37. Identification of wide compatibility genes in rice

So-called wide-compatible rice varieties (WCV) produce fertile F1 plants in crosses with both Indica and Japonica varieties. They are useful in exploiting heterosis in Indica/Japonica crosses. Such varieties were first identified in Japan (Ikehashi and Araki 1984). Subsequently, eight WCVs were also found at IRRI (Vijaya Kumar and Virmani 1987; IRRI 1988; Vijaya Kumar and Virmani 1988).

We made experiments to identify "wide compatibility" genes in five of the eight WCV cultivars, viz., Dular, Palawan, Moroberekan, BPI 76 and N22. Nine crosses were made among these five cultivars in 1988 dry season to obtain 15-20 F1 seeds per cross. In 1988 dry season, seven of the F1 hybrids were testcrossed with a Japonica variety, Taichung 65, producing 50-100 seeds per cross. The three-way cross and single cross F1 populations were grown in 1988 wet season, and spikelet fertility in two primary panicles was observed on an individual plant basis. When all F1 plants from a three-way cross showed fertilities coniparable to those of single-cross F1 plants, the Prst two varieties involved in the three- way cross were considered to carry the same genes for wild compatibility.

Spikelet fertilities in two single crosses between "WC" cultivars ranged from 61 to 96% (Table 1). Accordingly, we classified plants showing spikelet fertility above 61% as fertile and those below 60% as semi-sterile or partly sterile. Three- way crosses involving Dular/Palawan, Moroberekan/Palawan and BPI 76/Moroberekan produced no semi-sterile plants (Table 1), suggesting that the wide-compatibility genes in these three varieties are at the same locus and probably identical.

The genetic basis of intervarietal F1 sterility in rice is a complex. In addition to cytoplasmic male sterility, there are at least two different hypotheses both supported by experimental evidence (Oka 1988, p. 188-195). One is the hypothesis of sporogametophytic interaction of alleles at the same locus; F1 plants with S^a/S are semi-sterile as gametes with S^a are eliminated. If there is allele S^b which does not interact with S^a nor with S, plants with are "wide-compatible". The other is the hypothesis of duplicate gametic lethals; the F1 plants with S\1\/ + \1\ S\2\/ + \2\ are partly sterile as gametes with s\1\s\2\ deteriorate. In this case, plants with +\1\/ +\1\ +\2\/+\2\ are "wide-compatible" as they do not bring about F1 sterility in crosses with both S\1\/S\1\ +\2\/ +\2\ and +\1\/ +\1\ S\2\/S\2\. Many of wild-rice (0. rufipogon) strains seem to be "wide-compatible" in this way. There may be two or more loci for S, S^a and S^b, and two or more sets of + \1\/s\1\ + \2\/s\2\. Yet, according to these hypotheses, "wide-compatible" varieties showing no F1 sterility in crosses with any self-fertile ones are expected to liave the same genotype. The occurrence of a few semi-sterile segregants in the three-way crosses as presently conducted is not explainable without assuming a particular interaction between sterility-gene systems or complementary modifiers changing the effects of sterility genes.

Ikehashi, H. and H. Araki, 1984. Varietal screening for compatibility types revealed in F1 fertility of crosses in rice. Jpn. J. Breed. 34(3): 304-312.

IRRI-International Rice Research Institute, 1988. Annual Report for 1987, p. 174. Manila.

Oka, H. I., 1988. Origin of Cultivated Rice. Elsevier/Jpn. Sci. Soc. Press, Amsterdam/Tokyo, 254 pp.

Vijaya Kumar, R. and S. S. Virmani, 1987. Screening for wide compatibility varieties in rice. Agron. Abst. Amer. Soc. Agronomy, Atlanta, p. 69.

____ and ____, 1988. Genetic analysis of wide compatibility trait in rice. Genome 30, Suppl. 1: 468.