Back in Ancient Rome, Pliny the Elder wrote (in his Naturalis Historia):
Democritus, who has declared that he was acquainted with every variety of the grape known in Greece, is the only person who has been of opinion that every kind could be enumerated; but, on the other hand, the rest of the authors have stated that they are quite innumerable and of infinite extent, an assertion the truth of which will be more evident, if we only consider the vast number of wines. I shall not attempt, then, to speak of every kind of vine, but only of those that are the most remarkable, seeing that the varieties are very nearly as numberless as the districts in which they grow.*
Ancient Greece was the origin of grapes cultivated to make wine, as far as the Romans were concerned. So, Pliny was suggesting, in part, that the number of grape varieties is associated with the lengthy time over which they have been cultivated.
However, he is suggesting something else, as well. It is also possible that the number of grape varieties is associated with the amount of land that is used for that cultivation. That is, the more area available, and thus the more micro-climates, then the more likely it is that different varieties will be grown (and thus different wines made). Pliny is thus considered to be the originator of the idea of terroir in wine-making.
I thought that I might look at these two things, together. As a source of data, I have used the Organisation Internationale de la vigne et du vin, which lists countries with vineyards larger than 50,000 hectares in 2020 (with 2016 as the latest available data for smaller areas). Since my question is about wine-making not grape-eating, I have excluded (in decreasing order of grape-growing area): China, Turkey, Iran, India, Uzbekistan, Afghanistan, Egypt, Algeria, Morocco, Tunisia, and Turkmenistan. I have used all other countries, except that the data for the number of varieties are missing from South Africa.
The remaining data are plotted in the first graph. Each dot represents one wine-making country, located with the number of reported grape varieties shown vertically. The horizontal axis represents the recorded vineyard area (in hectares) — note that the scale is logarithmic.
This might look like a bit of a mess; but, in fact, it illustrates both of the points made above: the number of varieties can be influenced either by cultivated land area or by length of cultivation.
To see this more clearly, I have annotated the graph for you, in the next figure. I have added labels for some of the countries, and a summary line illustrating the relationship between number and area.
Let's take the line first. This illustrates a strong logarithmic relationship for 15 of the countries (ie. excluding Georgia, New Zealand, Armenia, Moldova, Chile, Spain). It indicates that whenever we double the vineyard area then we get c.80 more grape varieties (technically, this accounts for 70% of the between-country variation in the number of varieties).** So, we may safely conclude that, yes, the number of wine-grape varieties often increases with the amount of land that is used for their cultivation.
Note that Italy and France are at the top of the line, with the greatest number of recorded grape varieties (454 and 383, respectively). It has often been claimed that Pliny also commented to the effect that Italy has more kinds of wine grapes than there are grains of sand on a beach. Alternatively, it has also been noted that the Roman poet Virgil said: “It would be easier to measure the grains of sand in Greece, despite the different varieties of grapes.”
Greece is no longer in the picture as the cradle of wine-making. This role is now acknowledged to have occurred at least 8,000 years ago, in the area now occupied by Georgia and Armenia, which are much further east. You can see the result of this in the graph, as they both have a large number of grape varieties even though these days they do not have a lot of grapevine area. (Aside: in Georgia the vineyards’ share of the total agricultural crop area is greater than in any other country.) Most of these varieties have been there for thousands of years, but are not ones in common use elsewhere. In Georgia, for example, the two most common grapes are the whites Rkatsiteli and Tsolikouri, followed by the red Saperavi — when was the last time you saw these on the label of a bottle of your wine (U.S. wine imports from Georgia and Armenia)?
That accounts for two of the exceptions to the area / number relationship — ancient history is important. New Zealand’s placement in the graph is probably explained by recent history, instead — experimentation in response to climate change, trying new varieties to see which ones work in the new climate regime. This is also being done in Australia, and also the USA. However, the graph indicates that it is not (yet) being done in Chile, the only other New World country that does not fit the area / number relationship. In the Old World, the current trend seems to be more towards re-purposing the current varieties, rather than trying new ones.
This leaves Spain as the remaining odd-one-out. Spain is well known for producing oodles of bulk wine from a few high-yielding varieties, much of which ends up in France, to help bolster what is labeled “EU wine”. This may be slowly changing, especially in areas such as Rioja. Nearby, Portugal is a very different place. In the graph, it is the unlabeled point furthest above the line. It has only 20% of the vineyard area of Spain, and yet it has 2.2 times as many grape varieties — this is much more “normal”, according to the graph.
* This same book is usually credited as the source of “...volgoque veritas iam attributa vino est.” [It has become quite a common proverb that in wine there is truth.]
** Remember: this relationship applies among the points, not outside them. It would, for example, be naive to think that doubling from 10 Ha to 20 Ha would add 80 varieties!