Abstract:
Nineteen F1 families were produced by pair-crossing 7 stoloniferous genotypes of Digitaria milanjiana. Progenies and parents were grown as spaced plants (0.5 by 0.5 m microswards) and plucked samples were taken from vegetative regrowth in 2 successive seasons for estimation of digestibility using the in vitro cellulase method. Stolon number, herbage yield (one, 9-week regrowth), flowering date and leaf width were also assessed. Parental clones differed significantly in plucked sample digestibility (range 4.1 units) and there was no significant clonexyear interaction. Heritability, based on parent-progeny regression, was 0.86. The occurrence in progenies of a low percentage (4.9%) of genotypes with plucked sample digestibility significantly higher than that of the higher parent suggested that genetic improvement of plucked sample digestibility may be possible. However, such segregates did not occur in the progeny of high digestibility parents. Digestibility of plucked samples was not correlated with any of the agronomic attributes measured. This, together with the high heritability for stolon number (h2 = 0.86) indicates that it is feasible to breed a genotype of D, milanjiana with high yield, high leaf digestibility and vigorous stolon development for clonal propagation. Genotypes with a mean plucked-sample digestibility of >70.0% occurred in 11 families. For these families highest and lowest digestibility genotypes were further sampled; only the top 2 fully expanded leaves were analysed. Conclusions derived from plucked samples were confirmed, showing that full-sibs differed genetically in leaf digestibility. Three pairs of full-sibs, which differed in leaf digestibility by 5.0-7.2 units, were selected for in vivo and physiological comparison.