Using additive main effect and multiplicative interaction model for exploration of yield stability in some lentil (Lens culinaris Medik.) genotypes

The additive main effect and multiplicative interaction (AMMI) analysis has been indicated to be effective in interpreting complex genotype by environment (GE) interactions of lentil (Lens culinaris Medik.) multi- environmental trials. Eighteen improved lentil genotypes were grown in 12 semiarid environments in Iran from 2007 to 2009. Complex GE interactions are difficult to understand with ordinary analysis of variance (ANOVA) or conventional stability methods. Combined analysis of variance indicated the genotype by loca- tion interaction (GL) and three way interactions (GYL) were highly significant. FGH1 and FGH2 tests indicated the five  significant  components; FRatio  showed  three significant  components  and F-Gollob detected  seven significant components. The RMSPD (root mean square predicted difference) values of validation procedure indicated seven significant components. Using five components in AMMI  stability parameters (EVFI, SIP- CFI, AMGEFI and DFI) indicated that genotypes G5 and G6 were the most stable genotypes while consider- ing three components in of AMMI stability parameters (EVFII, SIPCFII, AMGEFII and DFII) showed that genotypes G8 and G18 were the most stable genotypes. Also genotypes G2, G5 and G18 were the most stable genotypes according to AMMI stability parameters which calculated from seven components  (EVFIII, SIP- CFIII, AMGEFIII and DFIII). Among these stable genotypes, only genotypes G2 (1365.63 kg × ha-1), G11 (1374.13 kg × ha-1) and G12 (1334.73 kg × ha-1) had high mean yield and so could be regarded as the most favorable genotype. These genotypes are therefore recommended for release as commercial cultivars