Evaluation of Grain Yield and Yield Components using Descriptive and Multivariate Statistics

In order to identify the most important characteristics and affecting on grain yield in the first year (2012) of 10 maternal lines were crossed with male parents MO17. In the second year (2013) 11 parental lines and 10 F1 hybrids were studied in a randomized complete block design with three replications at Mahidasht research station of Kermanshah. Measured traits such as stem diameter, plant height and ear, leaf area, number of kernels per row, number of kernel rows per ear, moisture content, 500-seed weight, seed weight per 5 cob, cob percent, the net weight of the grain. Results of statistical analysis showed that the mean and mod of 13 characters per lines due to the inbreeding depuration were very low and in the hybrid were high. 36 and 15 correlation, of 156 correlations was significant at the 1% and 5% levels respectively. The remaining six characters, number of rows and number of kernels per row, the most direct effect on grain yield with 0.362 and 0.242 respectively. The most total effect accrued to the number of rows of seeds, plant height, and number of kernels per row and grain weight, with 0.807, 0.736, 0.677 and 0.644 respectively. Cluster analysis showed that 21 genotypes were evaluated in 4 different clusters. Lines 1, 5, 7, 8, 9, 3, 6, compounds 13, 16, 20, 17, 18 and 21 were assigned to two different and far apart groups.