|
The purpose of this study was to obtain the basic genetic
information between the conventional high input of pesticides
and fertilizer cropping system and the low input cropping
system which would be used in rice breeding program.
Experimental results are summarized as follows: 1. Significant
differences of agronomic traits of varieties were found under
low input management, This difference was due to genetic
inheritance. 2. Selection should be proceeded under low input
management. 3. Most of agronomic traits performed superior in
high input management under nature selection. Earlier heading
day in second crop was observed. 4. The selection index
computed in F5 generation showed that yield component and
important agronomic traits were different under different
genetic background and management. Panicle number had the
highest contribution than other traits in two managements. 5.
The heritability of traits from F2 to F5 populations under
different input managements were not consistent in cross
combinations and generations. The selection efficiency of
traits under two input managements was different. 6.
Correlation coefficiences among traits in hybrid populations
and input managements were differed significantly in strength
and direction in each generation. 7. The lines that selected
from high input did not adapte to low input management. It
should be selected directly when planted under low input
managemant. 8. Under high input managemant condition, the
expression of the lines selected from low input were superior
to the lines selected from high input management. 9.
Significant difference was found in camparing expected genetic
gains and actual values in each management. In general, the
actual value was higher than the expected values. 10. The
principal component analysis showed that the expression of
lines selected from different input management performed
differently.
|