A COMPARATIVE STUDY OF YIELD STABILITY ESTIMATES FOR SELECTION OF IDEAL COTTON CULTIVAR IN MULTI ENVIRONMENT TRIALS BY VARIOUS BIOMETRIC METHODS
High and stable yield under the variable farming conditions is required for cotton cultivars to become commercially successful. Thus testing of their yield performance under a wide range of cropping environments is required. MET (Multi environment trial) data analysis targets to predict future cultivar performance. The use of efficient and practical statistical methods is required in order to select the Ideal cultivar that could combine, stable and high yield. The evaluation of stability methods for selection of the Ideal variety was the aim of this research. One data set of cotton MET data used for stability and performance evaluation over 18 testing environments for two years, 2002 and 2003, and 9 locations in Greece. The multivariate newly released GGE Bioplot model based in principal components analysis (Weikai Yan, 2001) compared against four conventional methods for selection of ideal variety. Along to GGE Bioplot were used, deviation from regression line sd2 and determination index R2 (Eberhart and Russell, 1966), Shuklas stability index (Shukla, 1972), absolute CVs and real CVs, among environments. Significant correlation was found between Shuklass index and GGE bioplot. Significant also is the correlation for Absolute CV, deviation from regression line sd2 and determination index R2. Real CVs isnt correlated with any of the indexes. The GGE model is also effective in selection of cultivars exploiting specific adaptability and for grouping mega-environments.
Yield stability Multi environment trials genotype z environment interaction bioplot mega-environments cotton
D.Baxevanos S.Tzortzios
University of Thessaly, School of Agricultural Sciences, Lab.of Biometry.Fytokou Str., N.Ionia.38446 Volos.Greece
国际会议
北京
英文
333-337
2005-10-14（万方平台首次上网日期，不代表论文的发表时间）