Phenotypic variation and diversity assessment in maize landraces

Maize landraces are invaluable reservoirs of genetic diversity for crop improvement, yet systematic evaluation of their quantitative traits remains insufficient. In this study, 119 maize landraces were evaluated for nine quantitative traits related to yield and tassel morphology. Genetic diversity indices, principal component analysis(PCA), cluster analysis, and stepwise regression were applied to characterize phenotypic variation and identify core evaluation indicators. The landraces exhibited substantial genetic variation, with coefficients of variation ranging from 20.73% to 60.96% and Shannon-Winner index from 0.88 to 0.98. PCA extracted four principal components explaining 68.81% of the total variation, with the first component (24.17%) primarily driven by the number of secondary tassel branches. Based on comprehensive D values (0.033-0.960), ten superior landraces were identified, including Z102, Z7, and Z89. Cluster analysis classified the 119 landraces into three groups differing in tassel morphology, grain characteristics, and agronomic performance. Stepwise regression further identified four key indicators-number of secondary tassel branches, angle between primary tassel branches and main axis, tassel diameter, and number of spike rows-as the most informative traits for breeding selection. These results provide a practical framework for the efficient exploitation of maize landrace resources and the identification of elite breeding materials.