Identifying Promising Pearl Millet Hybrids Using AMMI and Clustering Models uri icon

abstract

  • A set of 27 pearl millet (Pennisetum glaucum (L.) R. Br.) ) hybrids that newly developedusing A1 cytoplasmic male-sterile lines, were evaluated over three (two wet and one dry)crop seasons (hereafter refer to as environments) in Randomized Complete Block Design(RCBD) with two replications to predict genotype by environment (G × E) interaction forgrain yield and its component traits, and to identify the high yielding stable hybridsthrough AMMI and cluster analysis method for possible adaption. Analysis of varianceshowed significant genetic variation for all studied traits exists. The Additive Main Effectsand Multiplicative Interaction (AMMI) analysis indicated that genotype, environment andG × E interaction highly significant for grain yield and other traits. However, G × Einteraction component explained very low magnitude (3.87%) towards total geneticvariation, while genotype alone contributed much higher magnitude (8.04%) in AMMImodel and found TNBH 05 45 was an ideal hybrid for all three environments for grainyield (34% over best control). Diversity analysis showed seven diverse clusters followingEuclidean distance coefficient of 0.91 and found TNBH 05 03 and TNBH 05 45 hybridsare promising. Based on these two models, TNBH 05 03, TNBH 39 and TNBH 05 45 wereidentified for stable performance per se in all the environments, and could be used forsubsequent advanced testing and hybrid breeding programmes for possible release withinregions

publication date

  • 2017