JOINT REGRESSION VS AMMI ANALYSIS OF GENOTYPE-ENVIRONMENT INTERACTIONS FOR CEREALS IN ITALY

Joint regression and Additive Main effects and Multiplicative Interact ion (AMMI) models were compared for i) capacity of describing genotype -location (GL) and genotype-environment (GE) interaction effects (envi ronments = location-season combinations), assessed in terms of estimat ed variance of heterogeneity of genotype regressions and of the sum of the variances of significant interaction principal component (PC) axe s, and) repeatability between cropping seasons of measures of genotype stability across locations. These measures were Finlay and Wilkinson' s regression coefficient for joint regression, and the Euclidean dista nce from the origin of significant interaction PC axes (D) and the abs olute value of PC 1 score (\ PC 1 \) for AMMI. Shukla's stability vari ance (sigma(2)) was considered in addition. The study included three d ata sets for durum wheat, two for maize and one each for bread wheat a nd oat. Relationships between climatic variables and GL interaction oc currence were also assessed. AMMI proved distinctly more valuable in s ix data sets for description of GE effects and in four for description of GL effects over seasons. Its superiority was not crop-specific and tended to occur when more, distinct environmental constraints affecte d genotype responses. When both methods were appropriate, they provide d a similar ordination of sites and genotypes for GL effects. The mode ls that adequately described GL interaction over seasons generally pro vided also stability measures that were moderately repeatable between seasons. D was better repeatable than I PC 1 I and sigma(2) in a few c ases. Ordination or locations on GL interaction PC 1 tended to be cons istent both between wheat and between maize data sets having either no seasons or no genotypes in common.