Selection indices and character association studies in wheat (Triticum aestivum L.) under late sown condition

Abstract

The discriminant-function technique was used to create selection indices for 29 wheat genotypes (Triticum aestivum L.) growing under late-sown condition. The discriminant function technique was used to create 31 selection indices based on grain yield (q/ha) and its four components. Increasing the amount of characters in the selection index improves selection efficiency. The selection index based on five characters, that is, grain yield (q/ha), harvest index (%), test weight (1000 kernel wt. in g), number of tillers per plant, and biological yield (q/ha), exhibited highest relative efficiency, with genetic advance of 2619.76% and 17.34, respectively. The correlation coefficient study demonstrated a substantial positive relationship between grain yield (q/ha) and key traits, including test weight (1000 kernel wt. in g), biological yield (q/ha), harvest index (%), flag leaf area (cm2), and number of tillers per plant. Path coefficient analysis further highlighted that harvest index (%), test weight (1000 kernel wt. in g), number of tillers per plant, and biological yield (q/ha) have a significant beneficial direct effect on grain yield.