Use of biplot analysis and factorial regression for the investigation of superior genotypes in multi-environment trials

• Published in: European journal of agronomy 2005-03, Vol.22 (3), p.309-324
• Main Authors: Voltas, J., López-Córcoles, H., Borrás, G.
• Format: Article
• Language: English
• Subjects: Adaptation to environment and cultivation conditions; agronomic traits; Agronomy. Soil science and plant productions; ambient temperature; Biological and medical sciences; biplot analysis; cultivars; drought; environmental factors; filling period; flowering; Fundamental and applied biological sciences. Psychology; G × E interaction; Genetics and breeding of economic plants; genotype; genotype-environment interaction; germplasm evaluation; grain crops; Grain yield; Mega-environment; Morpho-physiological traits; phenotype; regression analysis; SAS codes; spring wheat; Triticum aestivum; Varietal selection. Specialized plant breeding, plant breeding aims; variety trials; Wheat; winter wheat
• ISSN: 1161-0301; 1873-7331
• DOI: 10.1016/j.eja.2004.04.005

For cultivar evaluation in recommendation trials, grain yield is the combined result of effects of genotype (G), environment (E) and genotype × environment interaction (GE). In this framework, GGE biplot and factorial regression analyses represent different approaches for GE investigation. The biplot facilitates a visual evaluation of ‘which wins where’ patterns, useful for cultivar recommendation and mega-environment identification. Factorial regression, alternatively, involves a description of cultivar reaction to the environment in terms of biophysical variables that directly affect crop yield. A combination of both methods can be used to assist in the detection and characterisation of superior genotypes. Initially, winning cultivars are detected using features of GGE biplot analysis. By applying a factorial regression model, these cultivars are subsequently characterised in relation to key environmental factors and physiological systems involved in yield determination. This strategy was illustrated in the analysis of Spanish winter and spring wheat recommendation trials for 2002. In both cases, GGE biplots identified three winning cultivars along with the corresponding subsets of trials where each winning cultivar showed yield superiority. Factorial regression then enhanced the understanding of circumstances under which winning cultivars performed better and identified phenotypic traits favouring a higher yield. For winter wheat, differences among environments in (i) pre-flowering thermal time, (ii) low temperatures prior to flowering and (iii) drought incidence during grain filling caused genotype-dependent responses in grain yield. In spring wheat, genotypes reacted differentially to changes in (i) pre-flowering thermal time, (ii) pre-flowering drought incidence and (iii) high temperatures during grain filling. The combination of relevant environmental variables allowed specific winning niches for each superior cultivar to be defined from an ecophysiological perspective. Such an outcome could be regularly employed in the future to delineate predictive, more rigorous recommendation strategies as well as to help define meaningful mega-environments for recommendation purposes in Mediterranean areas.