Assessing variability and detecting the effective traits on yield of chickpea (Cicer arietinum L.) promising lines under drought stress condition

Introduction
Chickpea is one of the most important pulses in the world and it plays a key role in feeding of people in Iran. Iran is one of the original centers of chickpea. This plant has a significant genetic diversity and has favorable conditions for breeding and introducing new cultivars in Iran. According to FAO reports (2014), the average grain yield of chickpea in Iran with a cultivated area of 565 thousand hectares and an average yield of 557 kg per hectare is very low compared to the average of the major countries producing chickpeas. In order to solve this problem, identification of the initial variety of lines and cultivars to start the program is of particular importance. In fact, three factors of heredity, high diversity and severity of selection are effective factors in increasing the response to selection. In this regard, there are various methods for estimating genetic variation in plant species, one of the most important of these methods is multivariate analysis. It is necessary to use these methods for the classification of germplasm and the analysis of the genetic relationships existing between the modifying materials. Among the various methods of multivariate analysis, the principal component analysis, factor analysis, cluster analysis and decomposition function analysis are among the most important of these methods. In fact, one of the important goals of an outbreak is to select the best genotypes. In order to achieve this goal, the studied population should have a significant variation in terms of the characteristics studied, which knowledge of this diversity requires evaluation of the germplasm.
 
Materials and Methods
This study included 20 chickpea (Cicer arietinum L.) promising lines were planted in a randomized complete block design with three replications in 2013 growing season. To evaluate the diversity of lines based on important agro-morphological traits and to achieve the desired goals, various statistical methods including principal component analysis, factor analysis, cluster analysis and decomposition function analysis were used. Genotypic coefficient of variance (GCV) and phenotype coefficient of variance (PCV) using genotype variance and phenotype variance, respectively, were calculated. All calculations and statistical analyzes such as component analysis, cluster analysis, and decomposition function analysis were performed using SPSS and Minitab software.
 
Results and Discussion
The lowest and the highest coefficient of genotypic changes were related to day physiology (2.66%) and 100-seed weight (49.76%). The results showed that the three main first components explained 76% of the total variance of the traits totally. The first and second components were named as "component of grain yield increase" and "vegetative growth component". The results of factor analysis were consistent with the results of the principal component analysis. In order to cluster the studied lines and for their grouping based on the studied traits, Ward method and Pearson’s square distance matrix were used. According to the cluster analysis, the genotypes were located in three distinct groups, so that the second and third clusters had the highest genetic distance and they were recognized as “yield clusters" and "vegetative growth cluster", respectively. Therefore, crossing between the two above mentioned-groups, will result to high artificial diversity in the future generations. According to the cluster analysis, the lines were divided into three separate clusters. Discriminant function analysis confirmed the cluster grouping completely. In a study, nineteen chickpea lines and five genotypes of wild chickpea (Cicer veticulatum) were classified into three distinct groups using cluster analysis, so that wild-type genotypes in one group and two lines of cropping lines were grouped separately from other lines. Ganjali et al, (2009), applied the multivariate bi-plot analysis to evaluate the variation in chickpea germplasm for drought resistance.
 
Conclusion
There was a remarkable genetic diversity for the current germplasm, therefore existence a high level of genetic diversity and a remarkable GVC for the trait of 100-seed weight indicating that this trait can be used as a suitable character for yield improvement in the germplasm and under this experiment conditions. Finally, selection for high levels of biomass, pod number per plant and seed number per plant, and a low amount of main branches number, will increase the grain yield in the future breeding program.