Genotype-environment interaction in Avena sativa L. employing ammi and factorial correspondence models

Resumo

The objective of this study was (i) to determine the presence of genotype-environment interaction (GXE) in Avena sativa l. for above dry matter yield of Schizaphis graminun tolerant and non-tolerant genotypes according to two different models: additive main effects and multiplicative interaction (AMMI) and factorial correspondence analysis; and (ii) to study and compare the results obtained with these models. Twelve genotypes and 12 environments were conducted at La Dulce and La Plata (Argentina) during three years (1993, 1994 and 1995). The environment (E), genotype (G) and GxE factors explained a 41.15%, 7.88% and 36.36% of the total sum of square, respectively. The first three axes of principal component analysis of AMMI were highly significant (p<0.001), explaining a 57.99%, 29.03% and 6.27% of interaction sum of square. The first three factorial correspondence coordinates accounted for 58.98%, 29.58% and 5.60% of the interaction sum of square. The relationships between tolerant genotypes-first clipping environments and non-tolerant genotypes-second clipping environments were reflected in the biplots of both models. The simultaneous use of AMMI and factorial correspondence analysis appeared as a useful methodology for the study of GxE interaction in the oat breeding for tolerance to Schizaphis graminum.