Introduction

Runner bean (Phaseolus coccineus L.) was introduced to Europe from Central America. It is a perennial species but in Polish climatic conditions (transitional, temperate climate) it shows annual habit of growth because its tuberous root system does not survive cold winters (Łabuda 2010). The tradition of growing runner bean on Polish territory goes back to the sixteenth century when it was initially grown as an ornamental and medicinal plant in monastery gardens. In the twentieth century, for many years Polish market was dominated by one cultivar “Piękny Jaś” but in the less agriculturally advanced southern and south-eastern regions numerous landraces were commonly cultivated by farmers (Łabuda 2010). Over the past 20 years substantial progress was made in runner bean breeding leading to the registration of seven new cultivars in the National Catalogue. Currently in Poland runner and common bean commercial cultivation makes more than half of the legume seed production for human consumption, however, economic importance of the runner bean is lower than that of the common bean. Runner bean is traditionally cultivated for the dry seeds. Consumers in Poland are accustomed to the specific taste of its large seeds. In Polish cuisine dry bean seeds are used for making soup, casseroles, salads, appetizers and variety of hot main course dishes. Climbing forms of runner bean are often grown in home gardens as an ornamental vine plant because of its magnificent inflorescences with white or red flowers.

Phaseolus collection of the National Programme of Plant Genetic Resources located at the National Centre for Plant Genetic Resources at Radzików (Poland), consists of valuable genetic resources for bean breeders and research. It includes nearly 3,000 accessions, of which 152 belong to P. coccineus. The accessions of runner bean are mainly landraces originated in Poland (68%), Ukraine (17%) and Slovakia (10%). Majority of landraces were collected in the south and southeast and to lower extent in central regions of Poland which are main areas for runner bean cultivation. During expeditions conducted in the years 1976–1979 regions of occurrence of old cultivars and landraces were determined. Most of them were located in the southern part of Poland, especially in the Tatra and Beskidy mountains and their forelands (Hanelt and Hammer 1977; Hammer and Hanelt 1979; Kulpa and Hanelt 1981; Hanelt and Schultze-Motel 1979; Kulpa and Jastrzębski 1986; Kulpa and Górski 1986).

Runner bean is a carrier of genes for resistance to many diseases caused by fungi (Ascochyta, Sclerotinia sclerotiorum, Fusarium, Phaeoisariopsis griseola, Colletotrichum) (Schmit and Baudoin 1992; Gilmore et al. 2002; Wallace and Wilkinson 1965; Abawi et al. 1978; Silbernagel and Hannan 1992; Mahuku et al. 2002, 2003), bacteria (Xanthomonas) (Miklas et al. 1994) and virus (Bean golden yellow mosaic virus) (Osorno et al. 2003). There is some evidence on successful experiments transferring the resistance genes from runner bean to common bean (Miklas et al. 1994, 1998, 1999; Wallace and Wilkinson 1965).

The aim of this research was to describe the level and structure of genetic diversity of three landraces of runner bean collected in Poland versus two commercial cultivars in order to assess their genetic potential for breeding and to complement earlier study by Nowosielski et al. (2002) on characterization of genetic diversity of domestic Phaseolus germplasm.

Materials and methods

Materials

Plant material used in this study consisted of five accessions of runner bean from the Polish gene bank collection. Three landraces included in this study were collected in Małopolska Highlands and Beskidy regions located westwards from the collection sites of landraces investigated by Nowosielski et al. (2002). The two commercial cultivars Eureka and Blanka were included in the analysis because they proved to be successfully grown at the time of the collection of landraces and remained in cultivation long afterwards (Table 1). Each accession was represented by 15 plants grown under greenhouse conditions.

Table 1 Descriptions of five runner bean accessions used in the study
Full size table

Molecular analysis

For AFLP (Amplified Fragment Length Polymorphism) analysis total DNA was isolated from leaf tissue of 6-week old individual seedlings using DNeasy Plant Mini Kit (Qiagen), in accordance with the procedure provided by the producer company. For each accession 15 individuals were analysed.

The AFLP™ Plant mapping kit (PE Applied Biosystems) was employed in the AFLP reactions according to the manufacturer’s protocol. The five most informative primer pairs EcoRI-ACA/MseI-CAT, EcoRI-ACC/MseI-CAG, EcoRI-AAG/MseI-CTA, EcoRI-ACG/MseI-CAG, EcoRI-ACG/MseI-CTG out of 64 combinations tested were used for selective PCR. All amplification reactions were accomplished using the Biometra T1 Thermocycler at the speed of temperature change 1°C/s. Electrophoresis was conducted on 36 cm long plates in denaturing polyacrylamide gel (4.5%) for 4 h at 2,400 V. The analysis was performed using fluorescent labeling and detecting technology using ABI PRISM 377 XL device (PE Applied Biosystems).

The length of the analyzed fragments, ranging from 35 to 500 bp, was determined using GENESCAN software with accuracy of 1 bp, and the minimum accepted height of the peak was 100 points. The detected DNA fragments were converted into the zero–one matrix, where 1 = presence and 0 = absence, using the MS Excel 2003 software.

Data analysis

Analysis of data was performed using seven measuring parameters: total number of detected loci (Loc), number of polymorphic loci at the 5% level (Loc_P) i.e., loci with allelic frequencies ranging from 0.05 to 0.95, proportion of polymorphic loci (PLP) at the 5% level expressed as a percentage, number of unique loci (Loc_U), proportion of unique loci (PUL) and number of loci/individual (Loc/n). AFLPsurv (Vekemans et al. 2002) software was used to estimate the number (Loc_P) and proportion (PLP) of polymorphic loci at the P = 0.05, Nei’s gene diversity (Hj); expected heterozygosity under Hardy–Weinberg equilibrium, and Nei’s genetic distances between accessions (after Lynch and Milligan 1994a, b).

Analysis of Molecular Variance (AMOVA) and the Principal Coordinate Analysis (PCoA) were performed, using genetic distance values, by the GenAlex 6.2 software (Peakall and Smouse 2006).

Results and discussion

Genetic diversity within landraces and cultivars

The five primer pair combinations used in this study produced 916 (98.4% polymorphic) scorable PCR products in the whole data set. Out of the total number of analyzed loci, 22.8% were common to all accessions, 14.5% were unique to BES 045, 13.2% to POLKIE99-035, 4.1% to Blanka, 2.5% to BES 003 and 2.1% to Eureka.

The analysis reveled fair amount of genetic variation both in landraces and cultivars (Table 2). The two dwarf Polish commercial cultivars Eureka and Blanka did not differ substantially from the landraces in terms of the number and the proportion of polymorphic loci, the number of loci/plant and the gene diversity parameter Hj. They had, however, much lower proportion of unique loci than two landraces. Modern cultivars, especially those of major cultivated species, display lower level of diversity and lower number of unique loci relative to landraces and natural populations resulting from intensive directional breeding manipulations (Beebe et al. 2001; Metais et al. 2002). Both aforementioned cultivars belong to fairly “old ones” as they were released in 1991/1994 by the same breeding company. While Eureka has been successfully cultivated ever since, Blanka following long period of its commercial use was withdrawn from the National Register of Cultivated Plants in 2008. Their relatively high genetic diversity, comparable to those of landraces, suggests that the breeding process leading to their release was rather moderate and most likely included domestic gene pool of runner bean. Poland has a long tradition of breeding cultivars on the basis of landraces, its origins date back to the second decade of last century (Kulpa and Hanelt 1981). The lowest values of the genetic diversity parameters for the landrace BES 003 are difficult to explain considering scarcity of information on its cultivation, collecting and seed handling in the gene bank. Gene diversity values (Hj) for all accessions fell below those for Italian landraces revealed by RAPD, ISSR, SSR and ET markers by Acampora et al. (2007), Spataro et al. (2011), Escalante et al. (1994). This finding may provide additional support to the hypothesis on evolutionary pathways of runner bean in Europe presented by Spataro et al. (2011). According to that, the European continent could be regarded as a secondary diversification centre for runner bean and the crop was initially introduced to Iberian peninsula and immediately afterwards to Italian peninsula from where it was distributed to the North and East. If this was the case then the migration northwards must have involved adaptation to the new environmental conditions and also selection by the consumers dietary preferences which most likely resulted in the reduction of variability and formation of new gene combinations.

Table 2 Parameters of genetic diversity (Loc-number of scored loci; Loc_P-number of polymorphic loci, PLP-proportion of polymorphic loci; Loc_U-number of unique loci; PUL-proportion of unique loci; Loc/n-number of scored loci/individual, Hj-Nei’s gene diversity)
Full size table

Genetic differentiation of landraces and cultivars

The AMOVA revealed significant variation among and within accessions at P = 0.01 level. Over half of the total variance (59.18%) was located within accessions leaving 40.82% to variability among accessions (Table 3), as it could have been expected for an allogamous species. However, disproportion between these two components was not as pronounced as that found by Acampora et al. (2007) using three marker systems (RAPD, ISSR and ET) in Italian and American accessions.

Table 3 Results of Analysis of Molecular Variance (AMOVA)
Full size table

In PCoA 78.96% of the total variance was accounted by the first three coordinates (49.8, 22.9 and 9.1%, respectively). The first and the second coordinates differentiated the landrace BES 045 form the other accessions. Remaining four accessions were intermixed with extensive overlap of BES 003 with Eureka and POLKIE99-035 with Blanka. The first and the third coordinates separated Eureka from Blanka, BES 003 and from POLKIE99-035 (Fig. 1). It was also shown that none of the accessions except for BES 045 presented distinct genetic makeup.

Fig. 1
figure 1

Principal Coordinate Analysis of AFLP data

Full size image

The distinctive gene pool of BES 045 landrace is ascribed to its genetic background. This landrace is characterized by red flowers and a purple-black seeds whereas the remaining accessions by white flowers and seeds. According to Santalla et al. (2002) red flower forms belong to Phaseolus coccineus var. coccineus, whereas white flower ones to Phaseolus coccineus var. albiflorus. Similar results with regards to genetic dissimilarity between red and white type flower forms were obtained by ISSR technique for nine landraces from central Italy (Sicard et al. 2005).

Nei’s genetic distance values ranged from 0.014 to 0.175 (Table 4). It also did not discriminate commercial cultivars from landraces. The relatively low genetic differentiation expressed by the low values of the genetic distance between accessions (average 0.048) except BES 045 landrace (average 0.167) is in agreement with the results of earlier study of three Polish landraces and two cultivars of runner bean by Nowosielski et al. (2002).

Table 4 Nei’s genetic distance between accessions
Full size table

It could be concluded from the results of this two studies that the genetic variation of P. coccineus cultivated in Poland might be relatively narrow. A low level of polymorphism among European accessions of runner bean was also observed by Hamann et al. (1995), Sicard et al. (2005) and Acampora et al. (2007). Low gene diversity and low Nei’s genetic distance values as well as intergradations among accessions in the PCoA analysis in our study support the hypothesis presented by Spataro et al. (2011).