Observed response and genetic variability in two maize populations after four cycles of reciprocal full-sib selection
Is Version Of
Effectiveness of four cycles of reciprocal full-sib selection for yield in two maize populations, BS10 and BS11, was evaluated and compared with that of ten cycles of mass selection for prolificacy. The original (CO) and advanced populations (Cn) and their respective crosses (Cn x Cn) were evaluated in seven environments. In separate experiments, unselected CO- and C4-derived S(,1) lines of each population were evaluated in four environments to determine if four cycles of reciprocal full-sib selection changed the genetic variance in BS10 and BS11;Yield gain per cycle of reciprocal full-sib selection, estimated by the linear regression entry means on cycles of selection, was significant only in the variety cross. Total response after four cycles of selection, however, was significant in both parental populations (7.0 and 7.9% in BS10 and BS11, respectively). Population response was confirmed by the higher mean performance of the unselected C4-derived S(,1) lines relative to that of CO-derived S(,1) lines of each population. Increase in midparent heterosis was small and improvement in crossbred performance (7.0%) could be attributed mostly to improvement in the parental populations;Mass selection for prolificacy was not effective for improving grain yield of the populations per se and their crosses. Possible explanations for the lack of response included inbreeding depression, genotype-environment interaction, and a threshold frequency of genes conditioning prolificacy beyond which other relationships (e.g., yield and ear size) tend to become more critical;Reciprocal full-sib selection for yield and mass selection for prolificacy increased ears per plant in the two populations and their crosses. Both methods also caused changes in days to silk and percentage of stalk lodging; the direction and magnitude of responses, however, differed between the two methods;Four cycles of reciprocal full-sib selection did not significantly change the genetic variance for grain yield in BS10 and BS11. Further gains, therefore, should be realized in subsequent cycles of selection. An increase in genetic variance for ears per plant was observed for BS10. Genetic variance for days to silk was significantly reduced in both varieties after four cycles of reciprocal full-sib selection for yield. Comparison of the grain yield distributions of CO- and C4-derived S(,1) lines showed that reciprocal full-sib selection has increased the likelihood of extracting productive inbred lines from BS10 and BS11. Reciprocal full-sib selection seemed to have emphasized selection for additive genetic effects.