Mitotic recombination has been postulated by some researchers as a mec
hanism to explain the lack of segregation for certain traits in plants
derived from tissue culture. Three genetic marker lines were used to
examine whether mitotic recombination could be detected from plants re
generated from rice (Oryza sativa L.) tissue culture. One line, CI 110
14, had two dominant genes An (awned) and Gl (pubescence) in linkage g
roup 12. The other two, CI 11020 and CI 245714, had two corresponding
recessive genes nn (awnless) and gl (glabrous). A total of 1788 R(1) p
lants regenerated from immature panicles of F-1 plants from the crosse
s Ct 11020 x CI 11014 and CI 245714 x CI 11014, and from the parental
lines, were evaluated from 1988 to 1990 in the greenhouse, and 360 R(2
) lines derived from these F-1 plants and parental lines were evaluate
d in the held. No mitotic recombination occurred for the genes investi
gated in this experiment. A high somaclonal variation frequency (47.5%
) for the awnless characteristic was observed. The variation was not d
ue to mutation of the dominant gene An, but was apparently caused by m
utation of an inhibitor gene. Results from these experiments suggested
that somaclonal variation was not always random and that specific loc
i had higher mutation rates during the somaculture procedure. Previous
suggestions that changes in DNA methylation pattern caused by the tis
sue culture process were not random were consistent with the data from
our studies.