Juan Carlos Caffarel and Mohamed Mergoum. North Dakota State Univ, Loftsgard Hall, P.O. Box 5051, Fargo, ND 58105-5051
The genotype by environment interaction is important for breeders in the selection process when there is a ranking change from one environment to another. When this occurs, the region can be subdivided into sub-regions, with more homogenous conditions that will reduce the genotype by location interaction (GL). The sites regression model (SREG) has been proved an effective means to determine regions with significant crossover interaction (COI) for yield (Navabi et al., 2006). In this study, an SREG model and its genotype-genotype by environment (GGE) biplot representation were used to study the association among locations and their representativeness for loaf volume. A genotype-by-trait (GT) biplot was used to study the association between loaf volume and 11 quality traits, looking for suitable traits for indirect selection. The results from this study were: i) the association among locations for loaf volume was non-repeatable across years, therefore, they must be treated as a single megaenvironment; ii) Langdon was the location less representative of the overall locations; and iii) there was no trait consistently associated with loaf volume across years suitable for indirect selection.