Study of diversity in a set of lentil RILs using morphological and molecular markers

Abstract

We propose the use of SRAP and SSR markers along agronomic or morphological traits in genotypes’ differentiation during breeding programs. 25 F5 recombinant inbred lines (microsperma and macrosperma) with good quality characteristics and high yield and a tester were used. All materials were sowed in a randomized block design with 3 replications, at a density of 140 plants per m2 . Different morphological traits were recorded and analyzed using ANOVA. Euclidean distances were calculated and a cluster analysis and a principal components analysis (PCA) were carried out. A total of 20 out of 25 SRAP combinations and 9 out of 17 SSR were assayed on the basis of their high polymorphism. Genetic distances were calculated according to Dice’s similarity index and a dendrogram was performed. A comparison between morphological and molecular data was carried out through the procrustes generalized analysis. The analysis of variance for morphological traits showed significant differences between genotypes for all variables. The analysis of principal components determined that two PCs explained 75% of the total variation observed among the genotypes and the cluster analysis allowed the conformation of two principal groups. The first one included four varieties belonging to the microsperma type, whereas the second group included all the macrosperma type Rils. Both groups showed differences also in yield and plant height. Molecular analysis showed that 66% out of 273 bands are polymorphic fragments from SRAP markers, indicating a high genetic diversity among Rils. The number of alleles per SSR locus ranged from 3 (SSR 33) to 7 (SSR 156) showing also the variability present. Two exclusive SRAP fragments generated by the combination me5-em1 (230 bp) were found in microsperma group allowing early selection of this trait and all characters associated with it. The correlation between Dice similarity index and Euclidean distance matrices was 87% indicating an excellent consistency between both types of markers and suggested that both types of traits would provide similar estimates on the variability and would be useful to differentiate Rils with similar origin