Additive Main Effect and Multiplicative Interaction Analysis in Bread Wheat-Derived Lines Across Environment

Abstract

Identification of high yielding stable genotypes is an integral objective of plant breeding programs. Testing of genotypes across environments is required to determine yield stability of genotypes. The specific objective of the current study was to analyze genotype by environment interaction (GEI) of grain yield for 50 genotypes using the additive main effects and multiplicative interaction (AMMI) model. Experiments were planted in an alpha lattice design with two replicates in Peshawar (E-1 and E-3), Hangu (E-2 and E-4) and Kohat (E-5) Khyber Pakhtunkhwa province, Pakistan during 2011/12 and 2012/13. Analysis of variance revealed significant differences among genotypes for all traits, while interactions due to genotype by environment were significant for all traits except days to emergence and 1000-grain weight. Significant GEI justified environment-specific as well as AMMI analysis to identify genotypes with specific and wider adaptation. The AMMI analysis revealed that the first interaction principal component analysis (IPCA 1) captured 64.0% of GEI sum of squares while the second interaction principal component analysis (IPCA 2) explained 25.8% of the interaction sum of square. The AMMI biplot identified G30 as a high yielding genotype followed by G19 and G49, whereas low yielding genotypes were G13, G8 and G7. Being close to IPCA1 axis, the most stable genotype with wider adaptability was G30 followed by G31 and G25. Based on AMMI stability value (ASV), genotypes G18 (2.15), G5 (2.78), G27 (3.72), G44 (4.31), G25 (4.43), G42 (4.57), G43 (5.78), G11 (5.82), G1 (7.66) and G29 (7.81) were found in the given order of relative stability. GGE biplot analysis explained 79.9% (PCA1=56.6 and PCA2= 23.3%) of the total variation. Genotype G19 positioning on vertex in sector E-3, E-4 and E-5, while G30 in sector E-1 and E-2 revealed their specific suitability to respective environments. GGE biplot identified environment E-4 as the most representative environment, whereas G49, G30, G22 and G45 as the high yielding genotypes. Shifted multiplicative model (SHMM) grouped genotypes into four clusters based on similarity/dissimilarity index for grain yield. High yielding and stable genotypes G19, G49 and G30 were placed in group B. Grain yield had positive association with tillers m-2 (r =0.73**), grain weight spike-1 (r =0.57**), biological growth rate (r =0.44**), grain growth rate (r = 0.80**), biological yield (r = 0.41**) and harvest index (r = 0.55*). The SHMM clustering and correlations of yield with other traits inferred that tillers m-2, grain weight spike-1, biological growth rate, grain growth rate, biological yield and harvest index contributed towards higher grain yield. Therefore, these traits could be used as selection criteria for the improvement of grain yield in bread wheat. Stability analysis identified G49 (Wafaq × Ghaznavi-98-3) as a high yielding stable genotype among breeding lines which can be commercialized after fulfilling procedural requirements