Classification of rapeseed genotypes based on quantitative traits using multivariate techniques

The genetic variability of 13 rapeseed genotypes created in Serbia was assessed at the Institute of Field and Vegetable Crops, Novi Sad, in three consecutive years. The genotypes were evaluated and classified based on 10 quantitative traits (plant height, height to the first branch, stem width, number of lateral branches, number of leaves, number of pods, seed yield per plant, 1000 seed weight, oil content and protein content), using principal component analysis (PCA) and two-way cluster analysis. The first four PCs with eigenvalues >1 contributed to 83% of the total explained variance. The PC analysis revealed that plant height, height to the first branch, number of lateral branches and number of leaves contributed maximum to the total divergence of the material. The genotypes were classified in two main clusters: the first cluster comprised three genotypes, while the other genotypes belonged to the second cluster which can further be divided to two subclusters. One genotype was not classified in these two clusters. The selected genotypes that have desirable traits may be used in hybridization programs for improvement of seed yield and traits that contribute to seed yield most.