The results of QDA indicated that analysis of individual seed mixtures did not increase our ability to detect differences between group responses. Therefore, SDA was applied only to response variable subsets without regard to seed mixture composition.