Report

Historical Genetics: The Parentage of Chardonnay, Gamay, and Other Wine Grapes of Northeastern France

See allHide authors and affiliations

Science  03 Sep 1999:
Vol. 285, Issue 5433, pp. 1562-1565
DOI: 10.1126/science.285.5433.1562

Abstract

The origins of the classic European wine grapes (Vitis vinifera) have been the subject of much speculation. In a search for parental relationships, microsatellite loci were analyzed in more than 300 grape cultivars. Sixteen wine grapes that have long been grown in northeastern France, including ‘Chardonnay’, ‘Gamay noir’, ‘Aligoté’, and ‘Melon’, have microsatellite genotypes consistent with their being the progeny of a single pair of parents, ‘Pinot’ and ‘Gouais blanc’, both of which were widespread in this region in the Middle Ages. Parentage analysis at 32 microsatellite loci provides statistical support for these relationships.

The wines of northeastern France, notably those of the Burgundy and Champagne regions, have been highly regarded for centuries. Like most of the world's finest wines, they are made entirely from old cultivars of Vitis vinifera L. The cultivars most strongly associated with this part of France are ‘Pinot noir’ and ‘Chardonnay’, which are used both for Champagne (1) and also for the best red and white wines, respectively, of the Côte d'Or in the heart of Burgundy. These two grapes are now grown in many of the world's wine regions. In the southern part of Burgundy, the red wines of Beaujolais are made primarily from ‘Gamay noir’. Several other cultivars, including ‘Aligoté’, ‘Melon’, and ‘Sacy’, are used in wines carrying regional Burgundy appellations.

Grapevines are propagated vegetatively, so that the individual vines of a cultivar are genetically identical to each other (except for somatic mutations) and to the single original seedling from which the cultivar originated. While some cultivars may have originated in the regions with which they are now associated, others are thought to have been introduced by traders or conquerors, most notably the Romans. Although a few varieties have been produced by controlled crosses since the mid-1800s, most V. vinifera cultivars in existence today are centuries old and are thought to have arisen by several mechanisms—domestication of wild vines, spontaneous crosses between wild vines and cultivars, and spontaneous crosses between cultivars (2). Such a spontaneous cross between two cultivated varieties gave rise to ‘Cabernet Sauvignon’, the most important cultivar of Bordeaux and arguably the most highly regarded red wine grape in the world today (3).

Any wild vines that were parents of today's important wine grapes cannot be identified, because they no longer exist. Parents that are themselves cultivars, however, may still exist in collections if not in cultivation. In a search for the parents of some important French wine grapes, we analyzed 322 cultivars of V. vinifera(4), including most extant old French cultivars.

Samples of 51 cultivars were obtained from the vineyards at the University of California at Davis and the rest were taken from the variety collection maintained by Institut National de la Recherche Agronomique at Domaine de Vassal, near Montpellier, France. After an initial screening of all cultivars at 17 microsatellite loci (5), we compared microsatellite alleles within all possible sets of three to identify pairs of cultivars that could have contributed the alleles of the third cultivar (6). A subset of cultivars was then further analyzed at 15 additional loci (7).

On the basis of 32 loci, 16 cultivars had microsatellite genotypes consistent with their being the progeny of a single pair of parents—‘Pinot’ and ‘Gouais blanc’ (Fig. 1) (8–10). For each of the 16 putative ‘Pinot’ × ‘Gouais blanc’ progeny, we calculated parentage indices to compare the probability of the observed progeny alleles if it had the putative parents to the probability of those alleles, if it had two random parents, or if the parents were close relatives of the putative parents. We show the detailed parentage indices for ‘Chardonnay’ inTable 1 and summaries for all 16 progeny cultivars in Table 2. The likelihood ratios show that the putative parents are 1012 to 1015 times more probable than two random parents and are 447 to 28,000 times more probable than either ‘Pinot’ or ‘Gouais blanc’ relatives.

Figure 1

Inheritance of parental microsatellite alleles by progeny cultivars for locus VVMD5 (top) and VVMD28 (bottom). Microsatellites were amplified from genomic DNA, electrophoresed in polyacrylamide gels, and visualized by silver staining. Lanes represent (from left to right) (P) Pinot noir, (G) Gouais blanc, (1) Aligoté, (2) Aubin vert, (3) Auxerrois, (4) Bachet noir, (5) Beaunoir, (6) Chardonnay, (7) Franc noir de la Haute Saône, (8) Gamay blanc Gloriod, (9) Gamay noir, (10) Knipperlé, (11) Melon, (12) Peurion, (13) Romorantin, (14) Roublot, (15) Sacy, (P) Pinot noir, and (G) Gouais blanc.

Table 1

Detailed parentage analysis for ‘Chardonnay’ (C) and its presumptive parents ‘Pinot’ (P) and ‘Gouais blanc’ (G). The likelihood of the progeny alleles for the putative parents versus alternative possibilities was calculated as in (3) and (21). For each locus, the likelihood ratio is the quotient of the probability of the progeny genotype if it had the presumptive parents and the probability of that genotype if it had the alternative parents. Observed allele frequencies are based on 322 genotypes. The calculations were also performed with upper 95% confidence limits of allele frequencies to compensate for sampling error.  

View this table:
Table 2

Summary of parentage analyses for 16 putative progeny cultivars of ‘Pinot’ (P) and ‘Gouais blanc’ (G). X and Y denote random unrelated cultivars. The likelihood ratio shown for each possible parentage for each progeny cultivar is the cumulative product of the ratio for each of 32 loci. Ratios are calculated from 95% upper confidence limits of allele frequencies.

View this table:

The origins of these varieties have been the subject of much speculation. While most are thought to be French, foreign origins have been suggested for some. ‘Chardonnay’, for example, has been proposed as coming from the Middle East, ‘Gamay noir’ from Dalmatia and ‘Sacy’ from Italy (11).

Because some of these varieties are morphologically similar, some authors have suggested that they are close relatives. The varieties are indeed related, but none of the specific relationships they proposed are the ones revealed by our data. In the 17th century, for example, ‘Aligoté’ was suspected to be a seedling of ‘Gouais blanc’ and ‘Chardonnay’. ‘Gamay blanc Gloriod’ was thought to perhaps be a seedling of ‘Gamay noir’. ‘Chardonnay’ was considered to be a white form of ‘Pinot noir’ (11). Based on their observations that wine cultivars historically associated with a region often share morphological similarities, several authors have proposed ecogeographic groupings for French wine grapes (12), some of which include varieties discussed here.

It is no surprise that these 16 varieties are the progeny of ‘Pinot’. No variety is thought to have been in Burgundy for a longer time, and it also has a long history in other parts of northeastern France. References to a variety that was probably ‘Pinot’ go back as far as the Roman agricultural writer Columella in the first century A.D. ‘Pinot’ may already have been present in the Burgundy region at the time of the Roman conquest (11).

The surprise is that the second parent is ‘Gouais blanc’, a variety considered so mediocre that it was banned (unsuccessfully) at various times in at least two regions and is no longer planted in France. The name “Gouais” derives from the old French adjective “gou,” a term of derision. In the Middle Ages, this variety was very widespread in northeastern France and probably occupied the majority of the vineyards, but only the mediocre sites, the better sites being reserved for more noble varieties such as ‘Pinot’ (11).

We believe that ‘Gouais blanc’ is not indigenous to northeastern France but was introduced from elsewhere because it is the same variety as ‘Heunisch weiss’, once widely grown in Eastern Europe (13). The third century Roman emperor Probus, a Dalmatian, encouraged viticulture in the provinces and is said to have given the Gauls a grape from his homeland. While it has been suggested that this grape was ‘Gamay noir’ (11), there is no evidence that this variety was grown outside of France until relatively recently, so perhaps Probus' gift to the Gauls was ‘Gouais blanc’.

We do not know how many separate crossing events produced these progeny but it is likely that each was produced at a different time and place. Historical references to the individual progeny cultivars point to beginnings in different locales ranging from the easternmost Loire Valley to Champagne to Alsace. Both parents were widespread in northeastern France and could have been in close proximity in all of these places. ‘Pinot’ was also grown in other parts of Europe in the late Middle Ages, so that crosses between the two parents could have occurred in places other than France. However, the progeny discovered in this study are all historically associated with northeastern France and not with other places, suggesting that the crosses occurred here. Other ‘Pinot’ × ‘Gouais blanc’ progeny may well exist in other parts of Europe.

Several of the progeny can be traced back to the early Middle Ages or before (‘Chardonnay’, ‘Gamay noir’, ‘Melon’). The others are mentioned in literature from 100 to 400 years ago. Several progeny varieties that were first described only in the late 19th century (such as ‘Gamay blanc Gloriod’ and ‘Peurion’) may have existed before this time. Although ‘Gouais blanc’ would have been a more probable parent earlier, when it was more widespread, even a relatively rare crossing event in the 19th century could have produced successful progeny cultivars. ‘Pinot’ and ‘Gouais blanc’ are clearly a good parental combination, whereas any other varieties growing in the vicinity would likely be ‘Pinot’ or ‘Gouais blanc’ relatives and would be less fit as a consequence of inbreeding depression.

Nine of the progeny cultivars have light-colored fruit (white-yellow to amber), four are blue-black or violet-black and two have intermediate pink berries. ‘Gouais blanc’ has yellow-gold berries. The parent of the dark-berried cultivars (‘Bachet noir’, ‘Beaunoir’, ‘Franc noir’, ‘Gamay noir’) must have been ‘Pinot noir’. But because ‘Pinot noir’ is heterozygous for berry color, we cannot say whether ‘Pinot noir’, ‘Pinot gris’ or ‘Pinot blanc’ was the parent in the other cases. However, ‘Pinot noir’ has always been the most common form so that it is the more probable parent.

Although ‘Gouais blanc’ has never been highly regarded and is maintained in only a few grape variety collections today, its genetic potential as a parent has been clearly demonstrated here, particularly in the case of ‘Chardonnay’, considered one the world's greatest wine varieties. Knowledge of parental relationships such as those reported here can facilitate rational decisions regarding the size of grape germplasm core collections, which are constantly threatened by economic constraints. A core collection containing ‘Pinot noir’ and ‘Gouais blanc’ will contain the same allelic resources as a larger collection that also includes cultivars that are the progeny of these two.

Although most grape cultivars are hermaphroditic and largely self-pollinating, grape is intolerant of inbreeding (14). In searching for parental relationships among more than 300 cultivated grape varieties, we did not find any that are the selfed progeny of another cultivar (15). Furthermore, despite their geographic proximity over a long period of time, none of the ‘Pinot’ × ‘Gouais blanc’ progeny cultivars seem to have produced any successful progeny. Our results suggest that heterosis has played a significant role in the emergence of successful wine grapes. ‘Gouais blanc’ is genetically relatively dissimilar to ‘Pinot’, sharing only 20 out of 64 alleles at the 32 loci studied, consistent with an eastern European origin for ‘Gouais blanc’. The large number of successful progeny these two parents have produced may be a consequence of their genetic distance (16). Modern grape breeding programs might benefit from the use of comparably distant parents.

  • * Present address: Plant Genome Mapping Laboratory, University of Georgia, Athens, GA 30602, USA.

  • To whom correspondence should be addressed. E-mail: cpmeredith{at}ucdavis.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article