12. Variation and inheritance of seed shedding in weedy rice

Lin-Hua TANG1, and Hiroko MORISHIMA2

1) Institute of Food Crops, Jiangsu Academy of Agricultural Sciences, Nanjing, 210014 China

2) National Institute of Genetics, Mishima, 411 Japan

The "weed rice" has long persisted as a weed of rice fields in some areas and seems to have a strong weediness (Tang and Morishima 1988). In outward look before heading, it is quite similar to coexisting cultivated rice. Seed shedding coupled with seed dormancy is characteristic of wild as well as weed rices. The degree of shedding observed in wild and weed rices is much higher than that exhibited by shedding genes in cultivated rice (Jodon 1955; Rao and Misro 1968).

To investigate the mode of inheritance of seed shedding and other weediness traits, a number of crosses were made using weed-rice strains collected from Korea (C9537; C9541), Nepal (C9520; C9491) and Thailand (SS18), and cultivated strains (108 and T65: Indica and Japonica testers, respectively; C9489 and C9493, from Nepal, both coexisting with C9491). Their F2 populations were examined in Mishima. To quantify the degree of shedding, the numbers of naturally shed seeds, seeds shed after hand-gripping and those persistent on the panicle were counted in two panicles per plant at maturity, giving scores 2, 1 and 0 respectively, and the mean score was taken as representing a given plant.

Different patterns of F2 segregation were observed as shown in Fig. 1, i.e., 1) Normal distributions ranging between parental values observed in crosses with Korean weedy strains (108 X C9537; T65 X C9541), 2) negative skew distributions observed in a cross with Thai weedy strain (T65 X SS18) and a cross between a Nepalese and a Korean strain (C952O X C9537-5). The skewed distribution is probably because naturally shed seed were all lumped under score 2 even though they could have varying degrees of shedding. 3) Discontinuous distributions suggesting 1:3 (T65 X C9520) and 1:6:9 (T65 X C9489) ratios found in crosses with Nepalese weedy strains, and 4) crosses with Nepalese cultivated strains which were non-shedding showed either no segregation (108XC9491) or transgression toward higher shedding (T65 X C9489; T65 X C9493; these data are not shown in Fig. 1).

From the above results, following inferences may be drawn. There are a number of loci controlling seed shedding with major and minor effects. When major genes are involved, shedding is dominant over non-shedding. The occurrence of shedding segregants from crosses between non-shedding parents suggests complementary action between loci.

Fig. 1. Frequency distributions of seed shedding in eight F2 populations.


Jodon, N. E., 1955. Present status of rice genetics. J. Agr. Assoc., China 10(N.S.): 5-21.

Rao, A. S. and B. Misro, 1968. Linkage sudies in rice. VII inheritance of genes governing long palea, red pericarp, grain shape and shattering of grain and their interrelationships. Oryza, Cuttack 5(1): 5-9.

Tang, L. H. and H. Morishima, 1988. Characteristics of weed rice strains. RGN 5: 70-72.