STATISTICAL-ANALYSES OF GLIADIN REVERSED-PHASE HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHY PATTERNS OF HARD RED SPRING AND HARD RED WINTER-WHEATCULTIVARS GROWN IN A COMMON ENVIRONMENT - CLASSIFICATION INDEXES

Reversed-phase high-performance liquid chromatography was used to anal yze gliadins extracted from grain harvested from 12 hard red winter (H RW) and 12 hard red spring (HRS) wheat cultivars grown in a common env ironment. Visual examination of the gliadin patterns did not distingui sh the two wheat classes by the presence or absence of any one particu lar peak. The peak heights at each time interval from each cultivar we re analyzed through cluster, principal component, and canonical analys es. Cluster analyses, based on closest (Euclidean) distances, produced five clusters plus six HRW cultivars that did not fall into any clust er. In the five clusters, two contained only HRS cultivars, two contai ned only HRW cultivars, and one contained both HRW and HRS cultivars. Principal component analysis showed that the first principal component (PCl) explained 21% of the total variation among cultivars, primarily separating HRW and HRS classes with only minor overlap. The first thr ee principal components together explained nearly half (44%) of the to tal variation. In these three major dimensions, there was greater scat ter within the HRW class than within the HRS class. Canonical analyses demonstrated that the correlation between PCI and the first canonical variable was 0.79, indicating that the cultivars and classes were in similar groups. Even though both PCI and the first canonical variable separated HRW and HRS classes, HRW cultivars occurred among the HRS cu ltivars in both analyses. Canonical discriminate analysis, based on gl iadin reversed-phase high-performance liquid chromatography, allocated all cultivars to their correct classes, except for the HRW cultivars TAM 105 and TAM 107.