The impact of genotype by environment (GXE) interaction on yield of selected southernpea (Vigna uniguiculata (L) Walp) genotypes in southeastern United States.

摘要

A series of studies was undertaken to characterize extent, pattern, and potential impact of genotype by environment (GxE) interaction on seed yield and yield components of selected southernpeas genotypes in southeastern USA. Effects of GxE were evaluated in three field experiments, conducted during 1995--1998. Germplasm consisted of southernpea genotypes of contrasting phenology and morphological traits. Structure of GxE effects on seed yield was investigated using, mean square decomposition, and Additive Main Effect and Multiplicative Interaction (AMMI) analyses. Genotype adaption across environments was assessed by application of regression analyses to the data set. AMMI revealed a highly significant genotype by environment interaction, most of which was partitioned into a genotype by location component of variance. The magnitude of this interaction, made it possible to categorize environments into two homogenous groups. Locations with similar IPCA scores consistently grouped together, suggesting that most GxE effects could be captured with fewer testing sites per year. Consequently, some testing sites could be made redundant as their inclusion does not provide additional information to the GxE matrix. Comparison of response patterns of genotypes in future environment showed Coronet, the Industry's standard variety was most responsive at high yielding environments, but changed rank with Arkansas Blackeye and C.T. Pinkeye when grown in low yielding environments. These genotypes performed well at high yielding environments, and thus, may be better all around genotypes than Coronet.;AMMI analyses showed a significant portion of the GxE interaction could be attributed to genotype x planting date (GxD) effects. Rank correlation of June and July seed yields was also highly significant and confirmed the results of AMMI, that the same genotypes were not likely to perform well at either planting date. Entry mean heritability of seed yield in June (0.72) was almost equal to entry mean heritability in July (0.71). This statistical response indicates that gain from selection will be equal at each planting date and consequently, selection could be made with equal reliability at either planting date.;Phenotypic and path correlation coefficients among seed yield and associated traits indicated that increased number of peduncles per plant and increased number of pods per peduncle lead to increased seed yields and as such, could be used as selection criterion for high yielding southernpea.