An investigation of genotype-environment interaction and stability for pea (Pisum sativum L) seed yield

Abstract

Eighteen pea breeding lines and three check cultivars were tested in two years with and without irrigation. The linear regression can explain only a small part of GEI so the use of regression technique was not possible for interpreting the data. In this case the ecovalence, stability variance and superiority measure stability parameters cannot describe properly the genotype's response. With the AMMI method it is possible to group properly the genotypes according to their response.