The inheritance of plant height was studied in six crosses of pearl millet using six generations at two stages of crop growth, i.e. seedling to panicle initiation and anthesis, in four environments representing various levels of moisture stress during summer (March-June) season. The F1 of all the crosses exhibited heterosis over better parent (BP), and the analysis of F2 and the back crosses indicated the prevalence of nonallelic interactions. Additive dominance model was inadequate in 27 of the 36 cases. Testing for four and five-parameter models, in case of inadequacy of three parameters, before direct switching over to six-parameter model appeared to be essential in order to have unbiased and representative estimates. Both additive and dominance effects were important with preponderance of the latter. The presence of epistasis, observed from the analysis of generation means, was concluded and confirmed by the existence of digenic interactions. Duplicate epistasis played a major role in character expression.
Pearl millet, inheritance, genetic parameters
