10. A gene complex responsible for seed shattering and panicle spreading found in common wild rices

Mitsugu EIGUCHI and Yoshio SANO

National Institute of Genetics, Mishima, 411 Japan

A high rate of seed shattering or shedding is generally observed in the common wild rices, suggesting that they have been selected for this character. Their seeds are shed as soon as they get ripened. In contrast, cultivated rices show a much lower rate of seed shattering although the rates vary among varieties. In the process of domestication of wild plants, genic changes have occurred to bring about the non-shattering character of cultivars. We are engaged in studies of those genic changes.

Materials used are two Indian wild strains, W107 (an annual type) and W149 (a perennial type) and a wild strain from Amazonas, Brazil, W1192. An isogenic line of Taichung 65 with markers g and lg (T65g lg) was used as the cultivated parent. The spreading of panicles was also observed. It characterizes wild rices and helps seed dispersal along with shattering. The F1 plants between T65g lg and the wild strains showed a high rate of seed shattering and panicle spreading like the wild parents, suggesting that these traits are controlled by dominant genes. The F2 populations between T65g lg and the two Asian wild rices suggested that 

================================================================
               No. of F2 Plants                  Chi-square
Cross         ======================             A:a      B:b
              AB  Ab  aB  ab  Total              (15:1)  (3:1)
================================================================
T65g lgXW107  66  29   6   3    104               1.0     1.9
T65g lgXW149  63  31   6   1     97               1.7     3. 3
================================================================
F2 phenotypes: A-shattering, a-non-shattering, B-spreading, b-non-spreading; all chi- squares were non-significant.

Table 2. Linkage relations between lg (ligulelessness), Ph (phenol reaction) and Spr3 (t; panicle spreading) observed in F2 from T65g lgXW107 (T65g lg-1g, spr/3, ph, and W107-+Ig, Spr/3, Ph)

================================================================
Linked  genes     F2 phenotype      Recombination  X2 for
=============  =================    valuea         observed and
A/a     B/b    AB  Ab  aB  ab                      expected no.
================================================================
lg   Spr/3(t)  69  13   3  19       0.13+-0.02      9.0 (P<0.05)
lg        Ph   75   7  12  10       0.23+-0.03      4.2 (P>O-1)
Spr/3(t)  Ph   68   4  19  13       0.21+-0.03      12.7(P<0.01)
================================================================
a-Estimated by Immer's (1930) method.
seed shattering was controlled by two dominat genes and panicle spreading by one dominant gene, which were independent (Table 1). The two gene for shattering may correspond to those reported by Kadam (1936).

The dominant gene for panicle spreading was found to be linked with both lg and Ph on Chromosome 4 (Table 2). The F1 plants from T65g lgX W107 and T65g lgXW1192 were backcrossed with T65wx, 6 and 8 times, respectively. In each generation, plants showing high rates of seed shattering and panicle spreading were selected for the next backcross. In each backcross generation, plants showing the two traits were more frequent than those showing only one of them. This suggested that the genes controlling the two traits were linked, even though such an association was not found in the F2 populations.

Near-isogenic lines of T65wx with genes for seed shattering and panicle spreading were selected from the BC\6\ (from the cross with W107) and BC\8\ (from cross with W1192) generations. Segregation patterns observed in late backcross generations showed that the selected lines each carried two dominant genes controlling seed shattering and panicle spreading, respectively, and that the two genes were tightly linked with a recombination value of about 2% (Table 3). The dominant gene for seed shattering would be one of two duplicate genes found in the F2. No dominant gene for seed shattering has been reported in the list of recommended gene symbols (RGN 6, p. 18). The dominant gene detected in this

Table 3. Linkage relations between Spr\3\(t) and Sh\3\(t) observed in BC\6\F\2\ and BC\8\F\2\ populations (recurrent parent: T65wx with spr\3\ and sh\3\)

================================================================
Donor  Genera-      No. of F2 plants     Recombina-     X2 for
parent  tion    ======================   tion valuea    obs. and
                  AB  Ab  aB  ab Total                  exp. no.
================================================================
W107   BC\6\F\2\   173   1  2   34  170    0.02+/-0.01     1.6ns  
W1192  BC\8\F\2\   171   1  6   68  246    0.02+/-0.01     3.0ns  
================================================================
A-shattering, a-non-shattering, B-spreading, b-non-spreading.
a-Estimated by Immer's (1930) method.
ns-Non-significant (df =2; Ab and aB being pooled).
study is designated as Sh\3\(t) tentatively.

Regarding panicle spreading, a recessive (spr\1\) and a dominant gene (Spr\2\) have been reported so far (RGN 6, P.20). Spr\2\ could be one of complementary dominant genes reported by Mitra and Ganguli (1932) in a cross between wild and cultivated strains (cf. Kinoshita 1984, p. 216). It remains unknown whether the dominant gene detected in this study corresponds to Spr\2\. This point cannot be resolved as the genetic stocks used by Mitra and Ganguli (1932) are not available at present. But our data suggested no complementary interaction of two dominant genes for panicle spreading. Tentatively, the gene detected in the present study is symbolized Spr\3\(t). The linkage relations of this gene and Sh3(3) with lg and Ph on Chromosome 4 were estimated as: lg-Spr\3\(t)-Sh\3\(t)-Ph. The same order of these genes was confirmed also in the cross of an isogenic line from W1192, suggesting that the Asian and American races of the common wild rice have the same gene complex.

The capacity to disperse seed is important for wild rices reproduced by seed, but it is undesirable for cultivars. At least two mutational steps are needed for establishing the gene complex suppressing seed dispersal. When such gene complex was selected during domestication, it might have bcen maintained through selection by man.

References

Immer, F. R., 1930. Formulas and tables for calculating linkage intensities. Genetics 15: 81-98.

Kadam, B. S., 1936. Genetic analysis of rice, I. Proc. Ind. Acad. Sci. 4,B): 224-229.

Kinoshita, T., 1984. Gene analysis and linkage map. In Tsunoda, S. and N. Takahashi (eds.), Biology of Rice, p. 187-274. Jpn. Sci. Soc. Press/Elsevier, Tokyo/Amsterdam.

Mitra, S. K. and P. M. Ganguli, 1932. Some observations on the characters of wild rice hybrids. Ind. J. Agr. Sci. 2: 271-279.