Fifteen genotypes of pearl millet (Pennisetum glaucum L.) were studied at two locations (Elrawakeeb and Shambat) of Sudan, during the years 2003 and 2004. The study was conducted to determine the interrelationship between yield and yield components and other quantitative traits, including days to 50%flowering, date to maturity, panicle length, dead part length/main head and plant height. The path analysis and phenotypic and genotypic correlation coefficients were calculated. The results showed

more »

... hat strong positive significant genotypic and phenotypic correlations were observed between grain yield/plant and harvest index (rp=0.754, rg=1.08), grain yield/plant and biomass dry weight (rp=0.639, rg=1.064) and grain yield/plant and number of seeds per panicle (rp=0.608, rg 0.820) and with other quantitative characters. However, grain yield/plant has negative association with days to 50% flowering, date to maturity and dead part length/main head. The path analysis indicated that number of fertile tillers/plant had the highest direct effect (0.512) on grain yield/plant. However, thousand-seed weight was greatly reduced by the negative indirect effects, through number of fertile tillers/plant and number of seeds/head. Based on the present results, it could be concluded that the number of fertile tillers/plant, number of seeds/head, thousand-seeds weight and panicle length/main head could be identified as the most important characters that associated with yield, and therefore can be used as selection criteria for yield improvement of pearl millet.