Study of grain yield stability in rice (Oryza sativa L.) promising genotypes

In order to determine the adaptation and stability of grain yield in 8 rice promising genotypes as well as two commercial cultivars (Khazar and Neamat), as control, a field experiment was carried out in three locations in Gilan province during 1997 and 1998 cropping seasons. The experimental design was randomized complete block with 4 replications and planting distance was 25 cm × 25 cm. Analysis of variance showed significant differences for grain yield among genotypes. Combined ANOVA was also performed following Bartlett test (for uniformity of error variances). Year and location effects were considered random while genotypic effect fixed. The F test was calculated based on expected mean squares. Results showed significant differences among the genotypes. The effects of location and year and genotype location and genotype year interactions were not significant. It means that the genotypes had similar response over different locations and years. However, the interaction effects of year location, and year location genotype were significant. Stability of genotypes were evaluated using six different methods including: variance (S2i) and coefficient of variability (CVi)of environment (parameter no.1), coefficient of regression of mean yield on environmental index (bi) (parameter no.2), variance of deviation from regression (S2di) (Parameter no.3) and within location variance (MSy/p) and C.V. (parameter no.4). Results of these methods were different more or less; however, genotypes no.7604 and 7606 as well as Khazar and Neamat were determined as stable genotypes, because low S2i, CVi and MSy/p and bi equal to unity and low and non- significant S2di. Consequently, due to higher mean yield compared to Khazar and the desired cooking and milling qualities, these two genotypes were selected to be released.