Multiple Regression Analysis of Grain Yield in Pearl Millet

Multiple regression equations expressing grain yield (Y) in terms of spike length (X₁), spike girth (X₂), 500 grain weight (X₃), number of ear bearing tillers (X₄), grain density (X₅); and also yield (Y) in terms of protein content (X₆), calcium (X₇), phosphorus (X₈), iron (X₉) and ash content (X₁₀) of grain, were fitted to 41 mean values of S₀, S₁ top crosses of one experiment and 81 S₀ parents and top crosses of another. The resulting equations were, Y=816.7+ 0.008505 (X₄) having R=0.4908 for S₀; Y=638.5+0.014952 (X₄) having R=0.7926 in S₁; and Y=899.9+0.020805 (X₁) + 0.1858 (X₃)+0.005995 (X₄) having R=0.6847 for the agronomic characters of the other experiment. In the case of quality characters observed only in the second experiment, the equation was Y=904.6 -0.06347 (X₆) -0.9048 (X₁₀) having R=0.4993. Thus, number of ear bearing tillers was found to be the most potent character for yield and the association in S₁ was more pronounced.