The interaction of wheat genotypes with a specific factor of the environment

Abstract

SummarySeveral alternative statistical approaches to the analysis of genotype-environment interactions are summarized, and their inter-relationships are presented mathematically and graphically. An experiment was designed to test the ability of these analyses to satisfactorily describe genotype-environment interactions and to locate genotypes causing or contributing significantly to such interactions in a situation where environments differed in only one factor, i.e. soil nitrogen status.Twenty-five lines of wheat were grown at each of 5 levels of applied nitrogen (0–270 kg N/ha) in a sandy, nitrogen-deficient soil, and their grain yields and yield components (ears per unit area, grains per ear, weight per grain) were determined. Designating the nitrogen levels as ‘environments’, linear yield responses, and deviations from linear responses, of individual wheat entries to environment mean yield were calculated according to the various statistical procedures previously outlined. The results suggested that the linear response to environment mean yield was similar for all wheat entries. However, further inspection of the data revealed that the responses of eight entries to environment mean yield departed significantly from linearity. Although several analyses satisfactorily identified these aberrant entries, classifying them as ‘unstable’, it was concluded that the degree of departure from linearity of response was not an adequate measure of instability. Furthermore, as previously suggested by Knight (1970), it was confirmed that the relative ranking of an entry for yield stability according to such a concept varied according to the average response of the group of entries with which it was being compared.The general effects of applied nitrogen on yield components were to increase the number of ears and the number of grains per ear, and to decrease grain weight. However, only the first of these effects occurred consistently for all entries. The number and weight of individual grains per ear varied with the nitrogen treatment, often unpredictably, in a complex pattern which differed significantly between entries.