COMBINING ABILITY OF MAIZE LINES IN TOP CROSSES WITH NARROW GENETIC BASE TESTERS

Abstract

The objectives were to study the combining ability of 30 lines from the maize breeding program of Instituto Agronômico, Sao Paulo State, (SP), Brazil, in top crosses schemes. The 30 lines and two testers (single cross hybrids) were crossed to obtain the 60 top crosses that were evaluated, with three commercial hybrids (checks), in two locations, Mococa and Campinas, in 2001/02, using a randomized complete block design with three replications. The following agronomic traits were evaluated: Plant height (PH), ear height (EH), lodged and broken plants (Ld + Br), and grain yield (GY) corrected to 14% moisture. The combining ability of the lines was estimated according to the Griffing's Method 2 (1956), adapted for partial diallels. The Mean Squares presented a significant (P< 0.01) effect for Lines and Lines /Tester in all the characters, except (Ld + Br) in Mococa, showing additive effects in their control. The cross effects were significant (P< 0.01) for all characteristics in Mococa and only for EH and GY in Campinas. The interaction effect (Tester x Line) was significant in Mococa (P< 0.01), indicating that non-additive effects for PH, EH and GY were important too. A significant number of three way cross presented GY similar to the checks.  The lines L 10, L 156, L 126, L 111, SLP 103 and IP 330 were outstanding for general combining ability and should be useful in maize breeding programs in tropical conditions.