GGE-BIPLOT Analysis of Durum Wheat Yield Trials

Abstract

Durum wheat (Triticum durum L.) is a vital crop in the rain-fed areas of Turkey. In this study, the grain yield of twenty durum wheat genotypes (sixteen advanced lines with four checks) tested across 8 rain-fed environments during the 2008-2009 cropping season was evaluated using GGE (Genotype (G) Main Effect + Genotype by Environment Interaction (GEI)) Biplot Approach. Environment (E) captured most (83 %) of the grain yield (GY) variation, while the portions attributed to G and GEI were only 6 and 11 %, respectively. In addition, most of the testing Es were highly correlated. According to GGE-Biplot analysis, ‘Which won where’ pattern partitioned the testing Es into three mega-environments (ME): the first ME with six Es with G13 (the highest yielder) as the winning genotype; the second ME encompassed one environment (E1, Konya) with G4 (the lowest yielder) as the winning genotype, and the last ME represented by one location (E6, Altintas) with G10 (the higher yielder) as the winning genotype. GGE-Biplot analysis showed that although the Durum Wheat Yield Trials were conducted in many environments, outcomes alike can be obtained from one or two representatives of each ME. On the other hand, no correlation of these MEs with their geographic location was observed. In conclusion, the presence of cross-over GEI underscores that efforts should be given to identify specifically adapted genotypes rather than broadly adapted ones tested on multi-environment trials (METs).