Evaluation of two cycles of selection among and within half-sib families in two populations of maize in the Northeast of Brazil

Abstract

During the planting periods of 1974/75/76, two cycles of selection were carried out among and within half-sib families in populations of maize (Zea mays L.) Dentado Composite and Flint Composite at different regions of the Brazilian Northeast. ln each cycle 500 half-sib families of each population were evaluated with simple lattice duplicated 10 x 10 and one replication per local. ln 1978, seeds of original I and II Cycles of each population were multiplicated using controlled sib crosses. In 1979, these six treatments with the other ten populations were evaluated in experiments using a randomized block design, with five replications and at two different locations, in order to observe the selection efficiency. The statistical analysis showed significant variation to selection cycles in Dentado Composite, but not for the Flint Composite. No significative interaction was shown for Cycles Dentado x Locations and Cycles/Flint x Locals. The interaction of "Restantes x Locations" was significant, confirming the sensibility of Centro-Sul materials to locations effects, which did not occur with populations that are under selection at the Northeast region. The observed genetic progress by annual cycle (linear regression coefficient) was 475 kg/ha (15.47%) for Dentado Composite, and 157.5 kg/ha (5.20%) for Flint Composite, being the latter statistically not significant within conventional probability level. The estimated gains in this study are indicative of broad genetic variability in the two populations, thus permitting substantial gains in subsequent cycles of selection.