The use of Cluster Analysis for Best Lines Selection in Maize at S6 Generation

In order to obtain pure lines of corn for direct sowing or their usage in crossing, it is necessary to provide pure genotypes. This experiment was carried out by the use of individual selection breeding method and the 198 S6 lines of maize were evaluated in 2007and 14 traits were measured on them. The results of descriptive statistics showed considerable variation among genotypes. The maximum amount of phenotypic variation among different traits related to lodging resistance, disease resistance and grain yield traits with a value of 41.61, 21.3 and 20.88 percent, respectively. Cluster analysis by Ward’s minimum variance, clustered lines into three groups and the Percentage of grouping accuracy was 94.4 with discrimination function analysis. The first group consists of 23 lines, the second group included 110 line and the third group involved 65 line. The traits mean of lines at the first and second cluster were higher than traits mean of total lines and selected the lines of these two groups and were taken to the next generation that combining ability are evaluated through line × tester analysis. Also the lines of these two groups due to high value of yield and other traits can be used in hybridization programs. Also, factor analysis based on principal components after varimax rotation showed that 65.87 percent of total variations between genotypes were determined by five independent factors. According to the first two factors, the lines were divided into five groups. The result of factor analysis and cluster analysis to classify lines partially matched and confirmed each other.