Components of variation, combining ability and heterosis in Ghanaian pearl millets [Pennisetum glaucum, (L), R.Br]

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Information on genetic variation is necessary in the breeding of plant populations for improvement. Two studies on genetic variation in pearl millet populations were conducted. In Study I, genetic variation and related parameters within four pearl millet populations, consisting of three local landraces (BLP, BSH and TY) and one base population (ICMV —bpl) were investigated by evaluating S1 progenies derived from each population at two locations in northern Ghana over two years. The main objectives were to determine the feasibility of selection within the populations and to recommend effective breeding strategies for their improvement. Significant genetic variations and low to moderate heritabilities were observed in the populations BLP, BSH, and ICMV bp —I, for grain yield, days to 50% blooming (maturity period), earhead length, plant height and 1000 grain weight and also for number of effective tillers per plant in BSH. However, genetic variation for most traits was not significant in the landrace population, TY. None of the landrace populations showed significant genetic variation for percent incidence of downy mildew disease, a major constraint to pearl millet production in Ghana. The results show that selection would be feasible, but with limited gains, within BLP, BSH and ICMV —bpl and that there would be little or no progress from selection within TY. The use of recurrent selection methods with progeny testing would be considered most appropriate, in view of the moderate heritabilities observed for most traits in the three populations. The genotype by location and genotype by year interaction variances were significant, which would require that superior genotypes to form improved varieties from the populations be identified by evaluating progenies at each specific location and within an adequate number of years, in order to develop adaptable improved varieties. The traits earhead length, plant height, 1000 grain weight and number of effective tillers per plant (only in BSH) had positive correlations with grain yield, while maturity period, incidence of downy mildew and incidence of chaffy tillers caused by stem borer had negative correlations with grain yield in most populations. Judging from these observations, tall plant height, long earheads, large grain size, short maturity period, low incidence of downy mildew and low incidence of chaffy tillers due to stem borer are factors with favorable effects on the yielding ability of most of the populations. In Study II, combining ability and heterosis among local and exotic pearl millet populations were investigated, by evaluating a set of 21 diallel crosses and their seven parents at two locations. The objectives were to determine gene action on traits, in order to determine appropriate breeding strategies and also to identify parents and crosses for population improvement and hybrid development programs. Only the SCA mean square was observed to be significant for maturity period, while both the GCA and SCA mean squares were significant for grain yield, earhead length and percent incidence of downy mildew, indicating that maturity period is governed solely by non-additive gene effects while all the other traits studied are governed by both additive and non-additive gene effects. However, the estimated Bakers’s ratio for grain yield was low, while those for the yield component traits were high, implying that grain yield is predominantly governed by non - additive gene effects, while the latter traits are predominantly governed by additive gene effects. In view of this, the breeding of hybrids, using the populations studied, would be a more productive approach for improving pearl millet grain yields, than population improvement. Three of the populations, namely TY, BLP and ICMV IS 90311 had the most significant GCA effects for desirable traits, indicating their suitability for use as parents in crossing to form source populations for improvement. Low to moderate levels of heterosis were expressed by the crosses for grain yield (up to 76%) and yield components. Crosses with the highest heterosis for grain yield and for most other traits were mostly the local-by-exotic crosses, while the local-by-local crosses had relatively low heterosis for most traits.
A thesis presented to the Department of Crop and Soil Sciences, of the College of Agriculture, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana, in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Plant Breeding and Genetics, 2005