Performances of Composite-Sibbed Lines of Corn Derived from Different Selection Methods and Their Hybrid Combinations

Abstract

Six S1 populations from six commercial single crosses of corn were used as the starting breeding material, with each selected independently under low plant density conditions (1.5 plant/m²) by two different methods of selection: 1) selfed family line selection (SFL); and 2) recurrent S1-full sib line selection (RFL). Breeding was carried out for three generations and the top six lines of each method were selected. Afterward, lines from the second method were separated into 18 sublines and each was subjected to RFL and MSL (mass sibbed line selection) for another three generations. Consequently, the best six lines from each selection method were subjected to further selection by the method from which each line was derived, but under high (5.3 plant/m²) and low (1.5 plant/m²) plant density conditions. There was no clear advantage in crosses of lines selected from high and low plant density conditions. However, most outstanding crosses were from diallel crosses within the RFL and MSL lines. One of the crosses was from lines of common parents, thereby substantiating the important role of additive gene action for the yielding ability of a hybrid. However, lines of common parents from RFL set 2 and RFL set 4 performed consistently well. Therefore, logically, RFL selection should be a good option for early
generation screening because it provides the chance for improvement of the line per se performances through recurrent cycles. However, the performance of final selected lines should be evaluated in crosses.