Linear discriminant analysis to differentiate sorghum germplasm: A crucial tool for breeding programmes

Abstract

A total of 86 Sorghum genotypes along with three checks (CO 30, CO 32, and K 12) were evaluated during Rabi 2021 season to identify variations and character associations among grain yield and yield component traits. The phenotypic data collected were used to create a statistical database and were analyzed using linear discriminant analysis (LDA) to identify and discriminate landraces for utilization in sorghum breeding. The LDA successfully differentiated the genotypes into three groups with an accuracy of 73.52%. The study revealed a significant level of variation among the genotypes, based on observations for nine quantitative traits. Further analysis using the LDA biplot showed that the genotypes within clusters 1 and 4 hold potential for future breeding programmes. Therefore, the observed phenotypic data can be useful for identifying and selecting appropriate accessions for sorghum improvement.

Keywords

• sorghum
• landraces
• clusters
• linear discriminate analysis(lda)