Utilizing Statistical Techniques for Assessing Genetic Variability and Diversity in Rice Germplasms

The nature of variation in quantitative (phenotypic) traits makes their genetic analysis through the classical Mendelian techniques impossible. Appropriate techniques for their handling were developed by statisticians; these techniques use parameters like mean, variance, covariance, coefficient of variations (CV) etc. Besides, the application of statistics in sampling and data analysis is a very significant and necessary pathway through which proper data after collection can properly be explored. In regards to agriculture especially in evaluating the rice germplasms through univariate and multivariate approaches is challenging and more authentic to apply. In multivariate analysis the system as a whole, having number of variables at a time taking into consideration for their interdependence, relationships etc. In fact, there is a vast area of study in multivariate analysis. However, the main aim of this chapter is to focus on the application of statistical techniques to find the variations among the germplasms and use some of the techniques that can be used in agriculture in clustering the rice germplasm especially to find the variability and superiority of donor for introduction in hybridization and crop improvement programme. Cluster analysis and Principal Component Analysis based on eighteen quantitative agro-morphological traits were studied in fifty-one rice landraces. It was found that in PCA analysis for evaluation of rice landraces, six components with eigen value >1 displayed 75.9% of the variability. Other PCs (7-16) had eigen values less than 1. The contribution of PC1 with eigen value of 3.755 was 23.47% of the total variability. Thus the use of statistical tools and techniques lays a fantastic way of analyzing data for finding the right way of finding the right donor and right trait under demand at mass level.