The genealogy of grape varieties

We are all familiar with the concept of a family tree (formally called a pedigree). People have been compiling them for at least a thousand years, as the first known illustration is from c.1000 CE (see my blog post on The first royal pedigree). However, these are not really tree-like, in spite of their name, unless we exclude most of the ancestors from the diagram. After all, family histories consist of males and females inter-breeding in a network of relationships, and this cannot be represented as a simple tree-like diagram without leaving most of the people out. I have written blog posts about quite a few famous people who have really quite complex and non-tree-like family histories (including Cleopatra, Tutankhamun, Charles II of Spain, Charles Darwin, Henri Toulouse-Lautrec, and Albert Einstein).

Clearly, the history of domesticated organisms is even more complex than that of humans. After all, in most cases we have gone to a great deal of trouble to make these histories complex, by deliberately cross-breeding current varieties (of plants) and breeds (of animals) to make new ones. For plants, we have also propagated genetic variants, thus making an even more complicated history (see my post Grape clones and varieties are not always what they seem).

In most cases, we have no recorded history for domesticated organisms, because most of the breeding and propagating was undocumented. So, the idea of reconstructing these histories is very different from the popular pastime of doing a bit of "family history research". Indeed, until recently it was effectively impossible. This has changed with modern access to genetic information; and there is now quite a cottage industry within biology, trying to work out how we got our current varieties of cats, dogs, cows and horses, as well as wheat, rye and grapes, etc.

Grapes

There are a number of places around the web where you can see heavily edited summaries of what is currently known about the grape pedigree. However, these simplifications defeat the purpose of this blog post, which is to emphasize the historical complexity. The only diagram that I know of that shows you the full network is one provided by Pop Chart (The Genealogy of Wine), a group that provides infographic posters for just about anything. They will sell you a full-sized poster of the pedigree (3' by 2'), but here I have provided a simple overview (which you can click on to see somewhat larger).

You can actually zoom in on the diagram on the Pop Chart web page to see all of the details. This allows you to spend a few happy hours finding your favorite varieties, and to see how they are related.

Perhaps the most interesting thing to note is just how many modern varieties have what the French call Gouais Blanc as one of the parents. This white variety is actually an indigenous Croatian variety known there as Štajerska Belina. Try to remember this the next time you drink a glass of Chardonnay or Gamay, both of which are crosses between this variety and Pinot Noir.

Why are there so many missing bits?

Reconstructing grape genealogies is often a tricky business, even with genetic data. Originally, this was done using morphological characters (the study of grape morphology is called ampelography), along with whatever historical records exist. However, these days we use DNA from whatever varieties and cultivars are available for sampling (see my post Grape clones and varieties are not always what they seem). Perhaps the biggest problem is that many of the cultivars are no longer known (there have been at least 10,000 of them recorded at some time in history), so that the pedigree is full of question marks representing unknown (unsampled) parents.

This has an odd practical consequence, which is that the time direction of the pedigree will be ambiguous whenever there is a missing parent. That is, we cannot work how which is the parent and which is the offspring among closely related varieties. Technically: estimates of what geneticists call identity-by-descent (IBD) are calculated based on the technique of linkage analysis for all pairwise comparisons of samples. However, complex inter-breeding schemes can generate IBD values such that we cannot distinguish parent-offspring relationships from sibling (brother-sister) relationships.

There are many examples of this problem in the study of the grape genealogy. As but one example, there is an online page associated with the book Wine Grapes (by Jancis Robinson, Julia Harding and José Vouillamoz). One of the web pages shows some of the pedigree diagrams from the book, and on this page there are three possible pedigrees for Syrah. The current genetic data cannot distinguish between these three possibilities (shown here).

This web page also happens to illustrate another aspect of drawing complex pedigrees. Sometimes, in order to avoid drawing intricate networks, and to make the diagram look more like a "family tree", authors resort to a standard subterfuge. The strategy is to show varieties multiple times in the diagram, to avoid drawing reticulate relationships (see my post on Reducing networks to trees, where I illustrate this using the pedigree of Zeus). This situation occurs in the pedigree for Pinot Noir (the authors note: "For the sake of clarity, Trebbiano Toscano and Folle Blanche appear twice in the diagram").

The smallest global wine production for 55 years?

The International Organisation of Vine and Wine (OIV) has recently indicated that the 2018-vintage global wine production is "back to normal" in size (Global wine harvest bounces back).

However, a year ago the OIV reported that the 2017 global wine production had fallen to its lowest level since the beginning of the 1960s. This decrease was the result of the inclement weather that extensively damaged vineyards in the world’s largest production area, western Europe (containing the world's three biggest producers: Italy, France and Spain).

As always, I wanted a picture of what it looks like to have "the lowest production for 60 years". Is this decrease only a little bit lower than the past few decades or a lot lower? A graph would obviously answer this for me. However, finding this graph turned out to be much more difficult that I thought. I expected that somewhere there would be a graph of annual global wine production, stretching back for decades, but I turned out to be badly wrong.

The trusty Statistical Compendium presents the production data averaged across each decade, only (this was graphed on the AAWE Facebook page after I started writing this post); and the OIV online data go back only to 1995. At least the FAO online data go back to 1961, although it is measured in tonnes of grape must not liters of wine.

Production

Eventually, Kym Anderson (of the University of Adelaide) kindly pointed out to me that the spreadsheet associated with the Statistical Compendium does actually have the annual data, all the way back to 1860. So, here is the graph that I have been looking for, with global wine production shown vertically (a hectoliter is 100 liters — for comparison, a dozen standard bottles of wine is 9 liters), and time proceeding horizontally. The pink line shows the 2017 production level compared to recent decades.

Obviously, there was a slow increase in production until World War II, with a big dip in production during the war itself. In contrast, there was a dip for only one year during World War I, although that year (1915) had the lowest production after 1863. After the second war, there was a rapid increase in production until 1980, followed by a decline until 1990. Production has remained steady since then.

As far as the 2017 production is concerned, all production levels prior to 1962 were lower (ie. 55 years earlier). However, both 1991 and 2002 came close to the 2017 level.

This result differs somewhat from the OIV reports, which originally agreed with this but then changed their previous low year to 1957 (eg. World wine output falls to 60-year low). However, the OIV values are an average of 5 million hectoliters greater than the values shown here.

Consumption

This graph leads to the obvious question about wine consumption. After all, we have been told for several decades about a so-called "wine lake" caused by world-wide over-production. Indeed, in the past the EU paid to have excess wine converted to spirits, and more recently it has been paying grape-growers to pull up vines, especially in southern France. The government of South Australia tried the same vine-pull tactic back in the late 1980s.

Here is the graph that compares consumption with production, as recorded in the Statistical Compendium spreadsheet.

This shows that consumption kept close pace with production until 1963, although the relevant consumption is usually the year (or two) after production (since the wines do not appear on the shelves until then).

After 1963, production has almost always exceeded consumption. This is the origin of the wine lake — indeed, in some years production far exceeded consumption. Both consumption and production dipped towards the end of last century, which is when the various governments started to address over-production. Since then, consumption has slowly increased again, although production has continued to exceed consumption. However, the two levels were very close in both 2012 and 2017.

Supply and demand caused concern in 2013, when it was reported by Morgan Stanley that "an excess of 600 million unit cases (almost one quarter of global consumption) in 2004, [was] reduced to just 1 million unit cases in 2012." At that time, it was suggested that "there's little indication that world production can keep pace with the oenophilic hordes."

The production levels of 2013–2016 eased this situation. However, the low production in 2017 has once again led to reports of a sharp decline in French wine inventories.

Welcome to the wine regions of Australia

A few weeks ago, I noted that the wine-making regions of Australia are rather poorly known, except perhaps in Australia (How aware are wine drinkers of wine regions?). This blog post is an attempt to at least partially redress this limitation.

Where they are

Let's start with the geography. I showed a map in an earlier post illustrating that Australia is the same size as the continental USA, and much larger than western Europe (The rise, rise, fall and rise of Australian wine). However, the wine-making regions are restricted to a rather small part of the continent, mainly the south-east but also part of the south-west, as shown in the map above.

The reason for this is explained in another post (Growing degree-days maps of Australia and the western USA). For most of Australia, the heat summation is way too high — this is the sum of the daily temperatures above 10 °C (50 °F) during the grapes' growing season. Only around the southern coast is the summation suitable for grape vines (see the map further down this post), although in the high-altitude south-eastern part the vines are absent because the heat summation is too low.

Most of the suitable vine-land area of Australia already has formally named Geographic Indication (GI) viticultural areas. Most of these are shown in the next map — you will need to click on it to see the details.

There is also the largest recognized region of all: South-Eastern Australia, which takes in the whole of New South Wales, Victoria and Tasmania, plus those regions of Queensland and South Australia in which grapes are grown (see the next map) — this includes everything except the vineyards in Western Australia. This is the most likely region to appear on labels for bulk Australian wine, which indicates that the grapes come from several of the large inland-irrigated grape-growing areas. However, it can also mean that the wines are blends from more than one of the smaller named regions.

If you want to peruse more detailed maps of the GI areas, then they are available on the Wine Australia web site (the map links are near the bottom of that page).

What types of regions are they?

Wine Australia produces a glossy booklet describing some of the more prominent regions — Discover Australia Wine: Regions and Varieties. These GIs include (arranged by state):

South Australia:
Adelaide Hills
Barossa Valley
Clare Valley
Coonawarra
McLaren Vale
Victoria:
Mornington Peninsula
Rutherglen
Yarra Valley
New South Wales:
Canberra District
Hunter Valley
Western Australia:
Margaret River
Tasmania

Note that, at the moment, the state of Tasmania is a single GI zone. This is Australia's coolest region, and the focus of a burgeoning sparkling-wine industry. There are seven quite different grape-growing areas within the state (see Regionality: the future of Australian sparkling), and it is thus quite likely that it will soon be sub-divided.

Australia has areas within all of the Winkler Index zones, based on heat summation, from Cool (I) through to Hot (V). Therefore, all wine types can be produced in at least one of the areas, and usually in several of them. The GI regions are shown in red on the map below, which also color-codes the five Winkler Index zones from blue (coolest) to orange (warmest).

The best-known region internationally seems to be the Barossa Valley, which makes warm-climate Shiraz wines that are very unlike the Syrah wines of the rest of the world. However, this fame seems to date mainly from the discovery by Robert Parker, in the 1990s, that these wines suit his palate particularly (within Australia, the move to prominence started in the mid-1980s). While the uniquely Australian Shiraz and Shiraz-Cabernet blends have taken the limelight, many of the wines actually have Mataro (Mourvèdre, Monastrell) also present, which has been referred to as "the secret ingredient of the Barossa reds", whether it is mentioned on the label or not (see A dark horse with exceptional pedigree).

By way of historical contrast, back in 1974 the wine merchant Dan Murphy published Classification of Australian Wine, in which book he noted: "Although no wine authority has ever written off the Barossa Valley as a top quality area, I cannot remember any one of them giving it a special accolade as a constant producer of world class table wines." Thirty years later, the wine writer James Halliday, in the Wine Press Club of NSW Annual Lecture for 2005, pointed out that the wines of the Barossa Valley usually don't win Australian wine-show trophies, either.

The Barossa is actually part of a larger collection of GIs that follow a range of hills (the Mount Lofty Ranges) that run north-south through the state of South Australia (see the next map). This is sometimes treated as the epicenter of Australia's international wine reputation, at least for those who like Parker-style wines. The various regions included are described in The 13 Wine Regions of the Adelaidean – Mount Lofty Ranges.

In spite of the reputation of this region for foreigners, the Australian focus in the past couple of decades has been on the development of cool-climate wine regions, especially for white wines. Most of these regions are in the states of Victoria and New South Wales (see the Winkler map above), and especially the emerging region of Tasmania. These areas produce wines that are rarely to the taste of Robert M. Parker Jr (for example, see the 2005 kerfuffle over the cool-climate Mount Mary Quintets wine — Sharp differences of opinion over Mount Mary; and The differences of opinion continue: Mount Mary Quintets vs. Robert M. Parker Jr. part 2).

Another important point to note is that some of the traditional wine companies produce multi-regional blends made to a particular style, rather than having a terroir-driven set of wines. The classic example is Penfolds, the premier brand of Treasury Wine Estates. Of their top 10 wines, 7 are multi-regional blends (and thus list "South Australia" or "South-Eastern Australia" as their geographical indication). These style-driven wines are not necessarily to everyone's taste, of course; and many people (including myself) suspect that they are aimed at the wine-investment market rather than the wine-drinking market.

For further reading, there is one book specifically describing the wine-making regions of Australia, but this now more than 15 years old, having been first published in 1999 and then updated in 2002. This is John Beeston's The Wine Regions of Australia. Alternatively, there is the slightly more general Wine Atlas of Australia, by James Halliday, with editions in 1992, 1998, 2006 and 2014.

It is also worth noting that, in order to understand the nature of the regions in as much detail as possible, it was recently announced that: "The vineyards in Australia’s 65 wine regions will be accurately mapped for the first time using high-resolution satellite images and advanced machine learning, in a national census of Australia’s wine-grape area" (AI to map Australia’s 65 wine regions row-by-row). This would be a first in the New World, to compete with the detailed maps of the wine-making regions of Europe.

Wine tourism

Tourists in wine regions are a big deal in Australia. Marlene Pratt has an interesting description of the various types of wine tourists in Australia (wine lover 15%, wine interested 55%, wine curious 17%, plus 12% along for the ride). When I was young, you could simply arrive at almost any winery during business hours and someone would look after you. However, these days about one-third of wineries charge a small fee, which will be deducted from any purchase you make.

Earlier this year, Wine Australia conducted a cellar-door and direct-to-consumer survey, which showed that cellar doors are the driving force behind direct-to-consumer sales in Australia, accounting for 44 % of direct-to-consumer revenue, ahead of wine clubs and mail orders (Wine Australia release full survey findings on Cellar Door direct wine sale opportunities).

As for which regions are the most popular to visit, The Wine Front web site conducted an online poll from March 2017 until now, asking: "What’s the best Australian wine region to tour (ie. from a pure wine touring perspective)?" The next graph shows you how the 476 voters responded.

This list is not necessarily based on the quality of the wines, since wine tourism and wine quality are not always closely related. To check this, we could compare this list to one for the number of classified wines in Langton's Classification of Australian Wine, which is based on the prices of the wines at auction. This is shown in the next graph.

Note that the regions do swap about a bit in their rank order between the two graphs, but they also do have a great deal of similarity — good regions to tour do generally produce good wine. The Langton classification is dominated by wines from the two regions immediately north of Adelaide (the Barossa Valley and the Eden Valley), but these do have one major limitation for tourists — they are pretty much closed on Sundays, unlike most of the other regions. Here is David Farmer's comment (Thoughts on Margaret River):

Margaret River for my money is now the most exciting wine region in Australia, both for the wine tourist and the fine wine buyer. The beaches and ocean scenery, the wealth of winery restaurants, cheese and boutique breweries, plus excellent scenery, makes it unbeatable. My home town of Tanunda in the Barossa Valley looks very dreary indeed, as does the Barossa as a tourist destination.

Lonely Planet has very recently released a tourist guide called Wine Trails: Australia & New Zealand, covering 30 of the wine regions (plus 10 from New Zealand).

Conclusion

Dan Murphy, in the book quoted above, also referred to: "the world of Europe and the Americas, the inhabitants of which, oddly as it may appear, seem to think that the vineyards of Australia are far more distant from them than we think the Americans are from us." So, if you ever happen to be in Australia, you should certainly check out some of these regions. In the meantime, you could do a lot worse than try some of their wines, instead.

How many grape clones are there per variety?

I recently noted that the twin ideas of "varieties" and "clones" are basic to grape-growing, whether they are used for wine-making, table grapes, raisins, juice, jelly, canning, or rootstocks (Grape clones and varieties are not always what they seem). This leads me to ask the question posed in the title.

The number of clones is often related to the genetic mutation rate of the variety. Pinot noir is therefore recognized as one of the most mutable grape varieties. Indeed, it is so mutable that we have produced several different clones that we actually treat as being varieties, because they can be used to make very different wines (as discussed in the previous post). Indeed, Pinot noir itself also has more recognized clones than do most other varieties.

This is illustrated in the first graph. The data come from the list of varieties and clones currently approved for viticulture in France (Catalogue of vines grown in France). Each grape variety is listed horizontally, with the number of approved clones counted vertically. Only the 38 varieties with at least seven clones are shown — there are a total of 269 approved varieties, and 950 clones, although 161 of the varieties (60%) have only one listed clone.

As you can see, Pinot noir is way out in front. Its history is, of course, associated specifically with Burgundy, although it is now an international variety. It is therefore interesting to note that the second variety in the graph is Gamay, the red-wine grape from Burgundy's nearest neighbor, Beaujolais. The people of this part of France have had a long time to keep an eye out for potentially useful mutants among their vines.

We can, of course, separate the graph into both red- and white-wine grapes, and this is illustrated in the next graph, which shows the 30 varieties of each type with the most clones. It is immediately obvious that red-wine varieties tend to have more clones than do white-wine varieties. Presumably, this reflects the historically greater interest that has generally been shown in producing "fine" red wines compared to white wines.

As usual in this blog, we can search for simple patterns among these data. Mathematically, the data from the first graph most closely fits a Power model, as shown in the following graph, where the added line shows the model. The most obvious place the model does not fit the data is the first two varieties — both Pinot noir and Gamay actually have fewer clones than "expected" from the model!

If we fit the Power model to the red and white varieties separately, as shown in the next graph, then it is interesting to note that the model fits the white varieties very well, although the first white-wine grape (Chasselas) has too few clones. However, the model fits the red-wine varieties rather more poorly, and especially Pinot noir and Gamay have many too few clones for this model . The latter results from the greater number of clones for the remaining red varieties — red-wine grapes generally have more clones than we would "expect" from a simple mathematical model.

So, we may conclude that white grape varieties have clone numbers that fit a simple mathematical model, but red varieties have more clones than can be "explained" by any such model. This is likely to be because red wines have been preferred by consumers, so that the vignerons have been more keen to find "better" clones for lots of the red varieties but only a few of the white varieties.

Some comments

Lest anyone thinks that there is some restriction on the number of grape varieties that can be used in France, it is important to note that new varieties are being approved all the time (eg. for 2018: New grape varieties in France, some forgotten, some hybrids). The matter of hybrids is of particular interest, because they have long been detested in France but are now being seriously considered as a solution to viticultural problems like downy mildew (see French wine's hybrid dilemma).

However, the situation in France is quite formal:

Grapevine clonal material in France is subjected to extensive testing and certification under the aegis of the Ministry of Agriculture. The Institut Français de la Vigne et du Vin (IFV) is a national agency certified to manage and coordinate the French national clonal selection program. The responsibilities now entrusted to the IFV were formerly performed by an entity known as the Etablissement National Technique pour l’Amelioration de la Viticulture (ENTAV), and include maintenance of the French national repository of accredited clones, and administration of the ENTAV-INRA authorized clone trademark to protect the official French clones internationally.

The list of approved French cultivars is important, because only cultivars from this list can be used in wines given legal protection under the Appellation d'Origine Contrôlée (AOC) scheme administered by the Institut National de l'Origine et de la Qualité (previously INAO). Recently, INAO has allowed other varieties from the list to be used in AOC wines as "grape varieties for climate and environmental adaptation", which means that issues such as climate change can be addressed by trialling new varieties.

There has been some concern expressed about restricted use of clones in many vineyards, especially outside Europe (eg. Genetic instability and how it affects wine). In the past, vineyards tended to be a mosaic of clones, but modern agriculture (not just for grapevines) has shown a strong move towards using single clones — those that are seen as being "the best" in some ineffable way. For example, Cabernet sauvignon clones 07 and 08 have been the basis of many of the California Cabernet vineyards for decades (see The guide to wine grape clones).

However, this approach is asking for trouble — when problems occur in the field they are likely to affect all of the plants at the same time, because the clones are genetically identical. Resilience to the vagaries of different seasons (especially weather, pests) is based on diversity, not lack of it. As has been noted (Myles et al., Proceedings of the National Academy of Sciences of the USA 108:3530-3535):

Vegetative propagation [has] discouraged the generation of unique cultivars through crosses. The grape currently faces severe pathogen pressures, and the long-term sustainability of the grape and wine industries will rely on the exploitation of the grape’s tremendous natural genetic diversity.

To this end, of the known varieties, 7,200 are grown at the Centre de Ressources Biologiques de la Vigne (CRB-Vigne) de Vassal-Montpellier, near the town of Marseillan in southern France.

Perhaps at issue here for the growers is that some clones will produce mediocre wines when grown in certain regions but not others. For example, it has been noted that the original Chardonnay clones used in Oregon came from California, where the climate is warmer and drier, and that this situation was much improved when clones were imported directly from Dijon in Burgundy, where the climate is more similar (see The role of grape clones in planting grapes)

Also, there is the matter of intellectual property rights — the use of some clones requires the payment of royalties (see How wine clones are already taking over). The Foundation Plant Services Grape Registry currently lists 594 grape varieties used for all purposes in the USA, which is more than in France, but many of these are proprietary. The latter do include, of course, the ENTAV-INRA authorized clones.