Is winemaking an art or science?

Coldstream in mid-winter is living up to its name. The township lies deep within the Yarra Valley, a region on Melbourne’s northeast fringes celebrated for its cool-climate wines. The hillsides are lined with vines – brown, gnarled and leafless but shimmering, this morning, in a thick cloak of frost.

The vines may lie dormant, but inside the cellars of Oakridge winery the annual ferment is seething with microbial life. A couple of batches are moving a bit slowly, says chief winemaker and Oakridge chief executive David Bicknell, but that’s par for the course. When the production of fine wine relies wholly on the vagaries of the indigenous microbial population, the process never runs like clockwork.

“With commercial yeast you get certainty – you can sleep at night,” says Bicknell. “But how do you make wine more interesting? You exploit the metabolic processes of different yeast species.”

Bicknell’s faith in wild yeasts adds stress at fermentation time, but the pay-off is multi-award-winning wines regularly acknowledged as some of the best in Australia. “The wines do taste different, even if there’s no way you can show that statistically,” Bicknell says. “The only way to really know is to taste.”

Exploiting the diverse and fluctuating populations of wild yeasts found on the plants, fruit and in the air of vineyards is “the new black” (not to mention red and white) in oenology. The practice is becoming more commonplace among artisan winemakers. Even some of the giant commercial wine corporations are investing in the method.

Wild fermentation, says Bicknell, represents the intersection of science, craft and philosophy. But it could also form the basis of a profound shift in the narrative of wine. The more we study winemaking’s microbes, the more it appears they might explain one of the wine industry’s most beloved, but vaguest, terms: terroir.

“Terroir is a wonderful marketing term,” says David Mills, a microbiologist at UC Davis, who studies microbes in wine. “But it’s not a science.”

The French word terroir is difficult to translate. The closest translation is “soil”, but that is just one of its components. Terroir connotes the unique sense of place – the soils, the topography and the microclimate. It’s what makes the wines of Bordeaux or Australia’s Coonawarra so distinctive, and so inimitable.

Sommeliers like Ren Lim, former captain of the Oxford University Blind Tasting Society (and a PhD biophysics student) will tell you merely from swirling a mouthful of Cabernet Sauvignon which Australian winery produced it.

“The ones from Margaret River often give off a more pronounced green pepper note, a note found commonly in Cabernets grown in regions which experience pronounced maritime influences. Coonawarra Cabernets are somewhat different and unique in their own way. They are often minty and have a eucalyptus or menthol note in addition to the usual ripe blackcurrant notes. The green pepper note is often suppressed under the menthol notes. Nonetheless, the Cabernet structure remains in both these wines.”

It’s a feat that Mills does not question. “I don’t doubt regionality exists, but what causes it is a whole other set of issues.”

Terroir has so far eluded science. But that may be about to change. And many places will be avidly watching this science. As climate change plays havoc with existing wine growing regions, new contenders to the wine industry – such as China – will stand to gain from demystifying the secrets of fine wine.

‘I don’t doubt regionality exists, but what causes it is another set of issues.’

It is one of history’s more enduring ironies that while people have been making wine for at least 9,000 years – and recording their experience at least since the times of Theophrastus, Aristotle’s successor, in Ancient Greece – much of the biochemistry remains mysterious. Though not for want of trying. Oenology uses cutting edge technology to support a global wine industry that produces more than 18 billion bottles a year and is forecast to be worth $327.8 billion by 2016. But while researchers have a good understanding of the chemicals that convey bouquet and palate, just how those unique qualities are formed and imparted to the wine is still more art than science.

Traditional winemakers attribute it to the terroir. It may be romantic but it’s a deeply rooted notion. In 160 BC, when the Roman statesman Cato the Elder wrote the earliest surviving text on farming, De Agri Cultura, he made much of choosing the right grape varieties to plant according to position and sunlight. Shortly after, the agricultural writer Columella said that when selecting a site for a new vineyard, “inquiry… must first be made into the excellence of the soil”.

There is no doubt that the soil impacts the growth of a grape vine. As for any crop, soil structure affects the amount of water and nutrients available. Vines in fact should not be too pampered: a little stress adds to the flavour of the wine. But these days, scientists debate as to whether the quality of the soil can in fact explain the unique quality of a wine.

“I say it does,” says University of Melbourne soil scientist Robert White. And his books, Soils for Fine Wines and Understanding Vineyard Soils, attest to that view.

For White, the big bold notes of a Heathcote Shiraz are explained by the deep red Cambrian soils of the Heathcote region. By contrast the lighter character of a Coonawarra Shiraz finds explanation in its famous 27 km long ridge of terra rossa soils, a shallow layer of reddish loam on limestone. “I wouldn’t say there’s no link between soil and wine composition. But it’s probably not a direct relationship,” offers Markus Herderich, the research director of the highly renowned Australia Wine Research Institute (AWRI). For instance part of the mystique of terroir is often attributed to the unique mineral composition of soils – where a mineral found in the soil appears to correspond to a mineral note in the wine. But according to Herderich, “The consensus among flavour scientists is that the minerals in soil don’t play a direct role in wine flavour.”

Many find that hard to believe. “If minerality in wine is a myth, then why does a Beaujolais-Villages have a granite-like bitterness, while a Beaujolais from chalky soil is all about acidity, and why a Cru Beaujolais from Morgon will often deliver a different character again that is supposed to come from the manganese in the soil?” asks wine educator Quentin Sadleron on his blog.

So if it’s not the soil or its minerals, what is imparting unique terroir to wine?

For winemakers like Bicknell it’s a no-brainer. It’s about the microbes. And recent research is backing him up.

Terroir is often attributed to the unique mineral composition of soils.

Most years in the cellars at Oakridge, fermentation starts slowly in the first few barrels. They usually pick up speed as the cellar fills up. “The more you put in there, the greater the population of wild yeast in the air,” says Bicknell.

For wild yeast winemakers, inducing the optimum population of microorganisms is critical – although it is by no means predictable.

“It’s a balancing act,” Bicknell says. “Some years, some batches just don’t want to ferment. They struggle, and no one seems to understand why.”

During the grape growing season, S. cerevisiae (yeast) strains are just one of the microbe species competing for a spot in the seething microbiotic communities found on grape skins, vine leaves, stems and soil. When grapes are first harvested and spontaneous fermentation begins, the chemistry of the brew changes and so does the population of microbes. Hopefully the wild yeasts start to predominate.

But winemakers like Bicknell are practising a dark art. They use their senses and experience – intuitively understanding that they are trying to shape a tiny ecosystem. But just what exactly that population is composed of has been a mystery.

Of course yeasts are the key drivers turning sugars into alcohol. But yeasts also produce so-called secondary metabolites, for instance the volatile esters that produce the rotting fruit note or the volatile thiols that produce the passionfruit note, so prized in a Sauvignon Blanc. Bicknells’ brew not only has many wild species of yeast, it also contains other types of fungi and bacteria – all producing unique chemical notes of their own. But he must be vigilant. If he makes a mistake, bugs like Brettanomyces can take over giving the brew a bouquet of mouse cage, wet dog and sweaty socks.

Researchers are at last getting a handle on the full composition of these microbial orchestras.

Different fungi and bacteria all produce unique chemical notes of their own.

In the sunny Mediterranean climate of Davis, California, David Mills studies microbes involved in wine fermentation. But his other research project profiles the microbes that populate an infant’s gut. That community – the so-called microbiome – can be altered by breastfeeding, increasing the infant’s resistance to disease-causing bacteria. Wherever one looks, the microbiome is making a big splash in human health – from the gut health of infants and adults all the way to autism. The population of our resident microbes outnumber us by 10 to one and it seems our body chemistry may be as much determined by their genes as ours.

So perhaps the same is true for wine. Its chemistry has long been credited to the multifarious family of grape-derived chemicals known as phenols and the alcohol produced by yeasts as they ferment the sugars. But other bugs are also contributing chemicals to the cocktail.

Until recently the true extent of the human microbiome was hidden: most of the bugs won’t grow in a culture dish. They were unveiled when techniques for reading their DNA signatures became fast and cheap. Mills and his colleagues decided to take advantage of those same DNA techniques to take a census of the microbes – yeast, fungi and bacteria – that were found in freshly crushed grapes that had spontaneously begun to ferment.

They tested 273 “ferments” sourced from wineries spread across different Californian geographic regions. And they found differences. For instance, the set of microbes associated with Chardonnay ferments sourced from Napa Valley differed to those from Sonoma and those from the Central Coast. And those differences, like terroir, were relatively stable across two different vintages.

While wine-makers have long recognized that microbes other than yeast play a role in wine-making – lactic acid bacteria, for instance, are used to remove a tart green apple flavour caused by malic acid – Mills says their study uncovered many species that weren’t considered to be involved in wine-making.

And in their paper the authors weren’t shy about suggesting their findings might provide at least part of the answer to terroir, alluding to “the existence of non-random ‘microbial terroir’ as a determining factor in regional variation among wine grapes”.

The study, published in January 2014 in The Proceedings of the National Academy of Sciences, attracted enormous attention. As Mills told the New York Times back then, “The reason I love this study is that it starts to walk down a path to something we could actually measure.… Someone has to prove that something about terroir makes it to the bottle, and no one has done that yet.”

But the Mills team have not yet proven that either.

“We’ve proven that what enters the winery are different microbial sets from different regions. At least some of these microbes are known to influence wine flavour. So it’s led us to a hypothesis that the regional sets might produce regional flavours.”

Proof may be on its way however. Mills’ team are now looking at a correlation between the full complement of chemicals that appear in a wine and the microbiome that was present in the ferment.

A long way from Davis, the vineyards of New Zealand are providing back-up. The country is famed for the crisp, fruity notes of its Sauvignon Blanc. Here University of Auckland biologist Matthew Goddard is testing the idea that it is the regionally different populations of wild yeasts and other fungi that provide some of the answer to terroir.

Like Mills, he tested grapes across the country for the DNA fingerprints of fungi (but not bacteria). However, his study, which came out in April 2014 in Environmental Biology, was a more comprehensive ecological survey. He not only sampled what was present on ripe grapes, he also took the juice and ferments from these, plus swabs from the bark and fruit of the grapevine, and from the soils beneath vines and the surrounding native forest.

And like Mills, Goddard found that ferments sourced from different regions had different populations of fungi in terms of their abundance and types of species. But he also went on to examine the populations of wild yeast responsible for activating ferments. This work, published in 2014 in The ISME Journal, showed that distinctive populations matched up with what was present in the vineyards they came from. And those vineyard ecologies in turn matched those of the surrounding forests. “So the concept flows from microbes in forests and vineyards to the ferments of fruit,” says Goddard.

“Alongside Mills’ paper, the evidence is stacking up. Different regions have different microbes,” says Goddard. And as far as the findings with wild yeast: “It’s the first objective evidence we have for natural regional variation in a species that is key to wine-making. This opens the door to the microbiome as a potential component of terroir.”

So do the different wild yeast populations produce different wines? Goddard and his team have recently conducted this test by fermenting batches of identical grape juice with regionally distinct yeast populations. But we still have to wait for the answer.

‘With these findings, someone can take advantage of the microbiome as custodians of the vineyards.’

Back at the Australian Wine Research Institute, Herderich and his team of more than 30 scientists are also puzzling over terroir. The variability that is possible in the output of a vine is astonishing. Herderich describes how the  compound responsible for the prized pepper note (see Wine Flavour Chart) in a Shiraz can vary 15-fold across vines from a single vineyard, and two-fold in grapes from the same vine depending on row orientation and whether the grapes received morning or afternoon sun. So the pepper note is unlikely to be elicited by something in the soil. But the plant might produce the pepper note in response to stress – perhaps a particular microbial onslaught?

“It’s one of the things we are testing,” says Herderich. Like the Californian and New Zealand researchers, they are doing their own census of the microbe populations and their contribution to unique wines.

“We think the studies on the microbiome are exciting as it potentially can affect wine in at least two ways – through its effects on grape composition and what microbes end up in the ferment. There is no doubt that more complex yeast populations can deliver more complex wines.”

With winemakers like Bicknell out there experimenting, he says they are looking for tools to help manage the microbiome in winemaking. “How wild is a wild ferment? Right now we are at the knowledge-building stage and metagenomics [sampling all the DNA signatures] allows us to measure microbes and biodiversity more elegantly.”

Will it take the art and mystery out of winemaking? The scientists response is emphatic NO!

“It’s a craft,” says Mills. “These findings on the identity of the microbes will be an additional tool.” And he adds far from detracting from terroir “they reveal the amazing beauty of it”.

Herderich emphasizes, “We’re interested in what makes wine unique and interesting – what creates that point of difference. It’s hard to do anything about the soil and the site once you’ve already planted a vine and waited 6-10 years for it to mature. With these findings, someone can take advantage of the microbiome as custodians of the vineyards.”

For Bicknell, the wild microbial populations of Coldstream are not only the key to his winemaking techniques, but also to his business model. Oakridge wines are all produced using natural fermentation. Indigenous microbial populations among the vines are tacitly encouraged to thrive through the much-reduced use of external inputs such as commercially produced enzymes, and through green-waste and water recycling.

The winemaker has worked closely with researchers at the AWRI and is regarded as one of the most compelling voices in the wild yeast movement. But he feels no need to know the precise identities of the day-to-day microbial species that thrive, churn, compete and die on his grapes.

“It’s just one of those things,” he says. “If it’s working, do you really need to know?”

For Bicknell, successful winemaking arises from a combination of craft, science, philosophy and what he terms “benign neglect”.

But successful winemaking is also a business. In this aspect, too, the gradual revelation of the role of microbes in the story of terroir has a positive impact.

“You can spend thousands of dollars on cultivated yeast,” says Bicknell. “But why pay for something when you can get it for free?”

The regional effect of climate change

Wine research has a newfound urgency. With studies indicating climate change may soon force large-scale change to viticulture practice, or wholesale relocation of vine-growing activities away from areas where wine has been produced for centuries, the quest to understand the mysteries of terroir has taken on a new impetus.

One study published in 2013 projected that climate change could devastate much of the world’s established wine-growing regions by mid-century, with the Australian wine industry hit particularly hard. Climate change biologist Lee Hannah and his team at Conservation International mapped the likely impact of climate change on Mediterranean-climate wine regions in Europe, the Americas, South Africa, China, Australia and New Zealand. Using a combination of 17 global climate models fed through two divergent climate scenarios – one with moderate warming, the other a more extreme temperature trajectory – the team modelled potential changes in conditions for viticulture up to 2050.

Under the lower warming scenario, the study concluded that the land currently used for viticulture could decrease by 19-62%, depending on country, by mid-century. The hotter scenario predicted a decrease range of between 25-73%. Vineyards in the affected areas would either be lost, or would require high levels of intervention to maintain productivity.

The modelling also predicted that new areas would become suitable for vine-growing as global warming continued – especially in China and New Zealand – but at a probable high environmental cost. Habitat loss was likely to be significant, while efforts to sustain established growing areas would lead to large increases in water use as irrigation and misting became necessary techniques to keep grape clusters cool.

Of all the world’s wine regions, Australia’s southern growing area – from Margaret River in the far west to the Hunter Valley in the southeast – is likely to be hardest hit, Hannah found, with almost three-quarters of vineyards predicted to need massive increases in water use to have any hope of maintaining viability.

Hannah’s paper, however, has attracted stern criticism from another group of wine scientists, who wrote a rebuttal in the same journal calling its measurements “too crude” and its conclusions unduly “alarming”.

Led by Cornelis van Leeuwen, a professor of viticulture at France’s National School of Agricultural Sciences of Bordeaux-Aquitaine, the dissenting team asserted Hannah’s figures and measurements were not precise enough to predict loss of established vineyards through climate change.

Using data collected from the German Rheingau and the French Burgundy and Rhine Valley wine regions, van Leeuwen countered that the average growing season temperatures in these areas were already significantly above the predicted viable cut-off levels Hannah used, and yet wine production continued apace.

“High-quality viticulture is sustained in these regions despite increased temperatures and dry farming,” he wrote, “because of both the evolution of consumer preferences and implementation of adaptive strategies by growers.”

The down-to-earth taste of down under

Some Australian wines taste distinctly different to their counterparts made in other regions of the world. Australian Shiraz-makers, for instance, can produce a more peppery blend than vintners overseas. Other red wines often display a “minty” note not found elsewhere. Less joyously, many winemakers are coming to terms with a degree of “smoke taint” that reflects rising wildfire activity over the past decade.

Curiously, however, until recently no one was quite sure how each of these flavours got into the wine in the first place. Now, work by the Australian Wine Research Institute in Adelaide and others is yielding vital clues – and in these cases, microbes aren’t involved.

As the processes responsible – both biochemical and mechanical – come to be better understood, winemakers will be able to exert greater control over flavours right at the time of where the vines are placed instead of having to rely on downstream adjustment by blending.

Shiraz accounts for around 45% of Australian red wine production. Descended from century-old Syrah grapevines, many Australian Shiraz varieties are known for their distinctive peppery aroma and flavour. The AWRI’s Markus Herderich calls this pepper note “quintessentially Australian”.

Until recently, however, its existence was a mystery – a matter of legend and marketing rather than evidence. The first breakthrough came in 2007 when a team led by Herderich purified Shiraz aroma extracts and then used their noses to pick out the peppery note.

The compound produced by the grape, named rotundone, gave the peppery aroma and flavour even at very low levels. It was later found to be the same chemical that flavours the berries of Piper nigrum – better known as black pepper.

The minty note found in some Australian white wines has long been associated with the presence of eucalypt trees in or near vineyards, but the exact route by which the gum trees imbued the wine with the aroma was unknown. Discovering it was important – not all wine drinkers welcome a minty presence.

A volatile compound known as 1,8-cineole had long been identified as the source of eucalyptus aroma (you can smell it in a thousand cleaning products, sweets and home remedies), but the task for the scientists at the AWRI required identifying the route by which it entered grape ferments. One theory posited precursor chemicals  already present in the grapes were amplified by fermentation. However, although the precursor chemicals were identified, they were not plentiful enough to account for the 1,8-cineole levels recorded. The compound was obviously coming from nearby gum trees – there was an inverse relation between 1,8-cineole levels and distance from the nearest eucalypt – but by what mechanism?

In the end, an AWRI team led by Dimitra Capone reached an almost anticlimactic conclusion: eucalypt leaves and bits of bark blowing in the wind and coming to rest on the vines were transferring 1,8-cineole into harvested grapes and wine ferments.

Given the propensity for the bush around many of Australia’s wine-growing regions to burn, studies into smoke taint have taken on added urgency. Research into how best to manage smoke taint in grapes is taking place as a joint venture between the AWRI and Australian state government laboratories in Victoria.

Bushfire smoke is full of volatile compounds called phenols. Somehow they enter the grape and change into non-volatile glycosides. But given the right catalyst, they then convert back into volatile phenols and give the wine in which they are residing a distinctive “ashy” taste.

The mechanism through which the smoke-borne volatiles enter the grape is still unknown and the focus of much work. But the catalyst was recently discovered: it happens pretty much at the last stage of the winemaking process – the actual drinking. Enzymes present in the saliva of wine drinkers are the critical last catalyst in the expression of the ash note.

And while work into smoke taint continues apace, perhaps the eventual solution to the problem will rely as much on consumer education as on-vine mitigation.

Herderich recalls growing up in an area of Germany where grapevines grew thick on steep hills, around the base of which ran a railway. As a child he recalls steam trains running along the line, and his grandfather telling him how the wines made from the vineyards were celebrated for their “subtle smoky flavour”.

“And then the railway lines changed to electric,” Herderich recalls. “The wines were never the same after that.”