Utilizing Multivariate Analysis to Enhance Selection Methods for Seed Quality Characteristics in Rice (Oryza sativa L.)
Nimisha Choudhary
Department of Genetics and Plant Breeding, School of Agriculture, Lovely Professional University, Phagwara-144411 (Punjab), India.
Satya Prakash *
Department of Genetics and Plant Breeding, School of Agriculture, Lovely Professional University, Phagwara-144411 (Punjab), India.
Shubhronil Ghosh
Research and Development Cell, Lovely Professional University, Phagwara-144411, (Punjab), India.
*Author to whom correspondence should be addressed.
Abstract
The experiment was carried out at the genetics and plant breeding laboratory, lovely professional university, Jalandhar (Punjab). Forty-five genotypes of rice (Oryza sativa L.) were investigated. Analysis of variance showed significant variation among the genotypes for all seven quantitative traits under evaluation, indicated presence of variation in the populations. The genotypes showed high variability for most traits, reflecting high potential for selection to enhance yield. Genotypes Ram Lakshman, IET-22020, SHIVANTH, DDR-119 had better mean seed vigou index, while Ashoka 200, Ruchi Dhan, PR 131 and HUR-36 showed the maximum germination speed, revealed as early types of maturing and hence the promising lines of choice for any breeding program. The investigation recorded high GCV and PCV seedling dry weight, which was trailed by speed of germination, seedling length and root length, implying large genetic variation for these traits and thus effective direct selection. The traits had high heritability and genetic advance, showing that they are regulated by additive gene effects, making them ideal for selection. Correlation analysis indicated seedling vigour index had positive correlations with standard germination, shoot length, root length and seedling length. The same traits also had maximum direct effects on seed yield at both genotypic and phenotypic levels. Cluster analysis showed that the genotypes are classified in 8 clusters having the maximum intra cluster distance in cluster 8 followed by cluster V, III and IV. Principal component analysis resulted in five principal components, which accounted for 80.11% of the total variation.
Keywords: GCV, PCV, correlation coefficient, cluster analysis, principal component analysis