for yield potential under non-stress conditions are likely to be
limited to environments where stress is relatively mild and
mean yields are greater than 50% of the non-stress yields. Our
experiment support this observation, and indicate that the stress
level in experiment should be severe enough to cause between
65 and 85% reduction in the mean yield of the stress trial as
compared to the non-stress trial to make a clear distinction
between drought-tolerant lines and the lines with high-yield
potential. Most breeding programs screening for drought
tolerance fail to impose sufficiently severe stress in their trials
and as a result are not able to accurately select drought-tolerant
lines. The use of reliable tensiometers in the stress trials, the use
of a number of drought-susceptible and tolerant checks at
repeated intervals and monitoring leaf rolling and leaf drying in
these checks as well as in the experiment as a whole, and proper
monitoring of water table depth in lowland trials are the tools
that can assist breeders in managing the proper level of stress in
their screening.