Comparison of Parametric and Non-parametric Methods for Analysing Genotype× Environment Interactions in Sunflower (Helianthus annuus L.) Inbred Lines

This study was conducted to measure the grain yield stability of 15 sunflower (Helianthus annuus L.) inbred lines at Dry-land Agricultural Research Institute, using randomized complete block design with three replications on five environments during three consecutive year (2009-11). In this study, parametric and nonparametric stability methods were used. The genotype × environment interaction was significant (P 0.01) using the Hildebrand (1980) and Kubinger (1986) procedure for non-crossover interaction (no change in genotypic rank) and De Kroon and Van Der Laan (1981) tests for crossover interaction. The results of Wricke’s ecovalence (Wi ^2), Shukla’s stability variance statistic (σi ^2), Plaisted and Peterson mean variance (θi) and Plaisted interaction variance (θ(i)) recommended line G10 as most stable line and the most non stable line was line G8. According to non-parametric stability parameters Si ^(1) and Si ^(2) statistics of Nassar and Huehn (1987), line G2 and according to Si ^(3) and Si ^(6), line G1 was most stable. The results of principal component (PC) analysis and correlation analysis of parametric and nonparametric stability statistics and seed yield indicated that Pi (Lin and Binns, 1988) and TOP (Fox et al., 1990) methods would be useful for simultaneously selecting for high yield and stability. A biplot of the first two PCs also revealed that the parametric and nonparametric methods grouped as three distinct classes that corresponded to different dynamic and static concepts of stability. Of the 15 lines tested, the high yielding and stable lines (located in dynamic region of the plot) were G5 and G6. As a result of the current analyses, the line G6 was released for use in conditions of Iran.