Strategy of multiple selection indices for discrimination of potential genotypes and associated traits for yield improvement in pearl millet [Pennisetum glaucum (L.) R.Br.]

Abstract

A study was conducted with 31 genotypes of pearl millet to identify the potential through correlation, path analysis and discriminant function analysis for yield and yield attributing traits at the Department of Millets, Tamil Nadu Agricultural University, Coimbatore during Kharif 2019. The top ranking genotypes for single plant yield was Cumbu 2, Kizikuppam local, PT6706, PT6582 and PT6581. High PCV and GCV were recorded for harvest index, single plant yield, biological yield, thousand grain weight, the number of productive tillers and total grains per panicle. The path coefficient analysis revealed that harvest index, biological yield, panicle length and flag leaf length showed a positive high direct effect on grain yield per plant, which can be utilized for the formulation of selection indices. The discriminant function analysis revealed that the maximum genetic gain for grain yield per plant can be obtained by selecting the traits like leaf sheath length, leaf length, flag leaf length, flag leaf width, the number of nodes, panicle length, the number of productive tillers, thousand grain weight, biological yield, harvest index and single plant yield in combination when path direct effect was used as economic weights. Selection criterion score values showed that the genotypes such as Cumbu 2, Kizikuppam local, Nattu Cumbu, PT6705, PT6580, PT6706, Cumbu 1 and PT6067 have outperformed others and can be utilized for further breeding programmes.

Keywords

• genetic variability
• path correlation
• discriminant function
• selection index
• pearl millet