REGULAR ARTICLE
Stability and adaptability analysis of rice cultivars using environment-centered yield in two-way ANOVA model

Markandu Anputhas, Sembakutti Samita, D. Sumith De. Z. Abeysiriwardena

Commun. Biometry Crop Sci. (2011) 6 (2), 80-86.
 

ABSTRACT
Identification of rice varieties with wider adaptability and stability are the important aspects in varietal recommendation to achieve better economic benefits for farmers. Multi locational trails are conducted in different locations / seasons to test and identify the consistently performing varieties in wider environments and location specific high performing varieties. The interaction aspect of varieties with environment is complex and highly variable across locations. Thus, the identifying varieties under these circumstances are difficult for varietal recommendations. However, several methods have been proposed in the recent past with the complex computation requirements. But, the aid of statistical software and other programs capabilities ease the complexity to a large extent. In this study, we employed one of the established techniques called variance component analysis (VCA) to make the varietal recommendation for wider adaptability for many varying environments and the location specific recommendations. In this method variety × environment interaction is portioned into components for individual varieties using yield deviation approach. The average effect of variety (environment centered yield deviation - D_k) and the stability measure of each variety (variety interaction variance - s_k²) are used make the recommendations. The rice yield data of cultivars of three month maturity duration, cultivated across diverse environments during the 2002/03 wet-season in Sri Lanka was analyzed for making recommendations. Based on the results the variety At581 gave the highest D_k value with wide adaptability selected for general recommendation. Varieties Bg305 and At303 also had relatively higher D_k and thus these two can also be selected for general cultivation purpose.