The objective of this work was to compare testers for discrimination and evaluation of the combining ability of 64 S2 families from the maize (Zea mays L.) hybrid CMS-39. The families were evaluated based on their performance per se and in crossing to three broad base testers. The testers were BR-201, a double hybrid, and two populations, CMS-50 and CMS-39. The topcrosses and S2 families were evaluated in four 8 x 8 lattices, with two replications, set up in adjacent areas on the experimental field of UFLA, Lavras, MG, Brazil, in 1996/97. General and specific combining ability, following partial diallel, and the heterosis of each topcross based on the performance of S2 families per se were estimated. Estimates of correlations were obtained from performance means of the S2 families and the hybrids with the different testers. It was only possible to discriminate the testers according to combining ability and heterosis estimates. Population CMS-50 was identified as the best tester.
Zea mays; breeding methods; hybrids; heterosis
