The first grapevine germplasm banks emerged in the late nineteenth century, after the appearance of phylloxera in Europe, from North America, in the middle of that century (Cabello et al., 1999). A significant loss of native plant material occurred due to the disappearance of millions of hectares produced by the attack of that insect. More recently, genetic diversity has suffered another drawback. New plantations with foreign material with low genetic variability have reinforced genetic erosion of native germplasm. Moreover, European Union incentives for restructuring and conversion, particularly in Portugal and Spain, have conducted to the loss of hectares of old vineyards and, most probably, also autochthonous minor cultivars. Germplasm banks assume a huge importance in the preservation of local cultivars that, due to their low rentability, have a reduced area of cultivation or that, inclusive, have no longer expression in viticulture areas and their existence is restricted to collections.

The number of different cultivars held in grapevine germplasm collections around the world is estimated to be approximately 10,000 (Alleweldt & Dettweiller, 1994). Among them only few hundred are cultivated for commercial wine production (Truel et al., 1980). The management of these collections requires attention to avoid redundancy, to track introductions that were wrongly assigned to a cultivar and to assist clone selection (Pelsy et al., 2010). So, identification of the plant material is crucial and represents the first step in germplasm management (Cipriani et al., 2010; Laucou et al., 2011).

Morphology has been the most used tool in the characterization of grapevine germplasm in most of the plant collections (Boursiquot & This, 1996; Ortiz et al., 2004). International organisations such as the OIV (Office International de la Vigne et du Vin) or the ex-IPGRI (International Plant Genetic Resources Institute, present Bioversity International) have published useful descriptors for the ampelography and comparison studies to be carried out with the germplasm material (OIV, 2009). However, this process is carried out on adult plants; a longer period is required before the identification of accessions and often it is not conclusive on the distinction of close cultivars. As many synonyms or homonyms exist for cultivars, passport data are not always sufficient to certify identities, and errors can arise (Buhner-Zaharieva et al., 2010; Laucou et al., 2011).

In recent decades, DNA-based methodologies were implemented, enabling easier and more accurate identification of Vitis germplasm. Pioneer and also recent research with Random Amplified Polymorphic DNA (RAPD) and Inter Simple Sequence Repeat (ISSR) molecular markers has been successful and widely applied to estimate genetic diversity among cultivated table and wine grapevine varieties, wild grapes and also rootstocks and to distinguish grapevine cultivars and clones, either alone (Moreno et al., 1995, 1998; Vidal et al., 1999; Ercisli et al., 2008; Karatas & Ağaoğlu, 2008; Tamhankar et al., 2008; Jing & Wang, 2013), combined both (Herrera et al., 2002) and combined with other markers (Gogorcena et al., 1993: RAPDs and Restriction Fragment Length Polymorphisms (RFLPs); Ulanovsky et al., 2002 and Pinto-Carnide et al., 2003: RAPDs and Simple Sequence Repeats (SSRs); Seyedimoradi et al., 2012: ISSRs and Directly Amplified Minisatellite DNA (DAMD).

The objective of the present study was to use RAPD and ISSR molecular marker systems to evaluate the content on redundancy, synonymies and homonymies in a group of 56 Portuguese accessions from two different grapevine germplasm collections, which represent all the cultivars used for ‘Vinhos Verdes’ quality wines production and also many cultivars of ‘Douro’ and ‘Porto’ DOC denominations.

In the present study, RAPD and ISSR markers were used to analyse the variability in 56 Portuguese accessions of V. vinifera cultivars. In several cases, for accessions with the same designation, different samples collected in two different Portuguese grapevine germplasm collections were analysed (Table 1).

The conservation of band profiles demonstrates the reproducibility of both techniques used in the study. RAPD and ISSR molecular techniques accumulate several advantages, namely, small amount of DNA required, facility of technical procedure, no fluorescence labeling, no need for information on template DNAs or the synthesis of specific primers and simultaneous amplification of several loci. So, their use is adequate to screen high number of samples in order to detect duplicated material, suspicious of erroneous identifications or synonyms and homonyms.

Few RAPD and ISSR primers (see Table 2) were needed to generate highly diagnostic and reproducible fingerprint. The RAPD primer OPO-07 and the ISSR primers UBC-888, UBC-889 and UBC-890 used in this study, revealed a high capacity for grapevine cultivars discrimination (Table 3) given the high number of total and polymorphic bands, which is in agreement with the results of Zietkiewicz et al. (1994) and Moreno et al. (1998) for the ISSR amplifications with 5’ three-anchored primers (UBC-888, UBC-889 and UBC-890).

Using RAPD and ISSR markers 37 and 36 different molecular profiles were obtained, respectively, within the 56 accessions analysed, at 0.95 coefficient of similarity (Figs. 2 and 3). Likewise, Lopes et al. (1999), studying 49 supposed different cultivars from the Portuguese National Ampelographic Collection and from the Grapevine Collection of Terceira (Azores islands) through microsatellite loci amplification, detected some synonym cases and only 36 different SSR profiles were found. Similarly, Laucou et al. (2011) analysed 4,370 accessions of the INRA grape repository at Vassal with 20 microsatellite markers and only found 2,836 SSR single profiles. Cipriani et al. (2010) analysed 1,005 grapevine accessions from CRA-VIT of Conegliano collection by amplifying 34 microsatellite loci, identified 200 groups of synonyms and only 745 unique genotypes and Buhner-Zaharieva et al. (2010) in germplasm analysed from the Movera Grapevine Germplasm Bank in Aragón, Spain, found in 36 autochthonous accessions only 24 SSR profiles, besides 33 misnamed accessions.

Combination of molecular and morphological characterization methodologies has led to a good management of grapevine genetic resources (Ortiz et al., 2004; Balda et al., 2014; Ferreira et al., 2015; Maul et al., 2015). In a characterization of V. vinifera L. accessions from the Spanish gene bank at Alcalá de Henares, Ortiz et al. (2004) using morphological descriptors, isoenzymes and microsatellites, reduced the number of different accessions from 621 to 177, which represents less than 30% of the initial number. In the scope of the COST Action FA1003, in 997 accessions of Eastern European cultivars analysed through ampelography and nine microsatellite markers amplification, only 659 unique profiles/cultivars were found (Maul et al., 2015).

Of the 56 accessions that were studied, Moscatel Galego Branco and Moscatel Galego Roxo are considered to be berry skin-colour mutants and so, ampelographically well distinguishable. In addition to the RAPD and ISSR markers used in the present study, as expected, also nuclear microsatellites amplification performed in these cultivars failed to distinguish this pair of cultivars (Ferreira et al., 2016). Other berry skin–colour mutants have been reported, namely colour variants of Muscat of Alexandria cultivar, as sharing the same molecular fingerprint at trueness-to-type established loci, and differing in specific genes related to anthocyanin biosynthesis (De Lorenzis et al., 2015).

Results highlight the genetic proximity between Sousão and Vinhão accessions. Sousão is the prime name of a cultivar grown especially in ‘Vinhos Verdes’ Region whose cultivar designation was modified to Sezão in the last review of the ‘Portuguese List of Varieties fit for Wine Production’ (DR, 2012). Vinhão has been reported as the synonym of the Spanish cultivar Sousón (Martín et al., 2006). However, a focus of confusion exists in Douro Region, where, frequently, is given the name Sousão to the cultivar Vinhão. Though, the observed separated RAPD clusters for Sousão and Vinhão groups are correct. Nevertheless, the fact that these accessions have a miscellaneous of names between the two RAPD clusters and that they cluster together in the ISSR marker analysis, suggest that Sousão and Vinhão accessions are genetically close.

Also Gouveio Douro (EVAG), Gouveio (UTAD) and Verdelho (UTAD) clustered together in both molecular markers systems (Table 4; Figs. 2 and 3; Fig. S1 [suppl.]). In this case, the accessions Verdelho and Gouveio Douro are most probably the cultivar officially designated Gouveio; in some localities of ‘Douro’ Region, Gouveio is designated as Verdelho (Pereira & Sousa 1990). Care must be taken to not confuse with the variety Verdelho, code PRT50317 (DR, 2012), a different variety, much used for ‘Madeira’ wine production with cultivation restricted to the island.

Pedernã is the local designation in ‘Vinhos Verdes’ DOC Region for the official cultivar named Arinto (Mota & Silva, 1986), and so the accessions Arinto Oeste (UTAD)/Pedernã (EVAG)/Pedernã (UTAD) have an identical profile (Table 4; Figs. 2 and 3; Fig. S1 [suppl.]). The designation of UTAD’s accession as ‘Arinto Oeste’ is surely due to the great importance of Arinto in the Bucelas DOC Region, located in west of Portugal, near Lisbon, where it is also known as ‘Arinto de Bucelas’.

RAPD and ISSR band patterns suggest a few cases of misidentifications between and within collections. The accession Arinto Bairrada (UTAD) revealed the band pattern of Baga (UTAD). This coincidence can be explained considering that Baga, in ‘Bairrada’ Region, is frequently designated Tinto Bairrada. The accessions Malvasia Fina (EVAG) and Bical (EVAG) were found to be very close but different from the accession Malvasia Fina (UTAD) that clustered at a distant similarity level. That suggests misidentification of Malvasia Fina in EVAG collection. Also, the accessions Azal Tinto (UTAD)/Azal Tinto (EVAG) and Rabigato (UTAD)/Rabigato (EVAG) revealed band patterns quite different (Figs. 2 and 3). Azal Tinto is the synonym of Amaral (Caño Bravo in Spain, Martín et al., 2006) and is mentioned as having several descendants (Castro et al., 2012; Lacombe et al., 2013). One of the Azal Tinto accessions may be in fact other genotype, eventually its relative. The different Rabigato profiles can be explained considering that there is a cultivar in ‘Vinhos Verdes’ DOC Region, Rabo de Ovelha, which is commonly designated Rabigato (Mota & Silva 1986).

A major obstacle to good management of grapevine germplasm banks is the persistence of synonyms and homonyms up to the current days in the viticulture worldwide. Some germplasm banks accumulate the responsibility of multiplication and commercialization of grapevine material. Efforts are being done and have to proceed in order to find an official name for propagation and distribution. The management of several grapevine collections has been carried out with the assistance of molecular markers, namely in Iberian Peninsula (Lopes et al., 1999; Ibáñez et al., 2003; Martín et al., 2003, 2006; Santiago et al., 2007; Santana et al., 2008; Veloso et al., 2010; Castro et al., 2011; Balda et al., 2014; Alifragkis et al., 2015) with great relevance in the correction of identification mistakes. In specific, RAPD and ISSR PCR-based fingerprinting are informative for estimating the extent of genetic diversity and patterns of genetic relationships among grape accessions in germplasm holdings (Dhanorkar et al., 2005; Ercisli et al., 2008; Karatas & Aǧaoǧlu, 2008; Zeinali et al., 2012).

Results evidence the necessity of grapevine material characterization in ampelographic collections, complementing morphological descriptors with molecular markers for duplicates and synonyms detection.