Terroir, and its expression from individual vineyards, is a topic that arouses a passionate response from almost every winegrower.  Do soils really matter, or is vine water supply the only important difference underlying seasonal or regional variation?  James Wilson, in his classic “Terroir, the role of geology, climate and culture in the making of French wines” favours geology, as do the owners of the 2,816 specific climats in Burgundy, each of whom, when you meet them, is eager to spend a couple of hours detailing the structures of their soil and sub-soil.  Expression of terroir may also be enhanced by using endogenous yeasts to control the ferment, or are the yeast company salespersons to be believed when they debunk this possibility? 

Some recent publications give solid data which may be used to judge the reality of terroir and some of the mechanisms by which it is expressed in wine:

• Influence of soil geochemistry on the chemical and aroma profiles of Pinot noir wines. Imre et al, (2012), J. Food, Agriculture & Environment 10, 280-288 (not currently available for public download, email john@maoripoint.co.nz and I will send you the PDF).
• Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate: Bokulich et al, (2014), PNAS 111, E139-E148 (http://www.pnas.org/content/111/1/E139.full).
• Monitoring Seasonal Changes in Winery-Resident Microbiota: Bokulich et al, (2013), PloS ONE 8, e66437. (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0066437#pone-0066437-g005).

Stephen Imre as part of his PhD thesis at the University of Auckland studied 3 closely adjacent sites at Bannockburn in Central Otago, on Mt Difficulty, Olssens, and Felton Road, all growing ungrafted 10/5 Pinot Noir on three distinctly different soils.  Microvinification, sensory analysis and some detailed chemical analyses were used to compare the wines.  Soils at each site were analyzed in great detail:  they ranged from sandy-silty loam with some clay at Felton Road, a thin layer of soil over schist gravels at Mt Difficulty, to fluvial gravels at Olssens.

Wines from each site were distinctly different, with different profiles of aroma compounds, tannins, and polyphenols, and differences in colour.  The authors conclude that as these compounds are grape-derived, it is likely that they reflect differences in the soil component of terroir.

An aside on high through-put gene-screens. Conventional microbiological assays rely on a swab applied to a dish of culture medium.  Living microorganisms grow until there are enough to be seen and identified.  If the culture conditions are not appropriate they don't grow and can't be seen.  A gene-screen is based on non-selective amplification of all the DNA in a sample.  The amplified DNA is fragmented and the fragments sequenced.  The amplified DNA can be applied to a gene-screen chip which has 20,000 or more known gene sequences on it, or the sequences can be directly run through a computer program to recognize the genes expressed and relate these to different organisms. These are very powerful techniques with a wide range of applications, for example identifying all the organisms in a sample of sea-water, or of soil.

Bokulich et al used a gene screen approach to compare fungal and bacterial populations on 273 samples of Chardonnay, Cabernet Sauvignon and Zinfandel grapes from 4 of the major grape-growing regions in California.  The grapes were destemmed and crushed, and samples taken from the fresh must before fermentation had begun.  Each grape variety within each sub-region had its own pattern of fungi and yeasts growing on their surface.  Each cluster of neighbouring vineyards had their own pattern in common, different from those elsewhere.  At the single vineyard level it was possible to detect seasonal vintage differences, more in bacterial than fungal populations, but these were less obvious when comparing regions.  That is to say, regional differences remained constant, year by year.  Most Saccharomycetes yeasts were of similar abundance, but other yeasts contributing to wine quality, including C. zemplinina, Lachancea thermotolerans, Hanseniaspora guilliermondii, and Pichia species, were expressed differently.  There also were characteristic differences in expression of organisms with negative  features for wine-making, such as Gluconobacter. The authors conclude that “the nonrandom distribution of key taxa with crucial impacts on wine quality supports the potential role of microbial biogeography in shaping wine terroir.” 

It is often suggested that “wild yeast” ferments actually are driven by yeasts introduced via contamination in the winery.  To test this, Bokulich and collaborators assayed microbial communities from all the surfaces and major equipment in a winery (the UC Davis research and teaching facility).  They found large populations of S. cerevisiae and other yeasts prior to harvest, but most of the surface communities were comprised of organisms with no link to wine fermentations. The yeasts present varied with time of year and past history of use of the facility.  Spoilage-related organisms were almost absent, indicating that normal winery practices are adequate for their control.  They suggest that the quantity of Saccharomycetes yeasts prior to harvest was such that there is a reasonable chance that they could contribute to a so-called “wild yeast” ferment.  It should be noted that all fermentations in this winery relied upon inoculation with S. cerevisiae.

How can we collate these results? Bokulich et al's study on endogenous winery organisms was performed earlier than their grape study.  The grape study reveals the organisms actually present in must at the start of fermentation, without defining their origin as winery or vineyard-derived. Different varieties of grapes crushed in the same facilities retained the patterns of microorganisms characteristic of the variety, not the winery. The UC Davis winery adds commercial Saccharomyces to all their ferments, so these are the yeasts that may accumulate there. The numerous studies demonstrating the presence of particular yeasts in different wineries are likely to be correct, but other than commercial yeasts inoculated into wines, those yeasts most likely entered the winery with the grapes.  We can conclude that endogenous yeast ferments will depend largely or completely on yeasts coming in from the vineyard, and that these depend upon terroir, with adjacent vineyards likely to be similar, and more distant ones to have their own characteristics.

Fear not, we will continue this discussion in future posts...