Introduction

Apples are the fruit species grown in the third greatest quantity in the world after citrus fruits and bananas, but global apple production involves only a few dozen cultivars. In the case of an ecological catastrophe (e.g. appearance of new pathogen species), the restricted range of cultivars could endanger the reliability and profitability of apple production (Lespinasse et al. 2000). Nowadays, the role of local cultivars and land races in large-scale production has become completely insignificant. In Hungary, there was a special cultivar structure at 1800s. These local cultivars have unknown origin, but they are typical of their area. Old and local apple cultivars have enormous value from the point of view of biodiversity (Szani 2007; Holler 2007). A survey undertaken in apple gene banks in 22 different countries have been demonstrated the number of preserved cultivars varied from 90 to more than 2000, and the percentage of indigenous cultivars in the collections ranged from 3 to 82 (Nybom and Garkava-Gustavsson 2009).

Some of these cultivars have good resistance or excellent inner quality, while others satisfy special consumer demands. The choice of fruit available on the market must be broadened if fruit consumption is to become more varied. This could be achieved not only by re-introducing old cultivars into cultivation but also by using them as crossing parent in the development of new cultivars.

Within the framework of the Hungarian apple breeding programme (Tóth et al. 1994), one of the main tasks is to collect, select, evaluate and preserve genetic resources. Of the old genotypes known to exist in the Carpathian Basin, almost 200 local cultivars have been collected in the department’s gene bank over the last two decades (Tóth 2005a). As in other parts of Europe (Tartarini et al. 2004; Kellerhals et al. 2012; Nybom et al. 2012), studies of these have been made in Hungary, and old cultivars with good resistance and acceptable quality have been selected (Tóth et al. 2005a, b, 2013).

Over the last decade, surveys have begun in many countries on the diversity available in gene banks, thus making it possible to assess the genetic cultivar that once existed in the cultivar use of a given region and to determine similarities and differences in the cultivar composition compared with other regions. In addition to phenological and morphological descriptions, it is also advisable to carry out genotypic analysis and this is reversely right (Pereira-Lorenzo et al. 2003; Routson et al. 2009; Szalatnay et al. 2009; Santesteban et al. 2009).

This work also casts light on accessions of the same cultivar that have been mistakenly documented due to the use of erroneous names (van Treuren et al. 2010). The diversity analysis performed in gene banks makes it possible not only to identify the cultivars but also to resolve pomological debates on questions such as synonymous cultivar names and the analysis of cultivar groups. This requires the application of identical analytical methods, which are now being elaborated by European Cooperative Programme for Crop Genetic Resources Network (ECPGR).

Some of the cultivars found in the gene bank could be of use not only in breeding but also in organic farming, so there is a great need for the re-evaluation of the old apple cultivars (Tóth 2005b; Papp et al. 2011; Tóth et al. 2013). Some of the old apple cultivars have become well adapted to the soil and climatic conditions of the Carpathian Basin producing good yields and highly appreciated fruit quality (Mitre et al. 2009).

It is important to examine old apple cultivars because the pomological descriptions available for old cultivars are not uniform. Many collections of cultivar descriptions use several names for one cultivar or the same name for several cultivars. Number codes, such as that elaborated by Vercier (1948) or by Brózik (1974), are useful in that they allow uniformly compiled descriptions to be made. Nowadays, in most cases, the guidelines given by Union Internationale pour la Protection des Obtentions Végétales (UPOV) or Community Plant Variety Office (CPVO) are applied, as the brief descriptions and number codes provide sufficient information for cultivars to be differentiated. Considering ECPGR and UPOV international standards, guidelines for the phenotypic description of the fruit genetic resources were developed by Szalatnay (2006) which are being practically applied. They have the advantage that they can be applied on an international scale and enable the results obtained by different research teams to be easily compared.

The aims of the present work were to provide assistance for gene bank analysis and to achieve better knowledge of cultivar traits, which could be useful both for scientists and growers. Old apple cultivars, either originating from the Carpathian Basin or of foreign origin but grown widely in this region, were examined to provide a characterisation of the gene bank collection and the determination of diversity using morphological markers laid down in the UPOV guidelines.

Materials and methods

Cultivars included in the study

Among the old apple cultivars/genotypes found in the Carpathian Basin, morphological characterisation using UPOV descriptions were prepared for 60 cultivars, not counting the control cultivars. Hungarian land races and variants retrieved from the English National Fruit Collection (Faversham, Kent, UK) and collected from orchards in Subcarpathia (Ukraine), Transylvania (Romania), the National Park Aggtelek (Hungary) and the Mecsek Hills (Hungary) were also included in the study for the purpose of cultivar identification and characterisation (Table 1).

Table 1 Origin of the investigated cultivars preserving in the gene bank of Soroksár (Hungary)
Full size table

Experimental location

The gene bank and cultivar collections serving as the basis for the research are located at the experimental station of the Department of Pomology, Corvinus University of Budapest, in Soroksár (47° 40′ N, 19° 15′ E). The orchard was established in 2001 on MM.106 rootstocks, with a free spindle crown form. In the larger part of the orchard (the first 10 rows), the trees are arranged in 2 × 2 blocks, i.e. a total of four trees/cultivar in a 5 × 2.5 m space, planted next to each other in two neighbouring rows, while in the remaining part (rows 11–13), three trees per cultivar are planted in blocks within a row, in a space of 3.5 × 1.2 m.

Morphological observations

Morphological descriptions were compiled based on the UPOV TG/14/9 (2005) guidelines. A total of 53 morphological and 3 phenological traits were observed in the dormancy and vegetation periods (Table 2). Data were recorded when the trees were bare, at flowering, at the end of intensive shoot growth and at fruit maturity. In addition to subjective observations, measured data were also recorded, based on the analysis of 20 flowers, leaves and fruit from each cultivar. The samples required for the analyses were collected randomly from three or four trees, collecting no more than one sample from each inflorescence or shoot. The analyses were performed between 2007 and 2011, and observations were made on each cultivar in at least 2 years. The final number code classification of the cultivars was determined on the basis of several years of observations and measured data. It is important not to average the categories determined over a number of years; the category characteristic of each individual cultivar is taken as that which occurs most frequently in the separate annual classifications.

Table 2 Descriptors using by the morphological characterisation according to the UPOV TG/14/9 guidelines
Full size table

Statistical analysis

The use of the UPOV number code makes it possible to perform statistical analysis on the data. The frequency of the expression categories for the individual traits, given as a percentage, was analysed using the PASW 18.0 statistical program package.

UPOV considers two cultivars to be distinct if there is a clear difference for at least one trait, indicated by a difference of one category for qualitative (QL) and pseudo-qualitative (PQ) traits and two categories for quantitative (QN) traits. In the course of the statistical analysis, a matrix was compiled in Excel using the number code descriptions of the cultivars, taking the nature of the traits into consideration. During the pairwise comparison of the cultivars, a value of 0 was assigned if a difference of one (QL, PQ) or two (QN) categories was not observed and a value of 1 if this condition was met. The 0–1 coding values were then added for each cultivar to obtain the similarity matrix used for further statistical analysis.

Hierarchical cluster analysis and dendrogram construction were performed using the Ward method, based on this matrix, using the R program (R Development Core Team 2008).

Results and discussion

Characterisation of the cultivars in terms of their number code classification

A complete number code characterisation of 60 cultivars was carried out using the 56 traits proposed by the UPOV TG/14/9 guidelines. Hierarchical cluster analysis was carried out on the basis of the similarity matrix in the R program using the Ward method. The resulting dendrogram is illustrated in Fig. 1. Six groups can be clearly distinguished on the dendrogram. The Batul cultivar group and the ‘Kanadai renet’ genotypes each form a separate main cluster.

Fig. 1
figure 1

Genetic relationships between the cultivars on the basis of the similarity matrix compiled from the phenotypic UPOV number codes (A from the UK, K from Subcarpathia)

Full size image

The I main cluster contains all the cultivars in the Batul group with the exception of ‘Zöld batul’, which was classified in a completely different cluster, at a considerable distance from the other ‘Batul’ genotypes examined. The II main cluster consists of the Kanadai renet group. No cultivar groups can be identified for the cultivars in the other four (III–VI) clusters.

Other Renet cultivars were found in all four groups (cluster III–VI). Three cultivars in the Sóvári cultivar group (‘Nemes sóvári’, ‘Zöld sóvári’, ‘Daru sóvári’) formed a separate small cluster in the cluster IV. In the same main cluster (IV), but at some distance from these, ‘Sóvári nobil’ could be found in a separate subcluster. Although regarded as being the same cultivar, ‘Nemes sóvári’ and the ‘Sóvári nobil’ sample included in the investigations were thus classified in the same main cluster (IV) but in different subclusters, so they cannot be called synonyms. The ‘Beregi sóvári’ genotypes were grouped in a completely different main cluster (VI) compared with the other Sóvári cultivars (cluster IV). The two ‘Beregi sóvári’ genotypes (from UK and Subcarpathia) exhibited no similarity based on the number code description (different subcluster in the main cluster VI), so they are presumably two separate cultivars.

Close relationships could be detected between ‘Nyári fontos’ and ‘Tordai alma’ (cluster III) and between ‘Entz rozmaring’ and ‘Húsvéti rozmaring’ (cluster IV). On the other hand, the ‘Cigány alma’ and ‘Simonffy piros’ cultivars, often considered to be synonyms, were grouped in two distinct clusters (V and VI). The number code description also indicated deviations between the two cultivars for almost 50 % of the traits.

It is interesting to note, however, that cultivars with a great tendency to exhibit parthenocarpy (‘Orbai alma’, ‘Tordai piros kálvil’, ‘Vajki alma’) formed a common cluster (IV) on the dendrogram based on morphology, although there was a great distance between the cultivars.

The genotypes examined included not only cultivars that originated in the Carpathian Basin but also a number of cultivars that are widespread in Hungary but are probably or certainly not of Hungarian origin, such as ‘Kanadai renet’ (‘Canada renet’), ‘Ontario’, Tartós Gusztáv’ (‘Gustavs Dauerapfel’), ‘Londoni pepin’ (‘London pepin’) and ‘Sárga szépvirágú’ (‘Yellow Bellflower’). On the dendrogram shown in Fig. 1, which is based on morphology, endemic cultivars cannot be clearly distinguished from cultivars of foreign origin, with the exception of ‘Kanadai renet’, which formed a separate group. In the work of Gasi et al. (2010), the international cultivars could be clearly distinguished from traditional cultivars on the basis of molecular data, but this was more ambiguous in the case of morphological data. Little information is available on the origin of the cultivars investigated. The site of origin is known or surmised for approx. three fourths of the cultivars. They originate from all over the Carpathian Basin, but the majorities are from Transylvania, while far fewer come from Subcarpathia (e.g. the Sóvári cultivars and ‘Vajki alma’), the Carpathian regions of Serbia (e.g. ‘Szabadkai szercsika’), the Great Hungarian Plain (e.g. ‘Jászvadóka’, ‘Simonffy piros’, ‘Harang alma’) or Transdanubia (‘Herceg Batthyány’) (see Table 1). The dendrogram (Fig. 1) does not reflect the regions, so the morphological similarity is not aligned with the information available on origin.

It is also clear from the dendrogram (Fig. 1) that, with the exception of the Batul and Kanadai renet genotypes, cultivar groups could not be distinguished unambiguously. The distance between the cultivars classified in each subgroup was great, reflecting the great biodiversity of the collection examined.

Relationships within cultivar groups

Batul group

The Batul cultivar group formed a separate main cluster (I) on the dendrogram (Fig. 1). On the basis of morphological data, all the Batul genotypes except ‘Zöld batul’ were found in the same small cluster. The cultivars Batul and ‘Mosolygós batul’ could not be distinguished on the basis of the results. Many authors have drawn attention to the similarity of the cultivars comprising the Batul cultivar group and the difficulties faced when trying to distinguish between them. Bereczki (1882) and Bordeianu et al. (1964) all noted that ‘Mosolygós batul’ resembled ‘Batul’ for a number of traits and could be a seedling of this cultivar.

On the basis of both morphological analysis, ‘Zöld batul’ was placed in a completely different cluster (III), at a great distance from the other Batul genotypes examined. Judging by present morphological and earlier morphological and molecular results and by differences in the flowering time (Király et al. 2012), ‘Zöld batul’ does not exhibit any direct relationship with the ‘Batul’ cultivar and is probably a distant descendant of the basic cultivar. In the future, it should thus be registered as a separate cultivar.

Renet group

The cultivars in this group could be clearly distinguished from each other on the basis of phenotypic traits. The multiplicity of the Renet cultivars is clear from the dendrogram (Fig. 1), where almost all of them were placed in different clusters (II–VI), while even those found in the same main cluster were distant from each other, in separate subclusters.

Pomologists classify the Renet cultivars in the same group on the basis of fruit traits, primarily their taste. Renets are characterised by dense, tender flesh and a characteristic spicy taste (Rapaics 1937). It is one of the largest cultivar groups.

Sóvári group

The Sóvári cultivars are grouped together on the grounds that they appear to be of common origin. Most sources in the literature speak of them as variants of the cultivar ‘Közönséges sóvári’. On the dendrogram compiled using phenotypic data, three members of the Sóvári cultivar group (‘Nemes sóvári’, ‘Zöld sóvári’, ‘Daru sóvári’) formed a small cluster, and Sóvári nobil was placed in the same main cluster (IV), while the Beregi sóvári genotypes were to be found in a completely different main cluster (VI).

Although ‘Nemes sóvári’ and the cultivar ‘Sóvári nobil’, maintained in the English National Fruit Collection, are considered to be synonyms, the results of morphological analysis indicated that they are not the same cultivar, so they cannot be regarded as synonyms. This was confirmed by marker analysis carried out by Király et al. (2012) and by the analysis of the S-genotype, performed by Halász et al. (2011), which revealed different SSR fingerprints and S-allele compositions for the two genotypes.

The complex analysis of Király et al. (2012) showed that the two Beregi sóvári genotypes (from the UK and Subcarpathia) could not be regarded as the same cultivar. The latter was found to be identical with the cultivar described by Máté Bereczki, so the specimen of ‘Beregi sóvári’ maintained in the English National Fruit Collection (Morgan and Richards 1993) should be counted as unknown genotype.

Identification of synonyms

The genotypes ‘Cigány alma’ and ‘Simonffy piros’, which are frequently mentioned as synonyms, proved on the basis of morphological analysis to be clearly distinct from each other. They were also placed in two separate clusters on the dendrogram. In the course of marker analysis, Holler et al. (2012) observed that the genetic fingerprint of ‘Cigány alma’ was identical with that of the ‘Roter Stettiner’ cultivar widely grown in Germany. On the basis of the descriptions published by Lucas (1875) and Bereczki (1887) it is possible that the cultivar in the Soroksár collection, described by Bereczki (1887) as ‘Cigány alma’, may be identical with ‘Roter Stettiner’.

‘Húsvéti rozmaring’ and ‘Entz rozmaring’ are considered by many authors (e.g. Angyal (1926), Brózik and Régius 1957) to be synonyms. The two Rozmaring cultivars due to the slight morphological deviation, either they really are the same cultivar or possibly cultivars originating from bud mutation.

Conclusion

The UPOV number code was used to give a uniform type of objective description for the cultivars in the investigated cultivar groups. It is currently hoped to re-register a number of old cultivars in Hungary, so information on the pomological characteristics of these cultivars will soon be of key importance. Phenotypic analyses were performed to determine the genetic variability of apple cultivars in the Carpathian Basin. Results of hierarchical cluster analysis based on the morphological data confirm a high diversity of the Carpathian apple cultivars. All cultivars except bud mutants could be differentiated from each other with the morphological markers applied. To clarify the similarity or distinctness of Batul bud mutants, further morphological or molecular analysis will be required. The genetic variability of the old cultivars was greater than that of commercial cultivars described on the literature, which demonstrate that very valuable genotypes are to be found among the old cultivars in Hungarian genetic resource collections. Resistant traditional apple germplasm could be used to increasing biodiversity or as resistance gene sources in future breeding programs. In our studies (e.g. Tóth et al. 2013 and a pending post-doc research), we try to select multi-resistant traditional apple cultivars for organic fruit growing.