Genetic Relationships Among Pinots and Related Cultivars

Genetic Relationships Among Pinots and Related Cultivars FERDINAND REGNER 1., ALEXANDRA STADLBAUER 2, CORNELIA EISENHELD 3, and HERWlG KASERER 4 Genetic analysis based on more than 30 SSR (= "simple sequence repeats") markers made it possible to investigate several cultivars of the Pinot family and to describe the relationship between them and Pinotrelated cultivars. The differentiation of Pinot blanc, Pinot gris and Pinot noir and different clones of them by SSR markers failed. Therefore it can be assumed that all Pinots are descended from the same genotype with only minor genetic differences. On the other hand polymorphism of Pinot clones could be obtained by applying 120 RAPD markers. Additionally, it was found that the cultivars Pinot noir precocce (early maturing) and Blauer Arbst have the same SSR profile as Pinots, and therefore they represent different types of the same cultivar. The cultivars Schwarzriesling (Mellerrebe), Farbklevner and Samtrot differ from the Pinot profile at the same loci. Therefore the cultivar Samtrot could be differentiated from Pinot as well as from Schwarzriesling by SSR markers and therefore can be considered a hairless mutation of Schwarzriesling. The profile of Schwarzriesling is very closely related to that of Pinot and together with Traminer it could be proposed that it is the origin of Pinot. Several other Pinot-related cultivars can be assumed as Pinot offspring. It can be supposed that Saint Laurent is a seedling of Pinot with unknown crossing partner. Aligote, Auxerrois, Melon, and Chardonnay represent white-berried seedlings of Pinot. Heunisch, an ancient cultivar could be proposed as the second partner for these crosses. KEY WORDS: genetic analysis, SSR, RAPD, microsatellites, gene pool, multiplex-pcr The origin of the Pinots is not clear. Because of morphological characteristics, the relationship to wild types (Vitis sylvestris) is assumed [2]. Another hypothesis claims that Pinots were spread by the Romans and are identical to V. allobrogica [1]. The Carolinger king Karl III brought the variety 'Cl~ivner' (old spelling for Klevner) from the Burgundy region to the area surrounding Lake Constance in 884 [1]. The first description of'pynoz' (old spelling for Pinot) was probably that of Eustache Deschamps in the 14 th century [8]. The age and importance of the Pinot group is indicated by numerous synonyms, definitions in most European languages, and many different types and clones. According to the current doctrine [1], the Pinot group consists of genuine Pinots (Pinot noir, Pinot gris, Pinot blanc), the Pinot noir mutant Samtrot, genotypes with slightly differing phenotypes such as Schwarzriesling (M~illerrebe, Pinot Meunier), Blauer Arbst and irregularly designated Pinots [1] such as Saint Laurent, Teinturier, Auxerrois, Chardonnay, and others. These misinterpreted Pinots [8] have since been verified as individual cultivars by ampelographic methods despite their morphological similarities to Pinot. An important step in differentiation of the Pinots was taken in 1958, when Chardonnay and Pinot blanc could be clearly differentiated as individual cultivars [9]. Despite clear ampelo- 1.2,3,4 HShere Bundeslehranstalt und Bundesamt for Wein- und Obstbau Klosterneuburg, Wienerstral3e 74, A-3400 Klosterneuburg, Austria. * Corresponding author [Fax: 0043-2244-29554, e-mail <reb.hblawo@eunet.at>]. Acknowledgements: We would like to thank the colleagues from the grapevine collections of Neustadt, Siebeldingen (Geilweilerhof), Geisenheim, Freiburg, Weinsberg, Senkvice and Pecs for providing independent reference samples of grapevine. We are grateful to J. Bowers and C. Meredith for the permission to use unpublished primer sequences. We would like to thank the colleagues K. Neumann and P. Schabner for reading over the manuscript. Manuscript submitted for publication 12 February 1999; revised 14 October 1999. Copyright 2000 by the American Society for Enology and Viticulture. All rights reserved. graphic descriptions, several misnamed Pinot cultivars could be found in commercial vineyards and some misidentified Pinots could even be found in germplasm collections. Molecular markers have been available for genotypic characterization for several years. While RFLP [3,4] and RAPD [10,11,14] markers offer only a limited use in genetic characterization, SSR (simple sequence repeats) markers are currently the most powerful tool [28] for characterization of genotypes and have advantages over RFLPs and RAPDs. AFLPs and cdna screening provide high resolution; however, distinction of cultivars and types as well as clones is not indicated in a clear manner [22]). RAPD analyses carried out with Pinots (noir, blanc, gris, and Meunier) did not allow successful differentiation [30] and was subject to problems with reproducibility [14]. Microsatellite markers are defined by short nucleotide repeats in the center region. Therefore mainly dinucleotide repeats (SSRs) [5,21,27] were used in this study. The repeat sequences effect highly polymorphic DNA fragments in different genotypes. Combining data of several polymorphic loci results in individual allelic profiles. The consistency of the results over time and different laboratories is given and data are comparable worldwide [12]. Therefore, microsatellite markers meet all the requirements for a reliable cultivar identification [28] and the determination of the origin of cultivars. Codominant inheritance of alleles allows us to follow segregation of specific alleles within a population and to understand the formation of gene pools and sometimes offers a view at evolutionary developments [12]. Microsatellite markers have been used for cultivar identification with grapevine for some years [15,26]. The six most polymorphic loci (VVS2, VVMD5, VVMD7, VVMD27, VRZAG62,

~ 8-- REGNER etal. VRZAG79) are sufficient to differentiate all grapevine cultivars analyzed so far [19]. Since more than 40 SSR markers have been developed by different laboratories [5,21,27], sufficient microsatellite loci are now available for the investigation of relationships between cultivars and their parentage. Accordingly, the theories about the origin of some cultivars, e.g., Mfiller Thurgau [15,20] had to be corrected. An hypothesis about the origin of Cabernet Sauvignon [6,20], Chardonnay, Silvaner, and several other cultivars [16] could be argued based on SSR alleles. This paper presents SSR data of several Pinots and Pinot-related cultivars. According to our findings, a hypothesis about the relationships of the Pinot family and related cultivars could be raised. The investigation is based on the analysis of more than 800 grapevine and rootstock genotypes including about 300 different cultivars and several Vitis species. Furthermore, RAPD analyses were performed in order to differentiate cultivars and clones. Differences among these genotypes illustrated the variability of Pinot clones. Materials and Methods The plant material (varieties and clones) used for this investigation was obtained from the collection of the HShere Bundeslehranstalt und Bundesamt ffir Wein- und Obstbau Klosterneuburg (Klosterneuburg, Austria). Independent reference samples were obtained from the collections of the Forschungsanstalt Geisenheim, Institut ffir Rebenzfichtung, Geilweilerhof, Staatliche Lehr- und Versuchsanstalt ffir Wein- und Obstbau Weinsberg (Germany), and from the collections of Pecs (Hungary) and Senkvice (Slovakia). DNA was extracted from young leaves according to Thomas et al. [26] and Regner et al. [18]. About 800 vine genotypes were analyzed by means of at least 12 SSR markers, and all possibly Pinot-related genotypes were analyzed with 34 SSR markers. The VVS (VVS1, VVS2, VVS3, VVS4, and VVS29) markers had been developed by Thomas and Scott [27] and the VVMD markers (VVMD5, VVMD6, VVMD7, VVMD8) by Bowers et al. [5] as well as Bowers and Meredith [personal communication]. Other VVMD markers include (VVMD14, VVMD17, VVMD21, VVMD24, VVMD25, VVMD26, VVMD27, VVMD28, VVMD31, VVMD32, VVMD36). The VRZAG (VRZAG7, VRZAG15, VRZAG21, VRZAG25, VRZAG29, VRZAG30, VRZAG47, VRZAG62, VRZAG64, VRZAG67, VRZAG79, VRZAG82, VRZAG83, VRZAGll2) had been developed in the course of our SSR investigations ofv. riparia [21]. The amplification of the SSR loci was performed according to the protocol of Smith et al. [23] specified by two-step cycles. General PCR protocol applied for these studies was two minutes denaturation at 94 C and 35 cycles with annealing phase for 30 seconds, (temperature between 45 C and 55 C), and denaturation for 15 seconds at 92 C. The annealing temperature for each locus was set according to the original protocols [5,20,24]. The final extension took place at 72 C for five minutes to stabilize the length of the fragments by adding an additional A base. Multiplex-PCR was successfully performed using loci with different size range. The reaction solution contained 20 pl of the buffer solution, which consisted of 16 mm (NH4)2SO4, 67 mm Tris-HC1 ph = 8.8, 1.5 mm MgC12, 0.01% Tween 20, 0.1 mm each dntp (GenXpress, Vienna) 0.2 ~M primer, 0.7 Units Biotherm Taq DNA polymerase (GenXpress, Vienna) and 50 ng genomic DNA of grapevine. Yield of DNA fragments was estimated by running an aliquot of the sample on a 2% agarose gel stained with ethidium bromide. The samples were denatured by heating in a formamide solution and were loaded together with a size standard (Genescan 350 Tamra, Appl. Biosys.) to a 6% polyacrylamide gel. Detection of the SSR fragments labelled with 6-FAM, HEX, and TET was carried out by an automated sequencer (ABI 373, Perkin Elmer, Vienna). Labelling with different fluorescent coloring agents facilitated the application of multiplex PCR. The fragment length of an allele is influenced by an internal equipment factor. This deviation, however, is stable and could be observed for all genotypes [19]. Using silver staining detection instead of an automated sequencing apparatus, we observed a shifting of one to three bases dependent on the microsatellite loci. Therefore, recently gained results show a constant deviation from prior published results [20]. The probability for a cross as the origin of a cultivar can be verified by means of likelihood ratio. The likelihood ratios of the proposed parentage were compared with those of other possible parent combinations at each locus. Cumulative likelihood ratios for the proposed parentages were calculated from the relative allelic frequencies (Table 1) in the 96 cultivars and their 95% upper confidence limits [6,20]. For each locus, the ratio of probability that the proposed vines are the putative parents was compared to the probability that two random cultivars are the parents. The cumulative likelihood ratio is the product of the ratio for each locus. The best score for confirmation of the proposed hypothesis is Table 1. Relative allelic frequencies and the 95% upper confidence limit of the frequencies derived from the genotypes of the 96 individuals at locus VVS2. Allele length Observed allele Upper 95% (bp) frequency confidence limit 129 0.005 0.004 131 0.01 0.005 133 0.3 0.023 135 0.042 0.01 137 0.1 0.015 139 0.05 0.011 141 0.005 0.004 143 0.18 0.02 145 0.015 0.006 147 0.01 0.005 151 0.21 0.021 153 0.026 0.008 155 0.042 0.01

GENETIC RELATIONSHIPS AMONG PINOTS- 9 a clear difference in likelihood ratio to alternative models. RAPD analysis was carried out using only true-totype genotypes as verified by SSR profiles. Decamer oligonucleotides were obtained from Operon Technologies, Alameda, USA (kit A1-20, C1-20, D1-20, El-20, and F1-20) and Metabion GmbH, Martinsried, Germany (BC:-301-302,-340,-349,-379; GTO:-3,-4,-5, GY:-60,- 103,-104-105,- 107,-109, M10, 05, O19, Q5). Amplification was performed in 20 ~tl of the buffer solution, which consisted of 16 mm (NH4)2SO4, 67 mm Tris-HC1 ph = 8.8, 1.5 mm MgC12, 0.01% Tween 20, 0.1 mm each dntp (GenXpress, Vienna) 0.2 I~M primer, 1 Unit Biotherm Taq DNA polymerase (GenXpress, Vienna), and 20 ng genomic DNA of grapevine. An Omnigene (Hybaid, GB) thermocycler processed 40 cycles of 30 seconds at 92 C, 90 seconds at 38 C and 60 seconds at 72 C. The arbitrarily amplified fragments were separated on a 2% agarose gel and detected by ethidium bromide staining. Results and Discussion RAPD analysis of Pinot: The real Pinot cultivars (Pinot noir, Pinot gris and Pinot blanc) [9] did not show any SSR polymorphism at any of the investigated SSR loci. Therefore, we agree to the previously attested 'close relationship' between the Pinots [29,30]. Due to the location of tandem repeat SSR markers in noncoding areas, these markers are not the proper tool to detect morphologically effective mutations such as different berry pigmentation. On the other hand, we could successfully determine polymorphism by genotyping different Pinot clones with RAPD markers. For this approach we used 118 RAPD markers to differentiate Pinot noir, Pinot blanc, and Chardonnay clones. Only a few primers could be selected that showed polymorphism. Remarkable polymorphism was found by applying Oligo-Pinot 1: CCCGATTCGG and Oligo-Pinot 2: GGTGCACGTT (Fig. 1). Using these primers to find polymorphism between Pinot blanc and Pinot noir, we observed that the frequency of polymorphism was not t 2 3 4 5 6 7 8 9 10 11 12 Fig. 1. RAPD fingerprints of Pinot noir clones and types obtained by amplification with Oligo-Pinot 2. The lanes contain the following Pinot noir types: 1 - genotype from the main collection of the HBLA; 2 - clone 543; 3 - clone Marienfeld; 4 - clone W210; 5 - clone 3/45; 6 - clone 5280; 7 - clone 16/9; 8 - Pinot Oberlin; 9 - clone Re 14/13; 10 - clone W10/5; 11 - clone 5257; 12 - type Blauer Arbst. higher between the color types than within one color type. Differentiation of the Pinot colors would meet the demands of growers and nurseries. However, according to our findings it seems easier to identify clones by using SCAR markers derived by RAPD than to predict the berry color based on DNA analysis. In general, polymorphism of clonal material is rare and difficult to detect. Increasing numbers of RAPD analyses, however, improve the chance of detecting variations at the DNA level. Having carried out a high number of RAPD analyses we observed polymorphism of Pinot clones, contrary to recently published monomorphic Pinot cultivars and clones [29,30]. A possible reason for the lack of variability may be the use of clones derived from the same region of Europe. Genotyping by RAPD markers is easier and cheaper than other methods. In comparison with SSR markers, RAPD markers offer unlimited use and independence from any sequence knowledge. Currently, RAPD is a proper tool to detect clonal differences. A disadvantage of the RAPD markers and other multilocus-markers such as AFLP is the limited reproducibility of the results. While the results may be repeatable in the same lab, reproduction usually fails when the same analysis is carried out in different labs. Therefore, these methods are limited to detection of polymorphism. SSR analysis of Pinots and Pinot related cultivars: Rare differences of SSR polymorphism could be found by genotyping some closely related Pinot cultivars such as Pinot noir precocce, Farbklevner, and Samtrot. Pinot noir precocce is an early maturing type with different agronomical characteristics such as low yield and sensitivity to blooming disorders. We could not detect any SSR changes in comparison to the main Pinot (noir, gris, blanc) profile. The cultivars Farbklevner and Samtrot, thought to be closely related to Pinot (noir, gris, blanc), differ from Pinot at the VVS2, VVMD6, VVMD31, and VVMD36 loci. With the exception of Samtrot being different at locus VVMD36, they all showed the same profile as the cultivar Schwarzriesling (Mtillerrebe). Furthermore Schwarzriesling was always clearly differentiated from Pinot in classical ampelography by comparing leaf pubescence. Farbklevner and Samtrot were misinterpreted as Pinots but actually are hairless types of Schwarzriesling [1]. The critical question, to which group of cultivars Schwarzriesling belongs, should not only be answered on the basis of SSR data, but by considering all available differences to Pinot. With respect to the age of the whole Pinot family, one could easily argue that Schwarzriesling may be a Pinot mutant with more extensive alterations in the genome. Nevertheless, the characteristic allele of Schwarzriesling (Table 2) at loci VVS2 (129 bp) and VVMD36 (240 bp) indicates that the cultivar is not derived from Pinot by mutation. The length of the specific allele is different from that in all other V. vinifera, and negates the former assumption [8] that Pinot might be the parental cultivar of Schwarzriesling. Since the differences in morphology, anthocyanins, wine components [2,13] and even in the SSR profile are too exten-

... ~ 10 m REGNER etal. Table 2. SSR alleles (length in base pairs) at several microsatellite loci of Traminer and Schwarzriesling demonstrate the Pinot origin. Each allele of Pinot was derived either from Schwarzriesling or Traminer. For instance, at the locus VVS1 Pinot consists of one allele with 183 base pairs inherited from Schwarzriesling and the second allele with 190 base pairs is inherited from Traminer. Locus Traminer Pinot Schwarzriesling VVS 1 162 : 190 183 : 190 183 : 190 VVS 2 151 : 151 137 : 151 129 : 137 VVS 3 214 : 220 214 : 220 214:220 VVS 4 168:175 168:173 168:173 VVS 29 171 : 171 171 : 179 171 : 179 VVMD 5 232:238 228:238 228:238 VVMD 6 199 : 206 199 : 199 199 : 206 VVMD 7 243:257 239:243 239:243 VVMD 8 141 : 143 141 : 143 141 : 143 VVMD 14 228:238 218:238 218:238 VVMD 17 220 : 220 212 : 220 212 : 220 VVMD 21 249:249 249:249 249:249 VVMD 24 214 : 218 216 : 218 216 : 218 VVMD 25 253:253 243:253 243:253 VVMD 26 249:251 249:255 249:255 VVMD 27 189 : 189 185 : 189 185 : 189 VVMD 28 237:239 221:239 221:239 VVMD 31 204:216 216:216 216:216 VVMD 32 241:273 241:273 241:273 VVMD 36 254:264 254:254 240:254 VRZAG 7 157 : 159 157 : 157 157 : 157 VRZAG 15 167 : 167 167 : 179 167 : 179 VRZAG 21 202:208 202:208 202:208 VRZAG 25 238:247 227:238 227:238 VRZAG 29 114 : 118 114 : 118 114 : 118 VRZAG30 151:153 151:153 151:153 VRZAG 47 169 : 169 165 : 169 165 : 169 VRZAG 62 189 : 195 189 : 195 189 : 195 VRZAG 64 140 : 164 140 : 164 140 : 164 VRZAG 67 128 : 134 128 : 154 128 : 154 VRZAG 79 246:252 240:246 240:246 VRZAG 82 254:274 254:274 254:274 VRZAG 83 190:202 190:202 190:202 VRZAG 112 236 : 242 242:244 242:244 I... Pinot pedigree ~i ii i ~ Pinot noir gris blanc "~i ii:~ Heunisch iii / r ~ ~i! / \ Melon \....,:,... ]... A,u:9:r~01S... ~i! I Chardonnay... Fig. 2. Proposal for the pedigree of the Pinots and some related cultivars. sive to be explained by clonal mutation, we consider a hybridization of Schwarzriesling with another cultivar to be the origin of Pinot. With respect to available data, only the cultivar Traminer could be the possible second parent of Pinot (Table 2). Only a combination of Traminer and Schwarzriesling includes all alleles of Pinot. It is interesting that Traminer shares 45 alleles (66%) with Schwarzriesling and 47 alleles (69%) with Pinot, 44 alleles of Traminer being present in both cultivars. Therefore, the dimension of the relationship between Traminer and Schwarzriesling is similar to that between Traminer and Pinot. Segregation of SSR markers shows that common alleles of both parental types were transmitted to Pinot at a high rate (97%). Because of these similarities at the genotype level the phenotype of Pinot has to be closely related to Schwarzriesling. Based on the SSR data of 96 cultivars (Table 3), which could have played a role in the descent of Pinot, the probability of the proposed cross versus other possible origins was calculated by likelihood ratios. Cumulative likelihood ratios (Table 4) exclude other parentages for Pinot and therefore support this hypothesis. Why does Pinot occur in three berry pigment types while Schwarzriesling has had stable pigmentation? The hybridization of a Traminer blanc or Traminer gris with Schwarzriesling could be the explanation for three different pigmentations. Due to heterozygosity all color types can occur if the alleles that carry the pigmentation information are inherited by the progeny. Therefore, a cross of Traminer with Schwarzriesling may contain the three-berry-color spectrum. The seeds of some V. vinifera varieties (Traminer, Pinot, and Riesling) are rounder and indicate a connection with the V. sylvestris gene pool [24]. With respect to our findings, that Traminer is derived by hybridization of two V. sylvestris genotypes [17], round seeds are typical for Traminer. The cultivar Schwarzriesling has more elongated seeds. The five-lobed leaf of Schwarzriesling transformed into the three-lobed and rounder leaf of Pinot. As a parental cultivar, Traminer would provide all these characteristics, verified as primary descriptors by the Office International de la Vigne et du Vin, Paris (OIV). Summarizing all these facts we can propose the following origin of Pinot: Pinot = Schwarzriesling X Traminer The analysis of other Pinot-related cultivars shows that Blauer Arbst is identical with Pinot without any deviations from the Pinot SSR profile. Despite several alleles which they share, Affenthaler and Teinturier (Pinot teinturier) are not directly related to Pinot as shown in Table 5. Meunier- assumed to be a synonym for Schwarzriesling- could not be satisfactorily characterized. Several Meunier samples from different collections showed differing genotypes. Although in some cases Meunier is identical with Pinot, we also detected identity with Schwarzriesling and one Meunier genotype, with an individual SSR pattern, representing an outcross from Pinot. The second partner for this Pinotoffspring could not be identified. Schwarzriesling was

GENETIC RELATIONSHIPS AMONG PINOTS- 11 Table 3. The 96 cultivars of Vitis vinifera with possible potential for the parentage of Pinot. Affenthaler N Burgunder N/B/G Fiano B Monica B Albana B Cabernet franc N Frankentaler B Mondeuse B Aligote B Cabernet Sauv. N Freisa N Morillon B Altesse B Calitor B Frehgipfler B Moscato B Amigne B Canaiola N Furmint B Petit Meslier B Ansonica B Carignan N Gamay N Petit Verdot N Aramon N Catarrato B Gordin B Picolit B Ardelanca B Cesar N Grenache N Pineau d Aunis N Aubun N Chardonnay B Heunisch B Piquepol B Auxerrois B Chasselas B Honigler B Plant vert B Barbaroux R Chasselas musq. B Humagne B Poulsard N Baroque B Chasselas Court.B Juracon B Riesling B Basilicum B Chenin B Kauka N Rousanne B Bellevue B Chicaud B KSIIner N Sauvignon B Bequinol N Cirfandli R Kurzstingler B Savagnin B Berbecel B Clairette B Lagrein N Schwarzriesling N Bermestia B Colombard B Laska N Silvaner B/N Bianchetta Tr.B Cortese B Madleine B St. Laurent N Arbst N Courtillier musq.b Malvasia B Traminer G Blaufraenkisch N Croatina N Marsanne B Trollinger N Bouquet B Dolcetto N Mauzac B Teinturier N Brachet G Durella B Melon B Verdesse B Brunello N Elbling B Merlot N Veltliner B Budai B Erbaluce B Molette B Wildbacher N also thought to have been spread from France like the Pinots. In a historical ampelography from 1837 [7], it was mentioned that in the 18 th century Pinot (Klevner) and Traminer were cultivated in Burgundy, Champagne, and the Alsace. The SSR profile of Saint Laurent clearly shows that it could be a progeny of Pinot (Table 5). The second parent is still unknown. Considering the loss of wild types and V. sylvestris genotypes [24] in this century, we may not find all genotypes with relevance to today's cultivars. Reconstruction of a Pinot pedigree: The white varieties that show a close morphological relationship to Pinot are Auxerrois and Chardonnay [8]. Their SSR profile is characterized by one identical allele at each locus in common with Pinot alleles. An obvious conclusion is that Auxerrois and Chardonnay were derived by outcrossing of Pinot. Aligote and Melon show the same kind of relationship to Pinot. The cultivar Heunisch could be identified as the second parent. Heunisch is a very old variety, but still available in some collections. The name Heunisch can be found in literature of the Middle Ages. It seems that Heunisch is as large a family as the Pinots. We could detect several Heunisch genotypes one of which was identical to the cultivar Gouais. Some genotypes of the Heunisch family might be the second parental vine of Aligote, Auxerrois, Chardonnay, and Melon (Table 6). The proposed cross Pinot X Heunisch for the origin of these cultivars is supported by high likelihood ratios. Differences in agronomical behavior can be explained by segregation of the corresponding chromosomes. The most frequently used designations for vines found in the literature of the Middle Ages [25] are Frankisch (Vinum franconicum) and Heunisch (Vinum Table 4. Likelihood ratios of the probability of the suggested parentages of Pinot, Auxerrois, Chardonnay, Melon, and Aligote versus other possibilities. Probability values were calculated from relative allele frequencies derived from our cultivar collection and from the 95 % upper confidence limits. The calculations are based on the data of 96 cultivars and 34 microsatellite loci (see Materials and Methods). Cultivar Suggested parents Cumulative likelihood ratios of the suggested parentage (1) X (2) versus X x Y a,b (1) x X a,c (1) x rel (2) a,d (2) X X a,c (2) x rel (1) Pinot (1) Schwarzriesling 6.4 X 1018 7.3 X 107 4.8 X 103 2.8 X 1012 3.3 X 103 (2) Traminer (3.6 X 1012) 6.5 X 104 (1.4 X 102) (5.7 X 108) (1.8 X 102) Auxerrois (1) Pinot 4.0 X 1017 1.8 X 1013 8.5 X 103 9.1 X 108 4.8 X 103 (2) Heunisch (3.1 X 101 ) (7.2 X 109) (6.2 X 102) (3.4 X 106) 2.9 X 102) Chardonnay (1) Pinot 9.1 X 1018 2.7 X 1013 2.0 X 104 5.9 X 109 5.3 X 103 (2) Heunisch (9.1 X 101 ) (1.3 X 109) (1.5 X 103) (1.2 X 10 6) (3.5 X 102) Melon (1) Pinot 2.7 X 1017 2.6 X 101 5.4 X 103 1.8 X 101 1.5 X 104 (2) Heunisch (6.1 X 109) (4.9 X 102) (4.2 X 102) (2.3 X 107) (6.8 X 103) Aligote (1) Pinot 1.8 X 1017 2.6 X 10 l 8.0 X 103 2.3 1012 2.1 X 104 (2) Heunisch (5.3 X 109) (4.9 X 106) (5.3 X 102) (2.8 X 108) (4.3 X 103) a Values in parentheses are the cumulative likelihood ratios calculated with the 95 % upper confidence limits for the allele frequencies. b X and Y are random unrelated cultivars. c The identity of one of the suggested parents is assumed and the other parent is unknown. The identity of one of the suggested parents is assumed and the other parent is a close relative to the second suggested parent.

~ 12 ~ REGNER etal. Table 5. Alleles of Pinot, some related red-berried cultivars and Affenthaler and Teinturier indicate that the latter are misinterpreted Pinots and not closely related to it. These cultivars have some loci without any allele in common with Pinot. Locus Schwarzriesling Pinot Samtrot St. Laurent Affenthaler Teinturier VVS 1 183:190 183:190 183 : 190 183 : 190 183:190 183:190 VVS 2 129 : 137 137 : 151 129 : 137 137 : 151 137 : 151 137 : 151 VVS 3 214:220 214:220 214:220 214:220 214:220 214:220 VVS 4 168:173 168:173 168:173 168:173 168:173 168:173 VVS 29 171 : 179 171 : 179 171 : 179 171 : 179 171 : 179 171 : 179 VVMD 5 228:238 228:238 228:238 228:238 228:238 228:238 VVMD 6 199 : 206 199 : 199 199 : 206 199 : 199 199 : 199 199 : 199 VVMD 7 239:243 239:243 239:243 239:243 251:257 239:251 VVMD 8 141:143 141:143 141:143 141:143 141:143 141:143 VVMD 14 218 : 238 218 : 238 218 : 238 218 : 238 218 : 238 218 : 238 VVMD 17 212 : 220 212 : 220 212 : 220 212 : 220 212 : 220 212 : 220 VVMD 21 249:249 249:249 249:249 249:249 249:249 249:249 VVMD 24 216 : 218 216 : 218 216 : 218 216 : 218 216 : 218 216 : 218 VVMD 25 243:253 243:253 243:253 243:253 243:253 243:253 VVMD 26 249:255 249:255 249:255 249:255 249:255 249:255 VVMD 27 185 : 189 185 : 189 185 : 189 185 : 189 185 : 189 185 : 189 VVMD 28 221:239 221:239 221:239 221:239 221:239 221:239 VVMD 31 216:216 204:216 216:216 216:216 206:214 216:224 VVMD 32 241:273 241:273 241:273 241:273 241:273 241:273 VVMD 36 240:254 254:254 254 : 264 254 : 254 276 : 294 240 : 254 VRZAG 7 157 : 157 157 : 157 157 : 157 157 : 157 157 : 157 157 : 157 VRZAG 15 167:179 167:179 167:179 167:179 167:179 167:179 VRZAG 21 202 : 208 202 : 208 202 : 208 202:208 202:208 202:208 VRZAG 25 227:238 227:238 227:238 227:238 227:238 227:238 VRZAG 29 114:118 114:118 114:118 114:118 114:118 114:118 VRZAG 30 151:153 151:153 151:153 151:153 151:153 151:153 VRZAG 47 165 : 169 165 : 169 165 : 169 165 : 169 165 : 169 165 : 169 VRZAG 62 189 : 195 189 : 195 189 : 195 189 : 195 189 : 195 189 : 195 VRZAG 64 140 : 164 140 : 164 140 : 164 140 : 164 140 : 164 140 : 164 VRZAG 67 128 : 154 128 : 154 128 : 154 128 : 154 128 : 154 128 : 154 VRZAG 79 240:246 240:246 240:246 240:246 240:244 256:262 VRZAG 83 190:202 190:202 190:202 190:202 190:202 190:202 VRZAG 112 242:244 242:244 242:244 242:244 242:244 242:244 hunicum). Aligote, Auxerrois, Chardonnay, and Melon are the combinations of these two very individual gene pools (Table 6). The Heunisch gene pool [25] was spread by the Huns and was an important source ofgermplasm. Frankisch was the definition of the endogenous vines derived from wild types. Traminer and Traminer-related cultivars such as Silvaner and Pinot represent the second important genepool of the Middle Ages [2]. In a simplified way, they can be seen as a combination of quality and quantity. Under moderate and cool climatic conditions Pinot is regarded as a vine that produces wine of high quality. In contrast, Heunisch vines provide high yields of minor quality. Both cultivars were planted at the same location, which increased the likelihood of hybridization. Many of today's cultivars are probably descendants of this random breeding [18]. The origin of several cultivars from a reconstructed cross is supported by calculations of cumulative likelihood ratios and a low probability for any other combination of these parentages (Table 4). The possibility that one of the putative parents is correct and the crossing partner is a close relative of the proposed parent (Table 4) was taken into consideration. These alternatives reached likelihood ratios close to the proposed parentage; however, none of the alternatives exceeded the proposed parentage. Therefore, cumulative likelihood ratios represent the ideal statistical method [6] to calculate the probability that both cultivars of the proposed cross are parental types. Usually, if the parental genotypes cover all alleles of the resulting cultivar, the probability for paternity is high. Mismatches of alleles, by contrast, decrease the probability for paternity. According to our findings, the designation of a genotype as a cultivar, type, or clone could be reevaluated. It would be easier to differentiate genotypes as distinct cultivars only if their main genetic backgrounds differ. Minor differences like variation within a cultivar should be defined as type or clone of a cultivar. Applying this definition to the Pinot cultivars, Pinot gris and Pinot blanc would only be defined as color types of Pinot noir, and Farbklevner and Samtrot would be hairless types of Schwarzriesling. Blauer Arbst and Pinot noir precocce

GENETIC RELATIONSHIPS AMONG PINOTS m 13 Table 6. Alleles of Pinot-related cultivars show the offspring relationship of the cultivars Auxerrois, Aligote, Chardonnay, and Melon derived from a cross of Pinot X Heunisch. Locus Pinot VVS 1 183:190 VVS 2 137:151 VVS 3 214:220 VVS 4 168:173 VVS 29 171 : 179 VVMD 5 228:238 VVMD 6 199 : 199 VVMD 7 239:243 VVMD 8 141 : 143 VVMD 14 218 : 238 VVMD 17 212 : 220 VVMD 21 249:249 VVMD 24 216 : 218 VVMD 25 243:253 VVMD 26 249:255 VVMD 27 185 : 189 VVMD 28 221:239 VVMD 31 204:216 VVMD 32 241:273 VVMD 36 254:254 VRZAG 7 157:157 VRZAG 15 167 : 179 VRZAG 21 202:208 VRZAG 25 227:238 VRZAG 29 114 : 118 VRZAG 30 151 : 153 VRZAG 47 165 : 169 VRZAG 62 189 : 195 VRZAG 64 140 : 164 VRZAG 67 128 : 154 VRZAG 79 240:246 VRZAG 83 190:202 VRZAG 112 242 : 244 Auxerrois Aligote Chardonnay Melon Heunisch 183:190 183:190 183 : 190 183:190 183:190 133:151 133:137 137:143 137:143 133:143 214:220 214:220 214:220 214:220 220:220 168:169 168:173 168:173 168:168 168:169 171:179 171:179 171:179 171:179 171:179 234:238 228:240 234:238 238:240 234:240 189 : 199 189 : 199 199 : 209 199 : 209 189 : 209 239:243 239:243 239:243 239:249 239:249 141:143 141 : 147 141:147 141 : 147 141:147 218 : 238 218 : 238 218 : 228 218 : 228 218 : 228 220:222 220:220 220:222 212:220 220:222 249:249 249:249 249:249 249:249 249:249 210:218 210:218 210:218 210:218 210:218 243 : 253 243 : 243 243 : 259 243 : 259 243:259 249:255 251:255 249:255 249:251 249:251 181 : 189 179:189 181 : 189 181 : 185 179 : 181 231:239 231:239 221:231 231:239 231:249 214:216 212:216 214:216 214:216 212 : 214 253:273 241:273 241:273 241:253 253:273 254:264 254:276 254:276 254:276 264:276 157:157 157:157 157:157 157:157 157:157 167:167 167:179 167:179 167:179 167:167 204:208 202:208 202:208 202:204 204:208 227:238 227:240 227:227 227:238 227:240 114:118 114:118 114:114 114:114 114:118 149:151 151 : 153 151 : 151 151 : 151 149:151 161 : 169 159:169 161 : 169 159:165 159:161 189 : 197 195 : 197 189 : 197 195 : 205 197 : 205 140:160 160:164 160:164 140:160 160:160 128:141 141:154 141:154 128:141 141:141 244:246 244:246 244:246 240:244 238:244 190 : 196 190 : 202 190 : 202 190 : 202 190 : 196 242:244 242:244 242:242 242:242 242:244 would represent types of Pinot noir with slightly different characteristics. Conclusions Genetic analysis by means of SSR markers showed no polymorphism with Pinot noir, Pinot gris, and Pinot blanc, whereas RAPD markers differentiated several clones. The cultivars Schwarzriesling and Farbklevner were identical and differed from Pinot. Samtrot could be verified as a mutation of Schwarzriesling with an individual genetic profile. Because of the original allele length at two SSR loci, we can say that Schwarzriesling is more closely related to V. sylvestris types than to Pinot. Our results suggest that Pinot was derived from a cross of Schwarzriesling X Traminer. The Pinot-related cultivar Saint Laurent is a Pinot offspring, and up to now, there has been no strict definition for Meunier as a homogenous cultivar. Affenthaler and Teinturier are not directly related to the Pinots despite having several common alleles. The cultivars Aligote, Auxerrois, Chardonnay, and Melon are descendants from a cross of Pinot Heunisch. Literature Cited 1. Ambrosi, H., E. Dettweiler et al. Farbatlas Rebsorten: 300 Sorten und ihre Weine. Ulmer Verlag Stuttgart (1994). 2. Bassermann-Jordan, F. Geschichte des Weinbaus, 3 rd ed. Pf&lzische Verlagsanstalt GmbH., Neustadt an der WeinstraBe, (1975). 3. Bourquin, J. C., L. Otten, and B. Walter. PCR-RFLP analysis of Vitis, Ampelopsis, and Parthenocissus and its application to the identification of rootstocks. Vitis 34:103-108 (1995). 4. Bowers, J. E., and C. P. Meredith. Genetic similarities among wine grape cultivars revealed by restriction fragment-length polymorphism (RFLP) analysis. J. Am. Soc. Hortic. Sci. 121(4):620-624 (1996). 5. Bowers, J. E., G. S. Dangl, et al Isolation and characterization of new polymorphic simple sequence repeat loci in grape (Vitis vinifera L.). Genome 39:628-633 (1996). 6. Bowers, J. E., and C. P. Meredith. The parentage of a classic wine grape, Cabernet Sauvignon. Nature Genetics 16:84-86 (1997). 7. Burger, J. Systematische Klassifikation und Beschreibung der in den 5sterreichischen Weing&rten vorkommenden Traubenarten. Herold Verlag, Wien (1837). 8. Galet, P. Cepages et Vignobles de France. Tome I1.

~ 14-- REGNER etal. L'Ampelographie francaise. 2 d ed. Imprimerie Charles Dehan, Montpellier (1990). 9. Galet, P. C~pages et Vignobles de France. Tome I1. L'Ampelographie francaise. 1 st ed. Imprimerie Charles Dehan, Montpellier (1958). 10. Grando, S. M., L. De Micheli, et al RAPD markers in wild and cultivated Vitis vinifera. Vitis 34(1 ):37-39 (1995). 11. Lodhi, M. A., N. F. Weeden, and B. I. Reisch. Characterization of RAPD markers in Vitis. Vitis 36(3):133-140 (1997). 12. Messier, W., L. Shou-Hsien, and St. Caro-Beth. The birth of microsatellites. Nature 381:483 (1996). 13. Peyron, D. Les composes phenoliques du Pinot noir de Bourgogne. Rev. Oenol. 86:22-25 (1998). 14. Regner, F., and R. Messner. Molekulare Differenzierung von Rebsorten mittels RAPD-Analyse. Mitteilungen Klosterneuburg, 43:160-164 (1993). 15. Regner, F., H. Steinkellner, et al. Genetische Charakterisierung von Rebsorten (Vitis vinifera) durch Mikrosatelliten-Analyse. Mitt. Klosterneuburg 46:52-60 (1996). 16. Regner, F., K. Sefc, et al Identification, characterization and relationship of grapevines by microsatellite analysis. VII Int. Symposium of Grapevine Breeding and Genetics, Montpellier 6-10.7 (1998). 17. Regner, F., A. Stadlbauer, and H. Kaserer. Considerations about the evolution of grapevine and the role of Traminer. VII Int. Symposium of Grapevine Breeding and Genetics, Montpellier 6-10.7. (1998). 18. Regner, F., A. Stadlbauer, and C. Eisenheld. Heunisch X Fr&nkisch, ein wichtiger Genpool europ&ischer Rebsorten (Vitis vinifera L. sativa) Vitic. Enol. Sci. 53(3):114-118 (1998). 19. Regner, F., and A. Stadlbauer. Genetic analysis of traditional grapevine cultivars - perception of viticulture. XXIV Congres Mondiale de la Vigne et du Vin, Mainz 5-9.7.1999 Section1:53-59 (1999). 20. Sefc, K. M., H. Steinkellner, et al Application of microsatellite markers to parentage studies in grapevine. Vitis 36(4):179-183 (1997). 21. Sefc, K. M., F. Regner, et al. Identification of microsatellite sequences in Vitis riparia and their applicability to genotype different Vitis species. Genome 42:1-7 (1999). 22. Sensi, E., R. Vignani, et al. Characterization of genetic biodiversity with Vitis vinifera L. Sangiovese and Colorino genotypes by AFLP and ISTR DNA marker technology. Vitis 35(4):183-188 (1996). 23. Smith, J. R., J. D. Carpten, et al Approach to genotyping errors caused by nontemplated nucleotide addition by TAQ DNA polymerase. Gen. Res. 5:312-317(1995). 24. Schuhmann, F. Berichte 0ber die Verwendung der Wildrebe Vitis vinifera L. var. Silvestris. Gmelin 72:213-218 (1971). 25. Schuhmann, F. Rebsorten und Weinarten im mittelalterlichen Deutschland. in Weinwirtschaft im Mittelalter. Band 9, Stadtarchiv Heilbronn (1997). 26. Thomas, M. R., S. Matsumoto, et al Repetitive DNA of grapevine: Classes present and sequences suitable for cultivar identification Theor. Appl. Genet. 86:173-180 (1993). 27. Thomas, M. R., and N. Scott. Microsatellite repeats in grapevine reveal DNA polymorphisms when analysed as sequence-tagged sites (STSs). Theor. Appl. Genet. 86:985-990 (1993). 28. Thomas, M. R., P. Cain, and N. Scott. DNA typing of grapevines: A universal methodology and database for describing cultivars and evaluating genetic relatedness. Plant Mol. Biol. 25:939-949 (1994). 29. Tschammer, J., and E. Zyprian. Molecular characterization of grapevine cultivars of Riesling -type and of closely related Burgundies Vitis 33:249-250 (1994). 30. Ye, G. N., G. Soylemezoglu, et al Analysis of the relationship between grapevine cultivars, sports and clones via DNA fingerprinting. Vitis 37:33-38 (1998).