Comparison among clustering in multivariate analysis of triticale using agronomical traits

H. Vural, R. Dogan

Abstract: This research has been done for determining the most suitable triticale genotypes in the triticale sowing duty system of the Marmara Region ecological conditions. 22 triticale genotypes were used in this two years long study, which has, randomize block experimental design with three replications. Data were analyzed by multivariate statistical methods. Regarding the plant height, the genotype no. 23 recorded the highest value; however, the genotype no. 21 recorded the lowest one. The genotype no. 3 recorded the highest spike length; however, the genotype no. 21 recorded the lowest values respectively. The genotype no. 3 recorded the highest value of grain number spike-1. The genotype no. 4 recorded the highest value of 1000 kernel weight. The genotype no. 21 recorded the highest test weight; however, the genotype no. 7 recorded the lowest values respectively. Regarding the grain yield, the genotype no. 13 recorded the highest value (6723 kg ha-1); however, the genotype no. 16 recorded the lowest one (5244 kg ha-1). According to the two years results, the differences between triticale genotypes were found to be statistically sinificant in all components observed. The differences between the years were proved significant in all components. According to the results of factor analysis, triticale genotypes have similar properties significantly. Therefore, more than one major component has not been found. However, it was found one principal component (PC1) by factorial analyses. In addition, twenty-two examined cultivars were separated in two main groups by cluster analyses.

Keywords: Breeding; factor analysis; triticale; yield components

Date published: 2019-03-29