Evaluation of relationship between grain yield and yield components in bread wheat cultivars using multivariate statistical methods

In order to investigate relationship between yield and its components in bread wheat, tested during 2011-2012 were conducted in Agricultural Research Station of Gonbadin randomized complete block design with three replications whit 10 bread wheat genotype (Morvarid, N-87-4, N-87-9, N-86-6, N-86-8, N-80-19, N-86-5, N-85-5, Kohdasht, Line 17). In this experiment, were measured 7 morphological traits. Analysis of variance genotypes for all traits was significant at 1% probability level. N-85-5 and Line 17 genotypes produced the highest and the lowest grain yield, respectively. Among all tested traits, the number of fertile spike had the highest correlation (0.999 **) with grain yield. Using stepwise regression, the number of fertile spikes entered the model as an effective attribute, justifying 99.8 percent of the data variation. In order to find the causal relationships, the path analysis was performed for grain yield The direct effect on the grain yield was related to the number of fertile spike traits. The results of cluster analysis by ward method showed that the studied cultivars were classified in two groups and the second group in terms of grain yield was the most (397.66). According to the results, can be said that N-85-5 genotype had the highest grain yield and trait number of fertile spike was identified as an effective trait for increasing grain yield.