Nonparametric statistics for evaluating yield of direct seeded rice genotypes in multi-environments.

Genotype (G) x Environment (E) interaction of 12 rice genotypes tested over six environments under direct seeded during optimal transplanting time for rice i.e. 15 th January and 15 th July of each year in dry and wet season respectively. The genotypes were grown in a randomized complete block design with three replications. The objective of this experiment was to identify stable high yielding genotypes. Combined analysis of variance showed genotype, environment and G×E interaction were significant. This indicates possibility of selection of stable genotypes across the environments. Huehn's non-parametric stability measures were used in both original datasets as well as after applying the correction to find out the stable rice genotypes. Based on original data genotype 12 was found stable whereas based on corrected data genotypes 11, 12 and 7 were found to be most stable under direct seeded rice. Principal component analysis was carried out to find out relationships among different stability measures. The first two principal components (PCs) PC1 and PC2 explained 83.58% (54.36 and 29.22 % by PC1 and PC2, respectively) of the total variance.