Mehdi Kakaei; Sayyed Saeed Moosavi
Volume 15, Issue 2 , July 2017, , Pages 21-38
Chickpea contains a high level of protein and plays very important role in the food security cycle in the worldwide, especially in the developing countries such as Iran which is one of the most important Asian countries in terms of chickpea production. Indeed, Iranian chickpea is planted ...
Chickpea contains a high level of protein and plays very important role in the food security cycle in the worldwide, especially in the developing countries such as Iran which is one of the most important Asian countries in terms of chickpea production. Indeed, Iranian chickpea is planted in 33 countries and its area of cultivation in Iran is approximately 650000 hectare (Ahmed Khan, 2009). Low grain yield is the most basic problem of chickpea in Iran and plant breeding is a economic and permanent solution for solving this problem. Genetic diversity, heritability and intensive selection are three effective factors for responding to selection of traits. As a consequence, multivariate statistical methods are conventional and suitable for identifying genetic diversity in plant breeding programmes; principal component analysis, factor analysis, cluster analysis and discriminate function are the most important multivariate methods generally used for different plants. Many researchers have used multivariate statistical methods in rice (Gholipour and Mohammad-Salehi, 2003), potato (Rabie et al., 2008), wheat (Zaki Zadeh et al., 2010), chickpea (Jeena and Arora, 2002), and Identifying the most effective genotypes and traits in chickpea and assessing the genotype diversity were the main objectives of the current research.
Material and methods:
In order to assess genetic diversity and identify the most important and effective traits of grain yield in 19 chickpea genotypes, the current experiment was conducted based on a randomized complete block design with three replications under well water conditions. The experiment was carried out in the Bu-Ali Sina University research field growing season and laboratory of Payame Noor University. The traits of chlorophyll index, plant height, branch number per plant, pod number per plant, seed number per plant, seed number per pod, 100-kernel weight, economic yield per unit area, biological yield per unit area and harvest index were measured. After trait measurement, principal component analysis (PCA), factor analysis (FA), cluster analysis (CA) and discriminate function analysis (DFA) were carried out in order to reach the aims of the research.
Results and discussion:
The results showed that the maximum coefficient of variation belonged to the traits of pod number per plant and harvest index, while the trait of 100-kernel weight had the minimum coefficient of variation. The results of principal component analysis showed that the three first components explained 68.9 percent of the total variance of the traits. The first and second components were known as the “grain yield” and “harvest index” components, respectively. In addition, factor analysis identified the three factors of “grain yield, “harvest index” and “plant vigour”. Cluster analysis based on the WARD method grouped the genotypes into four clusters. Meanwhile, discriminate function analysis confirmed the cluster analysis groups. The results showed a high genetic diversity among the lines. Lines number 12 and 18 were recognized as generally the best and the worst genotypes. In addition, the traits of pod number per plant, 100-kernel weight and harvest index had maximum effect on grain yield, while the trait of branch number per plant had a negative effect on grain yield in the current research.
According to the results obtained in the research, multivariate statistical methods are suitable and efficient methods for data reduction and indirect selection of grain yield that could successfully separate the efficient genotypes and the traits. The main objective was to assess the genetic diversity of 19 chickpea lines for their application in the new plant breeding programmes. According to the total results, the traits of pod number per plant were recognized as the most suitable traits for indirect grain yield selection. Line number 12, as a suitable line, had a maximum amount of the traits of pod number per plant, 100-kernel weight and harvest, while line number 18 (an unsuitable line) included the maximum amount of the trait of branch number per plant.