Y. Lokko, A.G.O Dixon and S.K. Offei
Thirty-six F1 crosses were produced in a Design II mating scheme involving 4 improved clones as females and 9 African landraces as males. The parents and their progenies were evaluated for their reactions to cassava mosaic disease (CMD) at 3 months after planting under natural infection in three environments in Nigeria (Ibadan, 1996 and 1997 and Mokwa, 1997). The objectives were to determine the relative importance of general combining ability (GCA) and specific combining ability (SCA), estimate heterosis, and compare line and topcross performance for resistance to CMD. The results showed that the genetic variation among crosses was predominantly due to GCA of the male parents. The one-degree of freedom orthogonal contrast (parents versus crosses) for mid-parent heterosis was significant. Five crosses had significant negative heterosis. Three landraces used as males (TME 8, TME 9 and TME 11) were the best general combiners for resistance. None of the improved lines used as females contributed significantly to the GCA effects for resistance. Line performance was significantly correlated with topcross performance, indicating that parental performance can be used to predict progeny performance.