Assessment of winter wheat advanced lines by use of multivariate statistical analyses

Abstract

This study was conducted to evaluate 49 advanced lines of winter wheat (Triticum aestivum L.) for their morphoagronomic traits and to determine best criteria for selection of lines to be included in future breeding program. The material was assessed in two years experiment at two locations, using RCBD design with three replications. Ten quantitative traits: plant height, number of fertile tillers, spike length, number of spikelets per spike, number of grains per spike, weight of grain per spike and per plant, fertility, biological yield and harvest index were evaluated by PCA and two-way cluster analysis. Three main principal components were determined explaining 71.391% of the total variation among the genotypes. One third of the variation is explained by PC1 which reflects the genotype yield potential. PC2 and PC3 explained 25.22% and 15.49% of the total variance, mostly in relation to the plant height and spike components, respectively. Biplot graph revealed strongest positive association between spike length, number of spikelets and biological yield and between number of tillers, weight of grains per spike and per plant. Two-way cluster analysis resulted with a dendrogram with one solely separated genotype, superior for all traits and two main clusters of genotypes defined with wide genetic diversity especially between the groups within the second cluster. Genotypes with high values for specific traits will be included in the future breeding programmes. Classification of genotypes and the extend of variation among them illustrated on the heatmap has proved to be practical tool for selecting genotypes with desired traits in the early stages of the breeding process.