The European grapevine, Vitis vinifera L., is the most commonly cultivated species of bunch grape. Indigenous to Eurasia, V. vinifera is one of approximately 60 Vitis species worldwide (Alleweldt and Possingham, 1988; Martinez-Zapater et al., 2010). Domestication of V. sylvestris C.C. Gmelin, the progenitor of V. vinifera, is proposed to have occurred approximately 7,000 years ago with initial cultivation occurring as early as 3,500 BC (Zohary and Hopf, 1993).
Initial classifications of grape cultivars separated V. vinifera into three groups: occidentalis, orientalis, and pontica (Negrul, 1938). Western European wine grapes with smaller berries were classified as occidentalis, orientalis cultivars were table grapes from western Asia with large berries, and pontica cultivars, having intermediate berries, were named after the ancient country from Asia Minor and arose from eastern Europe and near the Black Sea and are typified by Greek and Italian grapes (Aradhya et al., 2003; Negrul, 1938). These classifications have been generally supported by genetic analysis of structure within populations of V. vinifera. Finding that human use and geography dictated the bulk of cultivar relatedness, Bacilieri et al. (2013) utilized STRUCTURE software to analyze 20 nuclear microsatellite loci across 2,096 V. vinifera genotypes via a Bayesian approach. Three main groups isolated when Ks clustering was set to three indicated two wine groups, West and Central Europe (42% of attributed cultivars) and Balkans and East Europe (36%), and one table grape group derived from the East (22%). Additionally discovered admixed genotypes were suggested to arise from human interactions, both transfer of plant materials and breeding.
Cluster analysis of a separate subset of eight microsatellite loci within 366 V. vinifera accessions clustered cultivars into three groups: Mediterranean table grapes, Western European wine grapes, and Central European grapes (Aradhya et al., 2003). Groups within cluster varied in the extent they typified occidentalis, orientalis, and pontica descriptions, but generally the Mediterranean table grapes resembled orientalis with some pontica, Western European wine grapes represented both occidentalis and pontica. Further analysis of an Unweighted Pair Group Method with Arithmetic Mean (UPGMA) phylogenetic tree depicted four main clusters with slight crossover between the three original grape classifications. The majority of genetic polymorphism variation was discovered to lay within groups, rather than between them, with the first two PCA vectors only accounting for 26% of variation.
Genomic Resource Evaluation and Development
The grapevine genome is proposed as the result of an ancient polyploidization event either resulting as the merger of three genomes or following a hexaploidization event (Jaillon et al., 2007; Malacarne et al., 2012). Fairly small in size, at approximately 475 Mb, bunch grape genomes are treated as diploid in a contemporary breeding sense (Lodhi and Reisch, 1995). PN40024, a selection derived from ‘Pinot noir’ successively selfed to approach homozygosity, has been useful in whole-genome shotgun sequence assembly (Jaillion et al., 2007).
Barnaud et al. (2006) published the first study on linkage disequilibrium (LD) within the cultivated European grapevine, V. vinifera L. subsp. vinifera (sativa), their findings suggest that QTL mapping using genomic techniques may be useful, especially for wine type grapevines. They discovered LD occurred widely within five linkage groups, demonstrating a potentially restricted genetic base following domestication, with few recombination events. Rapid LD decay occurred in a natural population of 85 French Vitis vinifera L. subsp. silvestris selections, the progenitor wild relative of the cultivated European grapevine (Barnaud et al., 2010). Interestingly, following potential bottleneck events associated with the introduction of pests and diseases, genetic diversity was found to be reduced in wild grapevines when compared to cultivated. Using 160 SSR markers, two grouping clusters were estimated based on 81 Chinese native selections composed of 15 Vitis species and LD was estimated to occur up to 14.13 cM (Zhang et al., 2017).
Grapes are highly outcrossed and inbreeding intolerant, leading to highly heterozygous genomic structure (Olmo, 1979; This et al., 2006; Velasco et al., 2007). This heterozygosity is exemplified by comparison of homologous chromosomes revealing one million gaps, approximately 11% DNA sequence variation between Pinot noir homologous chromosomes (Velasco et al., 2007).
Early genetic linkage map development involving a 60 seedling double-pseudotestcross population of Cayuga White × Aurore indicated linkage groups ranging from 14 to 135 cM with 19 to 22 linkage groups based on 422 RAPD markers and additional RFLP and isozyme markers (Lodhi et al., 1995). More recent work revealed 19 linkage groups representing the 19 haploid chromosomes of Vitis spp., with linkage groups ranging from 48.3 to 95.1 cM in length (Troggio et al., 2007; Velasco et al., 2007).
Newer strategies for gene identification within grapes include association studies of large germplasm panels. Design of association panels is critical in grapevine, and perennial fruit crops due to the extensive time requirements and limited germplasm collections associated with fruit crops (McClure et al., 2014; Nicolas et al., 2016). Nicolas et al. (2016) proposed a 279 cultivar association panel for grapevines composed of selections combined to minimize the confounding