S. Veillet et al., COMBINED GENETIC-ANALYSIS OF PARTIAL BLAST RESISTANCE IN AN UPLAND RICE POPULATION AND RECURRENT SELECTION FOR LINE AND HYBRID VALUES, Theoretical and Applied Genetics, 92(6), 1996, pp. 644-653
The CNA-IRAT 5 upland rice population has been improved for 4 years by
recurrent selection for blast resistance in Brazil. In order to predi
ct the efficiency of recurrent selection in different test systems and
to compare the relative advantage of hybrids versus pure line breedin
g, a combined genetic analysis of partial blast resistance in the CNA-
IRAT 5 population was undertaken. A three-level hierarchical design in
inbreeding and a factorial design were derived from the base populati
on. Partial blast resistance of lines and hybrids was evaluated in the
greenhouse and in the field by inoculation with one virulent blast is
olate. The means and genetic variances of the hybrids and lines were e
stimated. Genetic advance by recurrent selection was predicted from es
timates of variance components. The inheritance of partial blast resis
tance was mainly additive but non-additive effects were detected at bo
th levels of means and variances. Mean heterosis ranged from 4%-8% for
lesion size and lesion density to 10-12% for leaf and panicle resista
nce. High dominance or homozygous dominance variances relative to addi
tive variance and negative covariance between additive and homozygous
dominance effects were estimated. A low frequency of favourable allele
s for partial resistance would explain the observed organisation of ge
netic variability in the base population. Recurrent selection will eff
iciently improve partial blast resistance of the CNA-IRAT 5 population
. Genetic advance for line or hybrid values was expected to be higher
testing doubled haploid lines than S1 lines, or than general combining
ability. Two components of partial resistance assessed in the greenho
use, lesion size and lesion density, could be used as indirect selecti
on criteria to improve field resistance. On the whole, hybrid breeding
for partial blast resistance appeared to be slightly more advantageou
s than pure line breeding.