RECIPROCAL FULL-SIB RECURRENT SELECTION IN MAIZE .1. DIRECT AND INDIRECT RESPONSES

Interpopulation recurrent selection methods can be used for developing new heterotic patterns and enhancing those identified previously. The objective of this research was to estimate the direct and indirect responses to eight cycles of reciprocal full-sib recurrent selection in maize (Zea mays L.) populations BS10 and BS11. The populations and their crosses corresponding to Cycles 0, 2, 4, 6, and 8 were evaluated at two levels of inbreeding (F = 0 and F = 0.5), and populations also were testcrossed to the respective Cycle 0 populations and unrelated inbred lines. Entries were evaluated in eight environments. Eight cycles of selection were effective for increasing grain yield of the population cross (6.5% cycle-1), BS10 (3.0% cycle-1), and BS11 (1.6% cycle-1). Standability, prolificacy, and grain moisture also were changed in the desired directions, and, in most instances, the responses were linear across cycles of selection. Direct responses were greater than indirect responses for grain yield and standability. The selected populations also performed better than did the unselected populations in crosses with unrelated testers. Changes in inbreeding depression and heterosis were associated with selection. Rates of direct and indirect responses for grain yield in the inbred populations tended to be greater than the rates of responses in the noninbred populations. Reciprocal full-sib selection increased the frequency of homozygotes for favorable alleles in the BS10 and BS11 populations, which would be improved sources of vigorous inbred lines with good general combining ability. Reciprocal full-sib selection seems to have increased the frequency of heterozygotes in the population crosses.