Microsatellite analysis of ancient alpine grape cultivars: pedigree reconstruction of Vitis vinifera L. 'Cornalin du...

Ancient and closely related grape cultivars from the Alps were analyzed with 50 microsatellite markers: 'Cornalin', 'Humagne Rouge' and 'Goron' from Valais (Switzerland); 'Cornalin', 'Petit Rouge' and 'Mayolet' from the Aosta Valley (Italy). Our results confirmed previous studies showing that the 'Cornalin' cultivars from Switzerland and Italy are distinct, and that 'Humagne Rouge' is identical to 'Cornalin' from the Aosta Valley. We propose the nomenclature 'Cornalin du Valais' and 'Cornalin d'Aoste' in order to prevent further confusion. At each locus, 'Goron', 'Petit Rouge', 'Mayolet' and 'Cornalin d'Aoste' all share at least one allele with 'Cornalin du Valais', strongly suggesting parent/offspring relationships. Alleles at 49 out of 50 microsatellite loci are consistent with 'Cornalin du Valais' being the progeny of 'Petit Rouge' and 'Mayolet'. The exception is a 10-base pair discrepancy at one locus, most likely the result of somatic mutation in one of the parents, since this parentage is supported by high likelihood ratios and historical data. We hypothesize that 'Cornalin du Valais' originated in the Aosta Valley through a natural cross and was then introduced into Valais centuries ago, probably via the Great St. Bernard Pass. Furthermore, 'Cornalin du Valais' is likely to be one of the parents of both 'Goron' and 'Cornalin d'Aoste', the respective second parents remaining unknown. This pedigree provides a convincing explanation for the allele-sharing patterns and is strongly supported by historical data. The present work is the first grapevine parentage study to deal with a multiple repeat unit discrepancy at a microsatellite locus. We suggest that the use of increasingly large numbers of loci in making parentage determinations leads to a corresponding increase in the probability of encountering a locus with intra-cultivar variability during the analysis. We therefore assume that a sole multiple repeat unit discrepancy is not sufficient to discard a parentage hypothesis.