Playing with Yeast Diversity: Not Quite Science

I feel as though I need to whisper this, but I think that yeast might be trendy. The furor (both supportive and critical) over natural wine is bleeding over into an interest in the “wild” yeasts that contribute to their distinction, and whether or not they’re really “wild.” Meanwhile, a super-saturated wine market means that producers are always and ever looking for ways to distinguish their product. Sometimes that means wrapping the bottle in paper or tying a twig around the neck or blending carbonated Moscato with vodka. But sometimes it means employing new and different yeasts to “add complexity” and finding a flavor profile that lies on the (hopefully) delicious edges just outside our genre expectations for grape and region.

Yeast have had a banner year. Last month, “microbial terroir” made the news when a beautiful paper from a team at UC Davis confirmed that microbes found in the vineyard change with location and climate; in other words, differences in vineyard microbe communities might contribute to differences in regional terroir. (The media largely overstated the conclusions of this study – the scientists haven’t yet linked differences in microbes to differences in flavor, but that’s the next step.) Earlier in 2013, the same folks made a microbe map of Davis’s student winery: they used DNA sequencing to identify every yeast and bacteria they could find on everything from the fermentors to the winery floor before, during, and after harvest. They found that, yep, the fermentation workhorse Saccharomyces cerevisiae hangs out year-round on winery surfaces (along with Hanseniaspora uvarum, one of those other yeasts that can be part of the first few days of a spontaneous fermentation before Saccharomyces takes over). Those findings are important because they back up the long-standing belief that wineries have yeasts-in-residence that will show up in fermentations whether the winemaker deliberately put them there or not.

Looking out for the little guys

Like Pope Francis and Miley Cyrus and GMO salmon and most other organisms that make the news, yeasts have been talked up in large part because they’re controversial. The natural-winemaking-is-meaningless camp has gotten a lot of mileage out of findings like that microbial winery map and other research showing that the yeasts that finish up fermentation are rarely the ones that the winemaker chose, purchased, and dumped into the vat. These findings say that winemakers don’t have much control: if a highly competitive commercial yeast is in the winery environment (maybe because it was used in a previous year or a different wine), it’s going to end up everywhere. If natural winemakers wax poetic about the complexity added by their unique, homegrown microbe communities that arise organically out of vineyard, but the yeast in their vats is mostly some old commercial variety that comes freeze-dried by mail-order, are they just delusional?

Not necessarily. Let’s lay aside for a moment the issues with how some of these studies were performed (the microbial terroir and winery map ones were meticulous; the commercial-yeasts-always-take-over ones not so much). Microbiology isn’t always a game of the biggest and strongest wins. We used to think that it was – that one super-powered microbe fought for dominance and beat the other microbes in the neighborhood into submission or death. We know better now, thanks to much better tools. DNA sequencing lets us find the little guys, the minor yeasts that set up shop on some corner and never make big names for themselves, but keep on surviving. It’s a matter of resolution. Do you look at Seattle and just see Microsoft and Boeing and Starbucks, or do you see the thousands of smaller companies at the corners of their little communities? There’s also some very well-reasoned speculation that the lower sulfur dioxide levels that contemporary winemakers are using are allowing for more extraneous microbe growth than happened in most of the 20th century, too. One of the problems with the studies that show commercial yeasts taking over spontaneous ferments is that a whole lot of sulfur dioxide added when the grapes came into the winery probably wiped out the majority of the grapes’ natural yeast population.

 Creating microbial complexity

Winemakers are trying to take advantage of those exciting, complex, multi-microbe fermentations, but many of them want to do so in a controlled way. They can follow recipes for spontaneous ferments that have worked in the past, but these always come with some degree of risk: even if a given strategy always works for the winery where it was developed, it very well may not work elsewhere. Qvevri wines – the traditional products of Georgia, made by throwing whole clusters of grapes into clay amphorae-like qvevri, waiting, and doing very little else – are an extreme example. Microbiologists have found that qvevri-style wines from a Swiss winery contained strong populations of Rhodotorula mucilaginosa and Pichia anomala all the way through fermentation (someone still needs to study Georgian qvevri wines). Both yeasts signify spoilage when they’re found in non-qvevri wine.

Now, maybe tasters would have rejected these wines as spoiled – regrettably, the scientists didn’t report any sensory analysis information about the finished wine. But we have it on the good faith of many experts that many qvevri wines are not spoiled and are, in fact, exquisite and extraordinarily complex. Pichia anomala can do good things for wine aroma: it produces glycosidases, enzymes that release aromatic molecules from natively non-smelly compounds in grapes). It can also make gobs of acetic acid and ethyl acetate, pushing wine unappetizingly toward vinegar. No one is likely to intentionally round up some Pichia anomala and toss it into a vat of merlot must in the hopes of making something reminiscent of a qvevri because they know the results will probably be undrinkable. But in the context of the qvevri – a very, very different environment physically, chemically, and microbiologically compared with a shiny steel tank in a modern wine-factory – maybe that little bit of Pichia is just right. This leaves winemakers who are looking for difference without being too radical in something of a pickle.

Even if they wanted to (and they probably don’t), they can’t adopt traditional methods like qvevri fermentation wholesale: the context of a contemporary winery is already different in ways that will probably affect whether those ferments succeed. They can experiment with throwing quirky yeasts in with their Saccharomyces, play around and find something that works for them, but without having a scientifically grounded way of predicting the results. Still, that’s what’s happening. The most common approach is to throw several different microbes into a must in an effort to mimic – in a more controlled and less scary way – the variety of a spontaneous fermentation. So what if one commercial Saccharomyces still ends up doing the bulk of fermentation? The very presence of those other yeasts, and their activity early on before increasing alcohol levels either kill them off or make them quiet, is enough to add that much sought-after “complexity.”

Scientific approaches to an ancient art

Scientists are catering to this obvious market need in two major ways. First, they’re coming up with new yeast strains that seem to play well with standard wine yeasts. Genetic engineering is one way to do this, because we can take a pretty ordinary strain of Saccharomyces and engineer it to host a gene (from some other, generally less-useful yeast) that converts a grape molecule into an aromatically attractive wine molecule. With the controversy over GMOs and the general conservatism of the wine industry, though, a more popular strategy might be good old-fashioned selective hybridization – the same technique that has given us Ruby Red grapefruit and donkeys and a slew of other plants and animals. We can force S. cerevisiae to mate with other Saccharomyces species to come up with new, different, and therefore potentially more “complexifying” yeast strains.

 The second approach is to study traditional, successful spontaneous fermentations to identify the yeasts that make them unique in the hopes of including them in a successful blend somewhere else. The past few years have seen something on the order of a gazillion papers published about identifying the yeast and bacteria species found in X unique or traditional wine from Y region. That’s a first step: identifying which yeasts are participating in which wines.

So all of this fantastic microbial diversity is happening. We more or less knew it was there before, but now we can document it. What remains frustrating is that we can point at the microbes and say “cool!” but we can’t work backwards to explain the finished wine in terms of what each microbe contributed. That’s a major gap. Lab studies can tell us what compounds any given yeast is theoretically capable of producing, but what happens in the context of a specific fermenting wine environment is inevitably different. We have plenty of microbiology data saying that mixed fermentations work (see here and here), but we don’t fully understand them. We don’t know how big or active a microbe population needs to be in order to affect wine quality. We know that what compounds a microbe makes depends on the other microbes in the mix, but we usually don’t know how any particular two microbes will play together. We haven’t yet made the connection between these minor microbes and wine flavors.

Until we can trace a yeast to the molecules it produces – in wine, not in a Petri dish – and then trace those molecules to our sensory perception of the finished wine, any experimenting with alternative microbes may lead to great winemaking art, but it isn’t really science.