yield under stress (rG); and the ratio of response to indirect
selection for grain yield based on secondary traits to response to
direct selection for grain yield under stress (CRY/DRY) for the
two populations (IR55419-04/IR 64 and IR55419-04/Way
Rarem) under upland stress in 2004 are presented in Table 4.
None of the secondary traits included in our study showed a
higher estimate for H than grain yield under stress. PE% and HI
showed a high positive rG with grain yield under stress. DTF
showed a high negative correlation with grain yield. LR and LD
recorded at weekly intervals showed little or no rG with grain
yield under stress. The CRY/DRY value was always less than
1.0, indicating that selection for these secondary traits would be
less beneficial than direct selection for grain yield under stress.
However, among the secondary traits, DTF, panicle exsertion
and HI showed a high value for CRY/DRY, indicating that PE%
and HI would be more reliable secondary traits for indirect
selection for grain yield under stress than LR and LD.
Atlin and Frey (1989) observed that greater gains are
expected from direct selection in low-input or stress environments
than from indirect selection under optimal conditions
unless H is greater in the non-stress than in the stress
environment and the rG exceeds 1.0. In our study, H for grain
yield under non-stress conditions was similar to that under
stress conditions and rG between yield under non-stress and
stress conditions was low to moderate and never approached 1.0
(Table 3). Similarly, none of the secondary traits had higher H
estimates than grain yield and rG between secondary traits and
grain yield under stress was not close to 1.0, except for HI, for