Cabernet Sauvignon, Merlot and Shiraz completely or nearly destroyed, amounting to a loss of 20 to 25 tons. Fortunately their premium blocks didn't burn down. General manager Wolfgang von Loeper says they carefully monitored the precious grapes left over and found that smoke taint levels were below the threshold at which taint can be detected sensorically. "We didn't treat the grapes any different, but we harvest by hand anyway."

The Cabernet and Merlot are intended for blends, while the Shiraz will be bottled as a single variety. Since they are now selling wine from the 2005 vintage, the 2009 vintage should reach the market in four or five years.

On the other side of the Schapenberg, Chris Joubert of Lourensford is also sighing with relief, since tests on the red wines have also shown guaiacol levels to be below the sensorically detectable level. "We only had fire damage in about half a hectare of vineyard, with fire lapping up the grass in between the vines and scorching some leaves." The real damage was in Lourensford's tree plantations, with 820ha lost. By the time the fire broke out, all the white grapes had been harvested with only the red varieties still on the vine. "We monitored them very closely, but we're very pleased with the recent laboratory tests."

It's four years down the line for Steenberg since the fires of 2005, and the wine has had some time in the bottle. The question is, did consumers notice anything? "We were more aware of the taint than our consumers, but some of the very discerning tasters did mention that they could pick up a strange smokiness," says Loubser.

"I have revisited some of these tainted bottles and I must say that with age it does seem as if the taint gets less. It could now be referred to as flinty or gun smoke."