Authors: *Gani, M. and Shinggu, C. P.
The treatments consisted of three seed rates (3, 6, 9 kg ha-1), two sowing methods (broadcasting, dibbling) and three poultry manure rates. The experiment was laid out in a split plot design with three replications. A factorial combination of sowing method and poultry manure rate constituted the main plot, while seed rate occupied the subplots.In all the years, there was a strong andpositivesignificant correlation between grain yield and plant height, leaf area index, number of tillers per plant, crop growth rate, panicle length, number of fingers per plant, productive tiller per plant, harvest index, 1000 grain weight, straw yieldhectare, panicle yield per hectare and negatively correlated with cumulative weed density and dry weight. The partitioning of correlation into direct and indirect contributions showed that, productive tillers per plantmadehighest direct contribution to grain yield of finger millet, while productive tillers per plant via harvest index, made the greatest indirect contribution to grain yield. Similarly, the highestindividual percent contributionsto grain yield was made by productive tillers per plant and the combined contribution was from productive tillers per plant via harvest index. Correlation and path analysis revealed that, productive tillers per plant and harvest index are theimportant yield contributing traits therefore due critical emphasis needs be given to these traits while selecting for grain yield improvement in finger millet.
Keywords: correlation, path co-efficient, finger millet
*Correspondence: E-mail: email@example.com