A vintner’s horror movie nightmare: The smoke is coming from inside the grapes

There are a lot of ways wines can go wrong. You may not have heard of this one yet, but winemakers fear its name: smoke taint.
There are a lot of ways wines can go wrong. You may not have heard of this one yet, but winemakers fear its name: smoke taint.
(Johanna Goodman / For The Times)

Some effects of a wildfire are visually arresting. Others are invisible, dormant for months, and then, suddenly, smellable. You may not have heard of it yet, but winemakers fear its name: smoke taint.

There are a lot of ways wines can go wrong. Maybe you’ve had a corked wine, where a fungus-exposed cork imparts a musty off-flavor like a dank basement. Smoke taint rears its head when grapes, kissed by environmental smoke as they’re growing, eventually yield a wine with unexpected smoldering flavors.

Those of us who enjoy mezcal or smoky amari might be intrigued — smoky wine sounds, potentially, like a new undiscovered natural wine trend. But smoke-tainted wine is less Laphroaig-writ-grapes and more like drinking from a well-used ashtray, with overpowering old-campfire and cigarette smoke flavors, according to Anita Oberholster, a cooperative extension specialist in enology at UC Davis who studies the phenomenon.


Last year’s summer and autumn wildfires up and down the West Coast mean that even those vineyards that seemed unscathed aren’t necessarily out of the woods yet — and winemakers and wine scientists in California, Oregon, Washington and British Columbia are working to try to understand smoke taint and its effects before expensive problems — like unsellable wines — come up.

Wine scientists began studying smoke taint in earnest in Australia in the aughts, after summer brush fires starting in 2003 led to millions of dollars’ worth of unusable grapes and wine. Smoke-tainted wines contain unusually high levels of odiferous, smoky-smelling products of burning wood called volatile phenols, including the molecules guaiacol, 4-methyl guaiacol and syringol. These volatile phenols are the same molecules that make bacon or Islay scotch taste smoky and campfires smell like campfires.

You’d be forgiven for thinking that there could be an obvious answer to smoke taint — if you don’t want smoke-reeking wine, then don’t make wine from smoke-reeking grapes. But sometimes, grapes that are spared immolation from wildfires mature normally and smell fine on the vine at harvest. It’s only once they ferment that the smell of their smoky bath months earlier returns in full force, seemingly out of nowhere. This smoke-flavored disappearing-reappearing act is the result of a diabolical biochemical accident caused by the interaction of smoke molecules with the natural enzymatic machinery of grapevines and yeast.

A little soupçon of the chemical ecology will help out here: Most of flavor comes from smell, and many of the smellable molecules in food and drink are biosynthesized inside plants as zippy, wafting chemical messages. That ripe fruit, grapes included, develop lush and tempting aromas is no accident: Those aromas are a message from the plant that the seed inside is mature and ready to be deposited somewhere else by whoever, or whatever, eats the fruit. A grape, or any other fruit, uses enzymes to manage these molecular flavor signals.

Smellable molecules are volatile and quite happily turn into a waftable, gaseous vapor. A group of enzymes called the glycotransferases helps keep them contained until the grape is ready to send its “I’m ripe, eat me!” Message out into the world. They’re very effective at grabbing a stray flavor molecule and attaching a stodgy, scentless, nonvolatile sugar molecule to it as an anchor. This assembly, an equally scentless glycoside, stays dissolved inside the grape, waiting to be cleaved apart by a different enzyme and unleashed to announce ripeness or some other message.

Where grapes and wine can run into trouble is that while these sugar-attaching enzymes are very effective, they are not very picky. A glycosyltransferase doesn’t care much (and can’t tell) if an aroma molecule it encounters was made inside the grape, or if it got diffused into the grape from the outside, especially if it has some structural similarity to molecules the grape does make on the regular. In the normal run of things, a grape makes some flavor molecules and a glycosyltransferase catches them and turns them into glycosides. In the case of wildfire smoke exposure, flavor molecules (this time, smoky ones, rather than ripe fruity ones) diffuse into the grape, get picked up by the glycosyltransferase enzymes — as if they were the grape’s own flavor molecules — and are packed away in hidden, scentless, glycoside form.


This means that a smoke-exposed wine grape might have very few smoky-smelling volatile phenols like guaiacol or 4-methyl guaiacol inside it in smellable form, but have a lot of their glycosides as “helpfully” synthesized by the glycosyltransferases. These compounds aren’t directly detectable by smell, but can convert to a smellable form later.

To turn grape juice into wine, you ferment it, and to ferment it, you use yeast. Yeast is a fungus, and fungi make lots of enzymes to digest stuff. Shiitake mushrooms make enzymes to digest wood; Aspergillus oryzae makes enzymes that help turn rice into sake and miso. Yeasts produce enzymes that are really good at cleaving apart glycosides and other molecules that include a sugar. This is one of the reasons wines have more flavor than grapes — these enzymes break down the glycosides of fruity, floral and juicy molecules that the grape had laid away in reserve, releasing their flavors for drinkers to enjoy. This also means that if a whole bunch of glycosides of volatile phenols are created after smoke exposure, the yeast will cleave them up too, creating a bloom of smoke flavors that weren’t fully evident in the grape.

This all adds up to a wild goose chase for winemakers: When they’re in smoke, volatile phenols smell smoky and get absorbed into grapes through the skin. Then, those smoky phenols stick around, but get transformed and locked away as glycosides by the internal workings of the grape, and lose their scent. Then, when the grapes ferment into wine, these scentless phenol glycosides get re-transformed into their normal forms, which smell smoky once again. Which means that smoke-exposed-but-not-currently-smoky-smelling-grapes plus yeast can equal really, really smoky-smelling wine.

It’s a vintner’s worst B-horror movie nightmare: The smoke is coming from inside the grapes.

In the relatively slow-moving academic world, the increasing and geographically widespread issue of smoke taint has lighted a fire under the wine science community to figure out its more complex and locally specific details, often getting very DIY in the process.

For a recent paper in the journal Planta, University of British Columbia assistant professor Wesley Zandberg and his collaborators studied smoke taint by deliberately exposing growing grapes to smoke. Since the field of empirical smoke taint research is pretty new and doesn’t have a tidy experimental kit you can buy off the shelf, they took a hands-on approach to creating the smoke and getting it onto the grapevines in controlled doses.


“We went down to Home Depot and got PVC electrical conduit and greenhouse plastic, then stitched it together so we could enclose a 60-foot block of 12 grapevines under one cover,” Zandberg explained. When so encased, he and his team used a commercial salmon smoker attached to a dryer duct to blow doses of smoke onto the vines.

After getting up in the dark for an entire summer — they needed to assemble the DIY greenhouse around the experimental vines, pump smoke onto them, and break the setup down before the sun rose to avoid a greenhouse-effect temperature stress on the grapevines — Zandberg and his collaborators cast a wide chemical net. They found that not only was smoke exposure creating the expected glycosides of volatile phenols, but that it changed other, seemingly unrelated parts of the grapes’ metabolism as well. Totally separate from the smoke flavors, molecules that contributed to tannins, color, mouthfeel and flavor in the grapes were being biosynthesized in different amounts in the smoke-exposed grapes than in their un-smoked control counterparts. This would be like if you or I ate a lot of garlic, and as a result got garlic breath — and got a sunburn and faster-growing hair at the same time.

If the first wave of smoke taint research was finding volatile phenols and the second wave was examining their glycosides, Zandberg and his collaborators also made sure to build their experimental design around a third wave: local flavors. To make the experimental smoke treatment, they used bark and needles from endemic ponderosa pine. We’re beginning to understand that smoke taint has a site-specific character all its own — an airoir, if you will. The chemical composition and flavor of smoke depends on what type of vegetation is burning, and how much lignin, cellulose and hemicellulose (the building blocks of woody plant tissue) it contains. Just as applewood, oak and mesquite fires produce different flavors of barbecue, wildfires on shrub land, a pine forest, or a stand of eucalyptus will all make different types of smoke and different types of smoke taint.

With the increasing sophistication of our smoke taint knowledge, in the coming years we may be able to intervene sooner and in more locally appropriate ways when wildfires threaten our grapes. In the meantime, if you want to taste a creative approach to dealing with smoke taint yourself, some wineries are finding that California’s smoke clouds have a silver lining for the open-minded — as is the case with Yosemite Cellars’ Rim Fire Red, a blend of smoke-exposed and unexposed grapes. Perhaps we’ll see more finessed smoky blends using 2018’s grapes. Check back in 2020 or so.