On-going on-farm microevolutionary processes in neighbouring cowpea landraces revealed by molecular markers

A knowledge of existing levels of diversity is fundamental for planning in situ (on-farm) conservation activities. Three neighbouring cowpea landraces (LRs) currently cultivated in central Italy were studied by amplified fragment length polymorphism (AFLP) and selectively amplified microsatellite polymorphic locus (SAMPL) markers to determine the distribution of genetic variation within and among them. The three LRs studied, although relatively similar, are highly different from one another as shown by the significance of the Fisher exact test for the genic differentiation and the absence of genotype sharing among them. Data obtained from the AFLP and SAMPL markers separately and their combined data revealed a relatively high level of diversity still present within the LRs. The more efficient SAMPL technique was better at discriminating between the plants than the AFLP markers. The three LRs studied appear to be structured as a metapopulation in which a substantial differentiation is maintained at the subpopulation level. A complex interaction of factors (drift, LR isolation, farmer selection, migration within LRs) explains the observed pattern of diversity. The results suggest that the best strategy for maintaining diversity in the area is to preserve each of the LRs observed on the farms from which it came.     

An improved genetic linkage map for cowpea (Vigna unguiculata L.) combining AFLP, RFLP, RAPD, biochemical markers, and biological resistance traits.

An improved genetic linkage map has been constructed for cowpea (Vigna unguiculata L. Walp.) based on the segregation of various molecular markers and biological resistance traits in a population of 94 recombinant inbred lines (RILs) derived from the cross between 'IT84S-2049' and '524B'. A set of 242 molecular markers, mostly amplified fragment length polymorphism (AFLP), linked to 17 biological resistance traits, resistance genes, and resistance gene analogs (RGAs) were scored for segregation within the parental and recombinant inbred lines. These data were used in conjunction with the 181 random amplified polymorphic DNA (RAPD), restriction fragment length polymorphism (RFLP), AFLP, and biochemical markers previously mapped to construct an integrated linkage map for cowpea. The new genetic map of cowpea consists of 11 linkage groups (LGs) spanning a total of 2670 cM, with an average distance of 6.43 cM between markers. Astonishingly, a large, contiguous portion of LG1 that had been undetected in previous mapping work was discovered. This region, spanning about 580 cM, is composed entirely of AFLP markers (54 in total). In addition to the construction of a new map, molecular markers associated with various biological resistance and (or) tolerance traits, resistance genes, and RGAs were also placed on the map, including markers for resistance to Striga gesnerioides races 1 and 3, CPMV, CPSMV, B1CMV, SBMV, Fusarium wilt, and root-knot nematodes. These markers will be useful for the development of tools for marker-assisted selection in cowpea breeding, as well as for subsequent map-based cloning of the various resistance genes.     

A Comparison of AFLP and RAPD Markers for Genetic Diversity and Cultivar Identification in Cotton

AFLP and RAPDmarkers were employed in sixteen diploid cotton (Gossypium sp) cultivars for genetic diversity estimation and cultivar identification. Polymorphism information content (PIC) and percent polymorphism were found to be more for AFLP markers as compared to RAPD markers. Average Jaccard’s genetic similarity index was found to be almost similar using either AFLP or RAPD markers. All the cultivars could be distinguished from one another using AFLP markers and also by the combined RAPD profiles. Cultivar identification indicators like resolving power, marker index and probability of chance identity of two cultivars suggested the usefulness of AFLP markers over the RAPD markers. AFLP and RAPD analyses revealed limited genetic diversity in the studied cultivars. Cluster analysis of both RAPD and AFLP data produced two clusters, one containing cultivars of G. herbaceum and another containing cultivars of G. arboreum species. Highly positive correlation between cophenetic matrices using RAPD and AFLP markers was observed. AFLP markers were found to be more efficient for genetic diversity estimation, polymorphism detection and cultivar identification.     

Intraspecific diversity and relationship between subspecies of Origanum vulgare revealed by comparative AFLP and SAMPL marker analysis

The genus Origanum is often referred to as an under-utilized taxon because of its complex taxonomy. Origanum vulgare L., the most variable species of the genus, is a spice and medicinal herb that is characterized by high morphological diversity (six subspecies). In this study, the relative efficiencies of two PCR-based marker approaches, amplified fragment length polymorphism (AFLP) and selectively amplified microsatellite polymorphic loci (SAMPL), were used for comparable genetic diversity surveys and subspecies discrimination among 42 oregano accessions. Seven assays each of AFLP and SAMPL markers were utilized. Effective multiplex ratio (EMR), average heterozygosity (H_av-p), marker index (MI), and resolving power (RP) of the primer combinations were calculated for the two marker systems. UPGMA and Structure analysis along with PCoA plots derived from the binary data matrices of the two markers depicted the genetic distinction of accessions. Our results indicate that both marker systems are suitable but SAMPL markers are slightly more efficient in differentiating accessions and subspecies than AFLPs.     

Development of S-SAP markers based on an LTR-like sequence from Medicago sativa L.

The Sequence-Specific Amplification Polymorphism (S-SAP) method, recently derived from the Amplified Fragment Length Polymorphism (AFLP) technique, produces amplified fragments containing a retrotransposon LTR sequence at one end and a host restriction site at the other. We report the application of this procedure to the LTR of the Tms1 element from Medicago sativa L. Genomic dot-blot analysis indicated that Tms1 LTRs represent about 0.056% of the M. sativa genome, corresponding to 16 x 10(3) copies per haploid genome. An average of 66 markers were amplified for each primer combination. Overall 49 polymorphic fragments were reliably scored and mapped in a F(1) population obtained by crossing diploid M. falcata with M. coerulea. The utility of the LTR S-SAP markers was higher than that of AFLP or SAMPL (Selective Amplification of Microsatellite Polymorphic Loci) markers. The efficiency index of the LTR S-SAP assay was 28.3, whereas the corresponding values for AFLP and SAMPL markers were 21.1 and 16.7, respectively. The marker index for S-SAP was 13.1, compared to 8.8 for AFLP and 9.5 for SAMPL. Application of the Tms1 LTR-based S-SAP to double-stranded cDNA resulted in a complex banding pattern, demonstrating the presence of Tms1 LTRs within exons. As the technique was successfully applied to other species of the genus Medicago, it should prove suitable for studying genetic diversity within, and relatedness between, alfalfa species.     

A SNP and SSR based genetic map of asparagus bean (Vigna. unguiculata ssp. sesquipedialis) and comparison with the broader species

Asparagus bean (Vigna. unguiculata ssp. sesquipedialis) is a distinctive subspecies of cowpea [Vigna. unguiculata (L.) Walp.] that apparently originated in East Asia and is characterized by extremely long and thin pods and an aggressive climbing growth habit. The crop is widely cultivated throughout Asia for the production of immature pods known as 'long beans' or 'asparagus beans'. While the genome of cowpea ssp. unguiculata has been characterized recently by high-density genetic mapping and partial sequencing, little is known about the genome of asparagus bean. We report here the first genetic map of asparagus bean based on SNP and SSR markers. The current map consists of 375 loci mapped onto 11 linkage groups (LGs), with 191 loci detected by SNP markers and 184 loci by SSR markers. The overall map length is 745 cM, with an average marker distance of 1.98 cM. There are four high marker-density blocks distributed on three LGs and three regions of segregation distortion (SDRs) identified on two other LGs, two of which co-locate in chromosomal regions syntenic to SDRs in soybean. Synteny between asparagus bean and the model legume Lotus. japonica was also established. This work provides the basis for mapping and functional analysis of genes/QTLs of particular interest in asparagus bean, as well as for comparative genomics study of cowpea at the subspecies level.     

A genetic linkage map of cowpea (Vigna unguiculata) developed from a cross between two inbred, domesticated lines

 We have constructed a genetic linkage map within the cultivated gene pool of cowpea (2n=2x=22) from an F₈ recombinant inbred population (94 individuals) derived from a cross between the inbreds IT84S-2049 and 524B. These breeding lines, developed in Nigeria and California, show contrasting reactions against several pests and diseases and differ in several morphological traits. Parental lines were screened with 332 random RAPD decamers, 74 RFLP probes (bean, cowpea and mung bean genomic DNA clones), and 17 AFLP primer combinations. RAPD primers were twice as efficient as AFLP primers and RFLP probes in detecting polymorphisms in this cross. The map consists of 181 loci, comprising 133 RAPDs, 19 RFLPs, 25 AFLPs, three morphological/classical markers, and a biochemical marker (dehydrin). These markers identified 12 linkage groups spanning 972 cM with an average distance of 6.4 cM between markers. Linkage groups ranged from 3 to 257 cM in length and included from 2 to 41 markers, respectively. A gene for earliness was mapped on linkage group 2. Seed weight showed a significant association with a RAPD marker on linkage group 5. This map should facilitate the identification of markers that “tag” genes for pest and disease resistance and other traits in the cultivated gene pool of cowpea. Received: 16 September 1996 / Accepted: 25 April 1997     

Genetic differentiation analysis of African cassava (Manihot esculenta) landraces and elite germplasm using amplified fragment length polymorphism and simple sequence repeat markers

Molecular‐marker‐aided evaluation of germplasm plays an important role in defining the genetic diversity of plant genotypes for genetic and population improvement studies. A collection of African cassava landraces and elite cultivars was analysed for genetic diversity using 20 amplified fragment length polymorphic (AFLP) DNA primer combinations and 50 simple sequence repeat (SSR) markers. Within‐population diversity estimates obtained with both markers were correlated, showing little variation in their fixation index. The amount of within‐population variation was higher for landraces as illustrated by both markers, allowing discrimination among accessions along their geographical origins, with some overlap indicating the pattern of germplasm movement between countries. Elite cultivars were grouped in most cases in agreement with their pedigree and showed a narrow genetic variation. Both SSR and AFLP markers showed some similarity in results for the landraces, although SSR provided better genetic differentiation estimates. Genetic differentiation (Fst) in the landrace population was 0.746 for SSR and 0.656 for AFLP. The molecular variance among cultivars in both populations accounted for up to 83% of the overall variation, while 17% was found within populations. Gene diversity (H_e) estimated within each population varied with an average value of 0.607 for the landraces and 0.594 for the elite lines. Analyses of SSR data using ordination techniques identified additional cluster groups not detected by AFLP and also captured maximum variation within and between both populations. Our results indicate the importance of SSR and AFLP as efficient markers for the analysis of genetic diversity and population structure in cassava. Genetic differentiation analysis of the evaluated populations provides high prospects for identifying diverse parental combinations for the development of segregating populations for genetic studies and the introgression of desirable genes from diverse sources into the existing genetic base.     

Comparative assessment of DNA fingerprinting techniques (RAPD, ISSR and AFLP) for genetic analysis of cashew (Anacardium occidentale L.) accessions of India.

Nineteen cashew accessions were analysed with 50 random primers, 12 ISSR primers and 6 AFLP primer pairs to compare the efficiency and utility of these techniques for detecting variation in cashew germplasm. Each marker system could discriminate between all of the accessions, albeit with varied efficiency of polymorphism detection. AFLP exhibited maximum discrimination efficiency with a genotype index of 1. The utility of each molecular marker technique, expressed as marker index, was estimated as a function of average band informativeness and effective multiplex ratio. Marker index was calculated to be more than 10 times higher in AFLP than in RAPD and ISSR. Similarity matrices were determined based on the data generated by molecular and morphometric analyses, and compared for congruency. AFLP displayed no correspondence with RAPD and ISSR. Correlation between ISSR and RAPD similarity matrices was low but significant (r = 0.63; p < 0.005). The similarity matrix based on morphometric markers exhibited no correlation with any of the molecular markers. AFLP, with its superior marker utility, was concluded to be the marker of choice for cashew genetic analysis.     

Identification and Characterization of Malaysian River Catfish, Mystus nemurus (C&V): RAPD and AFLP Analysis

This work represents the first application of the amplified fragment length polymorphism (AFLP) technique and the random amplified polymorphic DNA (RAPD) technique in the study of genetic variation within and among five geographical populations of M. nemurus. Four AFLP primer combinations and nine RAPD primers detected a total of 158 and 42 polymorphic markers, respectively. The results of AFLP and RAPD analysis provide similar conclusions as far as the population clustering analysis is concerned. The Sarawak population, which is located on Borneo Island, clustered by itself and was thus isolated from the rest of the populations located in Peninsular Malaysia. Both marker systems revealed high genetic variability within the Universiti Putra Malaysia (UPM) and Sarawak populations. Three subgroups each from the Kedah, Perak, and Sarawak populations were detected by AFLP but not by RAPD. Unique AFLP fingerprints were also observed in some unusual genotypes sampled in Sarawak. This indicates that AFLP may be a more efficient marker system than RAPD for identifying genotypes within populations.     

Microsatellite-AFLP for genetic mapping of complex polyploids.

In spite of the economical relevance of polyploid crops, genetic mapping of these species has been relatively overlooked. This is because of intrinsic difficulties such as the uncertainty of the chromosome behavior at meiosis I and the need for very large segregating populations. An important, yet underestimated issue, in mapping polyploids is the choice of the molecular marker system. An ideal molecular marker system for polyploid mapping should maximize the percentage of single dose markers (SDMs) detected and the possibility of recognizing allelic markers. In the present work, the marker index for genetic mapping (MIgm) of M-AFLP is compared with that of AFLP and SAMPL. M-AFLPs have the highest MIgm values (22 vs. 18.5 of SAMPL and 9.83 of AFLP) mostly because of their high power to detect polymorphism. Owing to their prevalent codominant inheritance, it is proposed that M-AFLP can be used for the preliminary identification of hom(e)ologous groups.     

AFLP markers linked to resistance against Striga gesnerioides race 1 in cowpea (Vigna unguiculata).

Amplified fragment length polymorphism (AFLP) analysis was used in combination with bulked segregant analysis (BSA) to identify molecular markers linked to two cowpea (Vigna unguiculata (L.) Walp.) genes conferring resistance to Striga gesnerioides race 1. After AFLP analysis of an F2 population derived from a cross between the resistant cultivar Gorom and the susceptible cultivar Tvx 3236, seven AFLP markers were identified that are linked to Rsg3, the gene conferring race I resistance in 'Gorom'. The distances between these markers and Rsg3 ranged from 9.9 to 2.5 cM, with two markers, E-AGA/M-CTA460 and E-AGA/M-CAG300, flanking Rsg3 at 2.5 and 2.6 cM, respectively. Analysis of a second F2 population derived from the cross between 'Tvx 3236' and the resistant cultivar IT81D-994 identified five AFLP markers linked to the race 1 resistance gene 994-Rsg present in 'IT81D-994'. The two markers showing the tightest linkage to the 994-Rsg locus were E-AAG/M-AAC450 and E-AAG/M-AAC150 at 2.1 and 2.0 cM, respectively. Two of the markers linked to 994-Rsg, E-AGA/M-CAG300 and E-AGA/M-CAG450, were also linked to Rsg3. The identification of molecular markers in common between the two sources of race 1 resistance suggests that either Striga resistance genes are clustered in these plants or that these loci are allelic. Mapping of the resistance loci within the cowpea genome revealed that three markers linked to Rsg3 and (or) 994-Rsg are located on linkage group 6.     

分子标记及其在核桃遗传多样性研究中的应用

核桃是我国重要的坚果和木本油料树种之一,具有重要的学术研究和经济价值。现代分子标记技术的迅速发展为核桃的种质鉴定、遗传育种、遗传多样性分析、亲缘鉴定等提供了崭新途径。本文主要介绍RFLP、RAPD、AFLP及SSR等几种分子标记技术的主要原理、特点以及在核桃遗传多样性方面的研究进展,分析了分子标记在核桃遗传多样性研究中的主要问题,并对其发展提出了展望。     

AFLP和RAPD标记技术在栉孔扇贝遗传多样性研究中的应用比较

AFLP和RAPD标记技术是近年来发展最快的基于PCR基础上的两种DNA标记技术,本文比较了两种标记技术在我国栉孔扇贝群体遗传多样性研究中的应用。共筛选20个RAPD引物和7个AFLP引物组合,检测到AFLP标记的有效等位基因数和平均多态信息量稍低于RAPD标记,但AFLP标记在每单位分析中扩增到的野生和养殖群体的多态性条带数(23．8,24．8)分别高于RAPD标记(5．6,5．6),AFLP多态性检测效率显著高于RAPD标记。AFLP和RAPD两种标记技术所揭示的野生种群与养殖群体间的近交系数、遗传距离两项指标均表明,我国栉孔扇贝养殖群体和野生种群之间尚未出现明显的遗传分化。研究结果表明：RAPD和AFLP这两种标记技术均可用于栉孔扇贝遗传多样性的分析,其分析结果是一致的。     

Analysis of genetic diversity through AFLP,SAMPL, ISSR and RAPD markers in <Emphasis Type="Italic">Tribulus terrestris</Emphasis>, a medicinal herb

Tribulus terrestris is well known for its medicinal importance in curing urino-genital disorders. Amplified fragment length polymorphism (AFLP), selective amplification of microsatellite polymorphic loci (SAMPL), inter-simple sequence repeat (ISSR) and randomly amplified polymorphic DNA (RAPD) markers were used for the first time for the detection of genetic polymorphism in this medicinal herb from samples collected from various geographical regions of India. Six assays each of AFLP and SAMPL markers and 21 each of ISSR and RAPD markers were utilized. AFLP yielded 500 scorable amplified products, of which 82.9% were polymorphic. SAMPL primers amplified 488 bands, 462 being polymorphic (94.7%). The range of amplified bands was 66 [(TC)₈G + M-CAG] to 98 [(CA)₆AG + M-CAC] and the percentage polymorphism, 89.9 [from (CT)₄C (AC)₄A + M-CTG] to 100 [from (GACA)₄ + M-CTA]. The ISSR primers amplified 239 bands of 0.4–2.5 kb, 73.6% showed polymorphism. The amplified products ranged from 5 to 16 and the percentage polymorphism 40–100. RAPD assays produced 276 bands, of which 163 were polymorphic (59%). Mantel test employed for detection of goodness of fit established cophenetic correlation values above 0.9 for all the four marker systems. The dendrograms and PCA plots derived from the binary data matrices of the four marker systems are highly concordant. High bootstrap values were obtained at major nodes of phenograms through WINBOOT software. The relative efficiency of the four molecular marker systems calculated on the basis of multiplex ratio, marker index and average heterozygosity revealed SAMPL to be the best. Distinct DNA fingerprinting profile, unique to every geographical region could be obtained with all the four molecular marker systems. Clustering can be a good indicator for clear separation of genotypes from different regions in well-defined groups that are supported by high bootstrap values.     

Genetic relationship among wild,landraces and cultivars of hazelnut (Corylus avellana) from Portugal revealed through ISSR and AFLP markers

The genus Corylus, a member of the birch family Betulaceae, includes several species that are widely distributed throughout temperate regions of the Northern Hemisphere. This study assesses the genetic diversity in 26 international cultivars and 32 accessions of Corylus avellana L. from Portugal: 13 wild genotypes and 19 landraces. The genetic relationships among the 58 hazelnuts (Corylus avellana L.) were analyzed using inter simple sequence repeat (ISSR) and amplified fragment length polymorphism (AFLP) markers. Eighteen ISSR primers and seven AFLP primer pairs generated a total of 570 unambiguous and repeatable bands, respectively, from which 541 (95.03 %) were polymorphic for both markers. Genetic similarity index values ranged from 0.239 for wild types and cultivars to 0.143 for landraces and wild types. The genetic relationships were presented as a Neighbor-Joining method dendrogram and a two-dimensional principal coordinate analysis (PCoA) plot. The Neighbor-Joining dendrogram showed three main clusters, and the PCoA analysis has shown to be congruent with the hierarchical analysis. Bayesian analysis clustered all individuals into three groups showing a good separation among wild genotypes, landraces and cultivars. The genetic diversity found on wild genotypes and Portuguese landraces may provide relevant information for the diversity conservation and it will be useful in breeding programs and to identify local selections for preservation.     

Identification, genetic localization, and allelic diversity of selectively amplified microsatellite polymorphic loci in lettuce and wild relatives (Lactuca spp.).

Selectively amplified microsatellite polymorphic locus (SAMPL) analysis is a method of amplifying microsatellite loci using generic PCR primers. SAMPL analysis uses one AFLP primer in combination with a primer complementary to microsatellite sequences. SAMPL primers based on compound microsatellite sequences provided the clearest amplification patterns. We explored the potential of SAMPL analysis in lettuce to detect PCR-based codominant microsatellite markers. Fifty-eight SAMPLs were identified and placed on the genetic map. Seventeen were codominant. SAMPLs were dispersed with RFLP markers on 11 of the 12 main linkage groups in lettuce, indicating that they have a similar genomic distribution. Some but not all fragments amplified by SAMPL analysis were confirmed to contain microsatellite sequences by Southern hybridization. Forty-five cultivars of lettuce and five wild species of Lactuca were analyzed to determine the allelic diversity for codominant SAMPLs. From 3 to 11 putative alleles were found for each SAMPL; 2-6 alleles were found within Lactuca sativa and 1-3 alleles were found among the crisphead genotypes, the most genetically homogeneous plant type of L. sativa. This allelic diversity is greater than that found for RFLP markers. Numerous new alleles were observed in the wild species; however, there were frequent null alleles. Therefore, SAMPL analysis is more applicable to intraspecific than to interspecific comparisons. A phenetic analysis based on SAMPLs resulted in a dendrogram similar to those based on RFLP and AFLP markers.     

Optimization of bulked AFLP analysis and its application for exploring diversity of natural and cultivated populations of red clover.

Landraces and wild populations of red clover (Trifolium pratense L.) may represent a significant yet poorly characterized genetic resource of temperate grasslands. A bulking strategy with amplified fragment length polymorphism (AFLP) markers was optimized to characterize 120 red clover populations in 6 different groups: Swiss wild clover populations, Mattenklee landraces, Mattenklee cultivars, field clover cultivars, Dutch wild clover populations, and Dutch landraces. Analysis of 2 bulked samples/population consisting of 20 plants each with12 AFLP primer combinations was found optimal for determining genetic diversity and relationships within and among red clover populations and groups. Swiss wild clover populations were clearly separated from all other red clover groups and variability within and among populations was shown to be particularly high in wild clover populations and Mattenklee landraces, emphasising their value as genetic resources for improvement of red clover cultivars, as well as for conservation and restoration of biodiversity. This study shows that the ancestry of red clover landraces is primarily found in introduced cultivars rather than in natural wild clover populations. In addition, the methodological considerations presented here may help improve diversity analyses using bulked samples.     

Genetic relationship of cowpea (Vigna unguiculata) varieties from Senegal based on SSR markers

Genetic diversity and phylogenetic relationships among 22 local cowpea (Vigna unguiculata) varieties and inbred lines collected throughout Senegal were evaluated using simple sequence repeat molecular markers. A set of 49 primer combinations were developed from cowpea genomic/expressed sequence tags and evaluated for their ability to detect polymorphisms among the various cowpea genotypes. Forty-four primer combinations detected polymorphisms, with the remaining five primer sets failing to yield PCR amplification products. From one to 16 alleles were found among the informative primer combinations; their frequencies ranged from 0.60 to 0.95 (mean = 0.79). The genetic diversity of the sample varied from 0.08 to 0.42 (mean = 0.28). The polymorphic information content ranged from 0.08 to 0.33 (mean = 0.23). The local varieties clustered in the same group, except 53-3, 58-53, and 58-57; while Ndoute yellow pods, Ndoute violet pods and Baye Ngagne were in the second group. The photosensitive varieties (Ndoute yellow pods and Ndoute violet pods) were closely clustered in the second group and so were inbred line Mouride and local cultivar 58-57, which is also one of the parents for inbred line Mouride. These molecular markers could be used for selection and identification of elite varieties for cowpea improvement and germplasm management in Senegal.     

Genetic diversity in cultivars and landraces of Oryza sativa subsp. indica as revealed by AFLP markers.

Genetic diversity among 49 Indian accessions of rice (Oryza sativa subsp. indica), including 29 landraces from Jeypore, 12 modern cultivars, and 8 traditional cultivars from Tamil Nadu, was investigated using AFLP markers. In total, nine primer combinations revealed 664 AFLPs, 408 of which were found to be polymorphic. The percentage of polymorphic AFLPs was approximately the same within the cultivars and landraces. Similar results were obtained when genetic diversity values were estimated using the Shannon-Weiner index of diversity. Genetic diversity was slightly higher in the modern cultivars than in the traditional cultivars from Tamil Nadu. Among the landraces from Jeypore, the lowland landraces showed the highest diversity. The present study showed that the process of breeding modern cultivars did not appear to cause significant genetic erosion in rice. Cluster analysis and the first component of principle component analysis (PCA) both showed a clear demarcation between the cultivars and landraces as separate groups, although the genetic distance between them was narrow. The modern cultivars were positioned between the landraces from Jeypore and the traditional cultivars from Tamil Nadu. The second component of PCA further separated medium and upland landraces from lowland landraces, with the lowland landraces found closest to the traditional and modern cultivars.