The objective of this experiment was to assess the genotypes by environment interaction for eight morphoagronomic traits in thirty F7:8 snap bean superior lines and evaluate the pod yield stability. The experiments were carried out in the cities of Bom Jesus do Itabapoana, Campos dos Goytacazes and Itaocara, located in the Northern and Northwestern regions of Rio de Janeiro State, Brazil. Experiments were designed in randomized blocks with two replications, using the traditional method (Yates & Cochran, 1938) and the methods proposed by Plaisted & Peterson (1959), Wricke (1965), Kang & Phan (1991) and Lin & Binns (1988), as well. Only the average fiber content in pods revealed no significant genotype x environment interaction, whereas most of the traits showed simple interaction, including pod yield. The results obtained using the methods of Plaisted & Peterson (1959) and Wricke (1965) agreed to a far extent and had a moderate estimate (0.6131) in relation to the traditional method. Both methods highlighted genotypes with pod yield close to the general average, i.e., around 7,546.46 kg ha-1. Kang & Phan's (1991) algorithm was more efficient in optimizing the stability estimates obtained using the methods of Plaisted & Peterson (1959) and Wricke (1965) than the traditional method. Lin & Binns' (1988) procedure revealed the most stable, responsive and productive progenies. Progenies 04, 03 and 10 stood out in all environments.
Phaseolus vulgaris; phenotypic stability; genotype by environment interaction
