Hiroko Morishima, Visiting Prof., Tokyo University of Agriculture, Tokyo, JAPAN (Rice Genetics IV. Proc. of 4th Intern. Rice Genet. Symp., 2000. Khush, G.S. et al., ed. 2001. Science Publishers, India & IRRI, Philippines)

        This chapter intends to explore some implications of rice evolution from the viewpoints of genetics and ecology. Core issues are (1) what genetic changes are associated with differentiation among and within species of cultivated rice and their wild relatives and (2) what factors are responsible for the domestication process. Genetic diversity among and within AA genome species is summarized. Second, four directions of differentiation within the Asian AA genome gene pool are clarified: differentiation from wild to cultivated type (domestication), ecotype differentiation from perennials to annuals in wild races, geographical variation in wild races, and indica-japonica differentiation in cultivars. Third, the genetic basis of the domestication syndrome is discussed. Our recent study demonstrated that mapped genomic locations of quantitative trait loci (QTLs) tended to cluster, reflecting the domestication syndrome as well as the indica-japonica syndrome. This phenomenon was explained by "multifactorial linkages." Domestication might be a process driven by conscious and unconscious selection of adaptive gene blocks distributed over the genome.

        In the past decade, a wealth of data provided by molecular markers, together with phenotypic, ecological, and archaeological data, significantly increased our evolutionary understanding of the genus Oryza. The target species dealt with in this chapter are diploid AA genome species - cultivated rice and its wild relatives. Several important problems such as the genetic basis of reproductive isolation are not included, but some new information obtained from our recent studies is discussed.

        Inter- and intraspecific genetic diversity in the O. sativa complex

        The two cultivated rice species, Oryza sativa L. and O. glaberrima Steud., belong to a species group called Oryza sativa complex together with the five wild taxa, O. rufipogon (sensu lato), O. longistaminata Chev. et Roehr., O. barthii A. Chev., O. glumaepatula Steud., and O. meridionalis Ng. This species complex was first defined as the diploid species having the genome A in common. Later, the genetic similarity among these species was confirmed using isozymes (Second 1991), restriction fragment length polymorphisms (RFLPs) of nuclear DNA (Wang ZY et al. 1992), cpDNA (Dally and Second 1990) and mtDNA (Second and Wang 1992), and amplified fragment length polymorphism (AFLP) (Aggarwal et al. 1999). This might indicate that nuclear and organellar genomes that are considered to undergo different evolutionary rates evolved concertedly at this level.

        Among these taxa, only O. rufipogon produces fertile F₁ hybrids with O. sativa and therefore these two species are considered to belong to a single biological species. Together with all circumstantial evidence, this suggests that O. rufipogon is the ancestor of O. sativa. Similarly, it leaves no doubt that O. barthii is the ancestor of African rice O. glaberrima.

Species relationships among the AA genome wild taxa

        Oryza rufipogon is widely distributed in tropical and subtropical areas in Asia and Oceania and tends to differentiate into perennial and annual types. O. barthii is the annual species found mainly in West Africa. Another African species, O. longistaminata, is strongly perennial. O. glumaepatula is distributed in Latin America and varies in perenniality. O. meridionalis is an annual species found in Australia. Taxonomically, O. longistaminata is distinguished by its strong rhyzomatous habit and O. barthii by its short ligule from other species. But Asian, American, and Oceanian taxa are barely distinguishable from each other by morphology only because of a lack of clear key characters. Therefore, those species are primarily classified on the basis of geographical origins. These five wild taxa are isolated from each other by various reproductive barriers.

        Nomenclature of the Asian AA genome wild taxon has been a controversial subject. The perennial and annual types are often designated as O. rufipogon Griff. and O. nivara Sharma et Shastry, respectively. Our group has lumped both types in a single species, O. rufipogon, since they are interfertile and variation between the two types is continuous (Oka 1988).

*Copied with permission of the author and courtesy of the International Rice Research Institute, Los Baños, Laguna, Philippines.

        With the hope of elucidating their phylogenetic relationships, many workers have studied genetic diversity among and within species. Nuclear gene markers demonstrated the distinct difference among species (Second 1985, Wang ZY et al. 1992, Doi et al. 1995, Ishii et al. 1996). A trend commonly obtained from those studies was that O. meridionalis was remotely related to other species that are relatively close. The variation pattern revealed in organellar markers was more complex (Dally and Second 1990, Second and Wang 1992, Ishii et al. 1988). The species relationships revealed were not always consistent among the different studies, perhaps partly because of the quantity and quality of the accessions used in the respective studies. Since wild taxa are polymorphic and heterogeneous, a good sample representing the whole genetic variation involved in the respective species should be used.

        Recently, Akimoto (1999) made a biosystematic study using 183 accessions of the AA genome species preserved at the National Institute of Genetics, Mishima. Figure 1 illustrates the rough sketch of his results. In isozyme and nuclear DNA RFLP, as perceived from previous studies, O. meridionalis (or O. meridionalis and O. longistaminata) is remote from other species that are distinct but relatively close to each other. O. glumaepatula showed closer similarity with O. longistaminata in isozyme and with O. barthii in RFLP. In organellar markers, species difference was distinct except for O. glumaepatula. Strains collected from northern South America together with those from Central America showed similarity with O. longistaminata in mtDNA, whereas those from the Amazon basin showed similarity with O. barthii in mtDNA as well as in cpDNA. Genetic closeness between O. glumaepatula and both African taxa was previously pointed out by Wang ZY et al. (1992), Doi et al. (1995), and Ge et al. (1999) based on a smaller number of accessions. On the other hand, similarity between some accessions of O. glumaepatula and O. rufipogon was pointed out by many authors, such as Dally and Second (1990), Doi et al. (1995), and Juliano et al. (1998). In Akimoto's study (1999), 57 O. glumaepatula accessions, excluding apparently weedy or introgressed types, did not show similarity with O. rufipogon. Yet, O. glumaepatula is undoubtedly a heterogeneous species and its phylogenetic status is still problematic.

        Fig. 1. Schematic diagrams showing genetic diversity among and within species of AA genome wild taxa based on isozyme (A), nuclear DNA restriction fragment length polymorphism (RFLP) (B), mtDNA RFLP (C), and cpDNA RFLP (D). (Redrawn from Akimoto 1999.)

        Interspecific relatedness was also discussed based on the precise analysis of particular genes including mobile elements (rDNA, Sano and Sano 1990, Cordesse et al. 1990; SINE, Hirano et al. 1994; CatA, Iwamoto et al. 1999; Adh, Ge et al. 1999; MITE, Kanazawa et al. 2000). The finding of such specific DNA will increase our understanding of the differentiation process at the molecular level. It should be noted, however, that gene lineage thus revealed does not necessarily reflect speciation.

        The amount of genetic diversity within species is important to understand the evolutionary status of the species. O. sativa has lower genetic diversity at the molecular level than O. rufipogon. This contrasts with the variation pattern observed at the phenotypic level, in which O. sativa is more diverse than O. rufipogon. The genetic diversity in O. glaberrima is much lower than that in O. sativa. This can be accounted for by the difference in breeding system of their respective wild ancestors.

        Among wild taxa, O. rufipogon generally showed the highest diversity, followed by O. longistaminata or O. glumaepatula (Table 1). This may be attributable to its variability in perenniality associated with allogamy, which conferred a potentiality to preserve genetic diversity in the populations (Barbier 1989). Evolution from perennials to annuals and from outbreeders to inbreeders is a general trend found in higher plants. Yet, precise phylogeny in this species complex remains unsolved.

        Table 1. Genetic diversity within species at isozyme and DNA levels. H and H' stand for average gene diversity and diversity index, respectively. (Modified from Morishima et al. 1992) ^aRecomputed from Second (1985). ^bAkimoto (1999). ^cSano and Sano (1990). ^dRecomputed from Dally and Second (1990). ^eKanazawa et al. (2000).

        Evolutionary trends in the Asian AA genome gene pool

        Rice has two primary gene pools, corresponding to O. sativa and O. glaberrima, which contain cultivated races and their wild and weedy relatives, respectively. Within the primary gene pool of O. sativa, four directions of differentiation are recognized (Fig. 2): (1) differentiation from wild to cultivated types, (2) differentiation from perennial to annual types in wild races, (3) geographical differentiation in wild races, and (4) varietal differentiation toward indica and japonica types.

        Fig. 2. Differentiation within the primary gene pool of O. sativa.

Differentiation from wild to cultivated types: domestication

        In seed crops, the cultivated type is characterized by nonshedding of seeds, rapid and uniform germination, efficient seed production, and determinate growth in comparison with the wild type. At the incipient stage of domestication, planting harvested seeds by man automatically selected this "adaptive syndrome of domestication" (Harlan 1975). This holds true in rice.

        Oryza sativa and O. rufipogon are genetically very close in spite of their clear phenotypic difference, and barely distinguishable by molecular markers. Wild and cultivated plants easily interbreed if grown nearby. Gene flow is mainly from predominantly inbreeding cultivated races to partially outbreeding wild races (outcrossing rate ranging from 10% to 60%, Oka 1988). Gene flow might have played an important role in diversification of the domesticates as in many other crops. Even now, natural hybridization between wild and cultivated rice occurs frequently and hybrid derivatives are found abundantly as weed types. We rarely find truly wild populations without introgression of genes from cultivated rice in tropical rice-growing areas.

        Perennial and annual types exhibit contrasting life-history traits that characterize fecundity/survivorship schedules of the individuals. The perennial (polycarpic) type shows vigorous vegetative growth, low seed productivity, late flowering, and a high outcrossing rate, whereas the annual (monocarpic) type shows the opposite characteristics (Oka 1988, Sano and Morishima 1982).

        Within the geographical range of O. rufipogon spreading in Asia and Oceania, distribution of the truly wild annual populations is restricted to tropical continental Asia. In this area, perennial and annual populations are allopatric because of their different habitats. The perennial populations prefer habitats characterized by deeper water and less disturbance. The annual populations are in temporary swamps that are parched in the dry season. These two types reflect differentiation of adaptive strategy resulting from natural selection of habitat conditions (Sano and Morishima 1982, Morishima et al. 1984).

        Perennial and annual types are phenotypically different but genetically close and no reproductive barrier exists. Life-history traits characterizing the two types vary continuously in nature and segregate continuously in an F₂ population, suggesting that those traits are controlled by multiple factors. Few markers tend to distribute differentially between perennial and annual populations. Polymorphism at some isozyme loci such as Pox1 and others (Morishima 1991), the presence or absence of the open reading frame (ORF)100 region of cpDNA (Chen et al. 1993), and the presence or absence of a locus of miniature inverted-repeat transposable elements - MITE (Kanazawa et al. 2000) were found to associate with perennial versus annual differentiation to some extent.

        In phenotypic characters, geographical variation is not so distinct in O. rufipogon. Polymorphism at molecular markers, however, revealed a trend of geographical differentiation. Our isozyme study demonstrated (Fig. 3) that the strains collected in South Asia (particularly on the west coast of India), Southeast Asia (including the east coast of India), and China tend to differentiate (Cai et al. 1996). Recently, similar geographical differentiation was observed in organellar markers (Akimoto 1999). The geographical differentiation in O. rufipogon is accounted for by "isolation by distance" and adaptation to local environments that occurred during the expansion of this species in Asia and Oceania. A reproductive isolation barrier does not seem to develop within O. rufipogon, though a more extensive survey is needed.

        Fig. 3. Scatter diagram of O. rufipogon accessions plotted by the first and second scores of factor analysis based on 29 polymorphic isozymes. (Cai and Morishima, unpublished.)

        The indica-japonica problem has been argued repeatedly as reviewed by Oka (1988). Various molecular studies consistently showed a distinct difference between these two major varietal groups in nuclear (Second 1982, Glaszmann 1987, Wang and Tanksley 1989, Nakano et al. 1992) and organellar genomes (Kadowaki et al. 1988, Dally and Second 1990, Second and Wang 1992, Ishii et al. 1993, Chen et al. 1993,). When typical indica and japonica types are compared, many genes and characters showed nonrandom association with each other to clearly separate the two types.

        When a large number of primitive cultivars are analyzed without a priori criteria, indica versus japonica differentiation is the principal variation, but some varieties that do not belong to either of the two types are also found. Such atypical cultivars are not necessarily intermediate on the axis distinguishing indica and japonica types or a recombined type, but seem to vary on a different variation axis. Himalayan hilly areas are known as the homeland of the atypical varieties (Sano and Morishima 1992). Further, some cultivars grown in low-lying deepwater areas in the Bengal delta are also atypical types (Glaszmann 1987, Hakoda et al. 1990). A recent study by Cai and Morishima (2000a) demonstrated that primitive cultivars collected in deepwater areas in Bangladesh included a japonica-like group and an atypical group, which did not belong to either type (Table 2). It was further suggested that geographical differentiation could precede seasonal ecotype differentiation (aman, aus, and boro).

Table 2. Classification of indigenous cultivars collected in Bangladesh deepwater areas. Numbers in parentheses show the number of cultivars collected in Khulna District. (From Cai and Morishima 2000a.) ^aBased on cluster analysis of 8 polymorphic isozymes.

        The four directions of differentiation mentioned above are conceptually independent from each other. Yet, they have proceeded, probably interacting with each other in the process of domestication. The evolutionary role of perenniality, geographical variation, and indica versus japonica differentiation will be discussed below in relation to the domestication process.

        Which is the immediate ancestor of domesticates, the perennial or annual type? It has been a subject of discussion whether the perennial or annual type is the ancestor of O. sativa. Sano et al. (1980) inferred that the perennial-annual intermediate type is most probably the immediate ancestor. "Intermediate type" implies the population that is habitually clonal and partially outbreeding but can propagate sexually if seed propagation is advantageous. Such populations are now mostly the secondary products of natural hybridization between perennial wild and cultivated plants. When the primitive perennial population was exposed to disturbed or dry conditions, the population genotype probably shifted toward wild annuals in natural habitats or a primitive cultivated type in man-made habitats through the "intermediate type." A trade-off which constrains energy allocation into sexual and asexual reproduction might have moved the plants toward higher seed yielders.

        Which geographical races of wild rice evolved into incipient domesticates? Older rice remains excavated to date are concentrated in the middle and lower basin of the Yangtze River in China. Though O. rufipogon is not distributed at present in this area, some evidence suggests the existence of wild rice in the past (Sato et al. 1991). It seems difficult, however, to determine the exact place of origin of cultivated rice until more archaeological evidence, particularly on wild rice, will accumulate.

        For genetic relationships between particular wild and cultivated rice, the following subjects are worth considering. First, Chinese wild rice strains, in particular those collected in the northern fringe of distribution of O. rufipogon, have some japonica-specific genes and characters (Second 1985, Sano et al. 1989, Morishima and Gadrinab 1987, Nakano et al. 1992, Cai et al. 1995). Second, some indigenous cultivars grown in deepwater areas in Bangladesh carry particular isozyme alleles such as Est10-4 and Amp5-4, which are rare in cultivars but not rare in wild populations (Cai and Morishima 2000). Third, the annual wild rice distributed on the west coast of India and primitive cultivars growing nearby showed high genetic similarity in isozymes, and both showed similarity with the japonica type in cpDNA as well as in rDNA according to Lolo and Second (1988).

        Are indica and japonica types monophyletic or dyphyletic? Oka and his group considered that indica and japonica types have diverged as domestication proceeded. This view is mainly based on the fact that wild rice has a potentiality to evolve into indica as well as japonica types (Oka and Morishima 1982) and that indica versus japonica differentiation was not found in O. rufipogon. A recent collection of O. rufipogon obtained from a broader geographical range, however, yielded a slightly different variation pattern. A tendency toward indica-japonica differentiation in terms of particular association among genes or characters was observed among wild strains (Second 1985, Morishima and Gadrinab 1987, Dally and Second 1990, Nakano et al. 1992, Sun et al. 1996a,b,c), though the degree of nonrandom association is much lower than in cultivars.

        The dyphyletic hypothesis postulates that indica and japonica types originated from different lineages of O. rufipogon (Second 1982, Sato 1996). Accumulated observations that indica and japonica types consistently showed a clear difference between each other and closer affinity with different wild accessions than with each other seem to support the dyphyletic hypothesis. Recent archaeological excavations in China and analysis of rice remains (phytolith, Wang et al. 1998; DNA, Sato et al. 1995) suggest that Chinese wild rice played an important role in the origin of japonica rice.

        I am inclined to the view that rice domestication has been a diffused process in both space and time. During a long period, prototypes of indica and japonica types have probably become two dominant groups and have accumulated marked differences in (?) keeping their respective genic constitution inherited from their founders.

        Genetic basis of the domestication syndrome

        Related species or ecotypes are differentiated by a particular pattern of association between states of different characters or between alleles at different loci. Such multilocus covariation is called gametic disequilibrium. In rice, variations between wild and cultivated types, between perennial and annual types, and between indica and japonica types are good examples of gametic disequilibrium. Nonrandom association is caused by various factors such as selection (coadaptation), linkage, pleiotropy, and founder effect, as discussed by Hedrick et al. (1978). It is difficult to dissect the underlying mechanism without elaborate experiments.

        In rice, linkage blocks harboring genes for internal barriers and fitness traits were detected by isolating the relevant chromosomal segments using the backcross method (Sano 1992). Key factors for gametic disequilibrium in quantitative traits can be elucidated to some extent by observing the shift of correlations in the hybrid generations. Parental associations that disappear in the F₂ are mostly due to selection for coadapted traits. On the other hand, those that remain in the F₂ or early generations are probably due to linkage and/or pleiotropy. Many studies showed that parental associations partly persisted in the F₂ (character coherence) though they were weak but significant. In the crosses of wild-cultivated, annual- perennial, and indica-japonica rice strains, we have experienced that parental associations in quantitative traits mostly disappeared in the F₂. This indicates that those associations are not due to linkage or pleiotropy of a few major genes.

        Advances in molecular genetics enabled us to dissect the genetic basis of quantitative differences between species or ecotypes, which has been studied only by statistical-genetics methods until recently. Targeting differentiation in the Asian AA genome gene pool, we performed QTL analysis in the cross O. sativa (indica)-O. rufipogon (perennial type) (Cai and Morishima, n.d.). Based on 125 recombinant inbred lines (RILs), 148 markers were mapped. Because this wild parent carried some japonica-specific characteristics, indica-japonica diagnostic traits segregated in this mapping population in addition to domestication-related traits. Among 34 quantitative traits examined, 22 traits revealed 1-29 putative QTLs, respectively. Several QTLs for domestication-related traits were mapped over the whole genome, showing a tendency to form clusters, each reflecting the domestication syndrome. The same situation was observed in QTLs for indica versus japonica diagnostic traits. In addition, some of the clusters joined the domestication syndrome clusters. Figure 4 presents examples of gene clusters. Several isozyme loci that serve as diagnostic markers to distinguish wild and cultivated types (Est10, Wang XK et al. 1992) and indica and japonica types (such as Cat1 and Amp2) were found in or near the QTL clusters. Similar cluster phenomena of domestication-related QTLs were reported by Koinange et al. (1996) in common bean and by Xiong et al. (1999) in rice.

        To determine whether QTL clusters are loosely linked loci or a single locus with a pleiotropic effect, more precise analysis is necessary. It is interesting to note that QTLs for mesocotyl length, which is considered to reflect endogenous hormone level, tended to join those clusters. This might suggest a possibility that the clustering phenomenon is partly due to the pleiotropic effect of an unknown key factor controlling various traits through hormonal regulation.

        In this [O. sativa-O. rufipogon cross, character correlations were generally weak among RILs that represent essentially F₂ variation. This implies that nonrandom character associations of the domestication syndrome are mainly due to natural selection for coadapted traits. The present analysis revealed several QTL linkages reflecting the domestication syndrome. Since they were located on different chromosomes, they segregated into RILs, resulting in decreased correlations. This phenomenon could be understood by "multifactorial linkages" advocated by Grant (1975). This phenomenon is inevitably caused by random distribution of multiple factors over the limited number of chromosomes that determine the differences in two or more quantitative characters.

        QTL clusters thus identified were mapped on the regions in which cultivar-derived alleles segregated at higher frequencies than expected. This coincidence could be explained by unconscious selection, which worked under cultivation pressure when establishing RILs favoring the combination of cultivar-derived gene blocks.

        The domestication process is undoubtedly a gradual process directed by man. In rice, Oka and Morishima (1971) demonstrated early that "cultivation pressure" brought about a rapid change in population genotype toward the cultivated type as suggested by Harlan (1975). Seed shedding is the critical trait disruptively selected in natural and cultivated fields. The genetic basis of seed shedding is simpler than other domestication-related traits. Eiguchi and Sano (1990) identified two loci conferring high seed shedding on O. rufipogon, one of which was linked to the spreading panicle locus on chromosome 4. Our QTL analysis detected five loci for seed shedding and one of them located on chromosome 1 was linked with another spreading panicle locus (Cai and Morishima 2000b). A spreading panicle is advantageous for wild rice to disperse seeds but disadvantageous for cultivated rice. It may be reasonable to infer that selection for nonshedding genes with relatively large effects played an important role as a trigger for domestication.

        Our experiments with another cross O. sativa-O. rufipogon showed that association among seed shedding, seed dormancy, and awn length tended to increase in later hybrid generations without deliberate selection (Table 3-1). In the cross propagated in bulk, the outcrossing rate was higher in the lines with wild characteristics than in those with cultivated-rice characters (Table 3-2). This suggests that flower characteristics to increase the selfing rate would be selected by linkage drag. Our QTL analysis indicated that loci governing those traits, including anther length, which is known to positively correlate with outcrossing rate, are linked with each other (Fig. 4).

Table 3. Changes in population parameters in cultivated-wild hybrids. **, * significant at 1% and 5% levels, respectively

Fig. 4. Examples of QTL clusters (chromosomes 1 and 8). Genomic regions affecting the respective traits (LOD>2.8) are shown by the bars to the right of linkage groups. Character code: SHA = seed shattering, DOR = seed dormancy, AWL = awn length, ANL = anther length, P/T = panicle number/tiller number, DTH = days to heading, BVP = basic vegetative phase, KCL = KClO₃ resistance, APH = apiculus hair length, GMS = germination speed, LTR = low-temperature resistance, PTB = panicle neck to lowest branch, MSL = mesocotyl length, SWD = spikelet width. The small spindle on the chromosome shows the approximate position of centromere. (Cai and Morishima, n.d.)

Pernes (1983) predicted that linkages of domestication-related traits could be adaptive, particularly in the outbreeding species, in which recovery of the cultivated type following frequent outcrossing between wild and cultivated types might be facilitated by such linkages. In rice, wild rice is partially outbreeding though current cultivars are predominantly inbreeding, and natural hybridization has played a significant role in rice evolution during and after domestication. The domestication syndrome could be molded by unconscious and conscious selection of adaptive gene blocks distributed over the genome.

1. Aggarwal RK, Brar DS, Nandi S, Huang N, Khush GS. 1999. Phylogenetic relationships among Oryza species revealed by AFLP markers. Theor. Appl. Genet. 98:1320-1328.
2. Akimoto M. 1999. Bio-systematics of the Oryza AA genome wild taxa (Oryza sativa complex) by morpho-physiological traits, isozymes and RFLPs. PhD thesis. Hokkaido University, Japan.
3. Barbier P. 1989. Genetic variation and ecotypic differentiation in the wild rice species Oryza rufipogon. II. Influence of the mating system and life-history traits on the genetic structure of populations. Jpn. J. Genet. 64:273-285.
4. Cai HW, Wang XK, Morishima H. 1995. Isozyme variation in Asian common wild rice, Oryza rufipogon. Rice Genet. Newsl. 12:178-180.
5. Cai HW, Wang XK, Morishima H. 1996. Geographical differentiation of annual type of O. rufipogon. 13:67-69.
6. Cai HW, Morishima H. 2000a. Diversity of rice varieties and cropping system in Bangladesh deepwater areas. JARQ. (In press.)
7. Cai HW, Morishima H. 2000b. Genomic regions affecting seed shattering and seed dormancy in rice. Theor. Appl. Genet. 100:840-846.
8. Cai HW, Morishima H. n.d. QTL clusters reflect character associations in wild and cultivated rice. Theor. Appl. Genet. (Submitted.)
9. Chen WB, Nakamura I, Sato YI, Nakai H. 1993. Distribution of deletion type in cpDNA of cultivated and wild rice. Jpn. J. Genet. 68:597-603.
10. Cordesse F, Second G, Delseny M. 1990. Ribosmal gene spacer length variability in cultivated and wild rice species. Theor. Appl. Genet. 79:81-88.
11. Dally AM, Second G. 1990. Chloroplast DNA diversity in wild and cultivated species of rice (genus Oryza, section Oryza): cladistic-mutation and genetic-distance analysis. Theor. Appl. Genet. 80:209-222.
12. Doi K, Yoshimura A, Nakano N, Iwata N. 1995. Phylogenetic study of A-genome species of genus Oryza using nuclear RFLP. Rice Genet. Newsl. 12:160-162.
13. Eiguchi M, Sano Y. 1990. A gene complex responsible for seed shattering and panicle spreading found in common wild rice. Rice Genet. Newsl. 7:105-107.
14. Glaszmann JC. 1987. Isozymes and classification of Asian rice varieties. Theor. Appl. Genet. 74:21-30.
15. Ge S, Sang T, Lu BR, Hong DY. 1999. Phylogeny of rice genomes with emphasis on origin of allotetraploid species. Proc. Natl. Acad. Sci. USA 96:14400-14405.
16. Grant V. 1975. Genetics of flowering plants. New York (USA): Columbia University Press. 563 p.
17. Hakoda H, Inouye J, Morishima H. 1990. Isozyme diversity found among Asian deepwater rices. Rice Genet. Newsl. 7:91-93.
18. Harlan JR. 1975. Crops and man. Madison, Wis. (USA): American Society of Agronomy. 295 p.
19. Hedrick PW, Jain SK, Holden L. 1978. Multilocus systems in evolution. Evol. Biol. 11:101-184.
20. Hirano HY, Mochizuki K, Umeda M, Ohtsubo H, Ohtsubo E, Sano Y. 1994. Retrotransposition of a plant SINE into the wx locus during evolution of rice. J. Mol. Evol. 38:132-137.
21. Ishii T, Terauchi T, Tsunewaki K. 1988. Restriction endonuclease analysis of chloroplast DNA from A-genome diploid species of rice. Jpn. J. Genet. 63:523-536.
22. Ishii T, Terauchi T, Mori N, Tsunewaki K. 1993. Comparative study on chloroplast, mitochondrial and nuclear genome differentiation in two cultivated rice species, O. sativa and O. glaberrima, by RFLP analysis. Theor. Appl. Genet. 86:88-96.
23. Ishii T, Nakano T, Maeda H, Kamijima O. 1996. Phylogenetic relationships in A-genome species of rice as revealed by RAPD analysis. Genes Genet. Syst. 71:195-201.
24. Iwamoto M, Nagashima H, Nagamine T, Higo H, Higo K. 1999. p-SINE1-like intron of the CatA catalase homologs and phylogenetic relationships among AA-genome Oryza and related species. Theor. Appl. Genet. 98:853-886.
25. Juliano AB, Naredo MEB, Jackson MT. 1998. Taxonomic status of Oryza glumaepatula Steud. I. Comparative morphological studies of New World diploids and Asian AA genome species. Genet. Res. Crop Evol. 45:197-203.
26. Kadowaki K, Yazaki K, Osumi T, Harada K, Katsuta M, Nakagahra M. 1988. Distribution of mitochondrial plasmid-like DNA in cultivated rice (Oryza sativa L.) and its relationship with varietal groups. Theor. Appl. Genet. 76:809-814.
27. Kanazawa A, Akimoto A, Morishima H, Shimamoto Y. 2000. Inter- and intra-specific distribution of Stowaway transposable elements in AA genome of wild rice. Theor. Appl. Genet. 101:327-335.
28. Koinange EMK, Singh SO, Gepts P. 1996. Genetic control of the domestication syndrome in the common bean. Crop Sci. 36:1037-1045.
29. Lolo OM, Second G. 1988. Peculiar genetic characteristics of O. rufipogon from Western India. Rice Genet. Newsl. 5:67-70.
30. Morishima H. 1991. Association between Pox-1 variation and seed productivity potential in wild rice. In: Rice genetics II. Manila (Philippines): International Rice Research Institute. p 1-9.
31. Morishima H, Sano Y, Oka HI. 1984. Differentiation of perennial and annual populations of the Asian wild rice Oryza perennis. Plant Syst. Evol. 144:119-135.
32. Morishima H, Gadrinab LU. 1987. Are the Asian common wild rices differentiated into the Indica and Japonica types? In: Hsieh SC, editor. Crop exploration and utilization of genetic resources. Proceedings of the Conference. Changhua (Taiwan): Taichung District Agricultural Improvement Station. p 11-20.
33. Morishima H, Sano Y, Oka HI. 1992. Evolutionary studies in cultivated rice and its wild relatives. Oxford Surv. Evol. Biol. 8:135-184.
34. Nakano M, Yoshimura A, Iwata N. 1992. Phylogenetic study of cultivated rice and its wild relatives by RFLP. Rice Genet. Newsl. 9:132-134.
35. Oka HI. 1988. Origin of cultivated rice. Tokyo (Japan): Japanese Science Society Press. 254 p.
36. Oka HI, Morishima H. 1971. The dynamics of plant domstication: cultivation experiments with Oryza perennis and its hybrid with O. sativa. Evolution 25:356-364.
37. Oka HI, Morishima H. 1982. Phylogenetic differentiation of cultivated rice. 23. Potentiality of wild progenitors to evolve the Indica and Japonica types of rice cultivars. Euphytica 31:41-50.
38. Pernes J. 1983. The genetics of domestication in cereals. La Recherche 14:910-919. (In French.)
39. Sano R, Morishima H. 1992. Indica-Japonica diferentiation of rice cultivars viewed from the variation in key characters and isozymes with special reference to land races from the Himalayan hilly areas. Theor. Appl. Genet. 84:266-274.
40. Sano Y. 1992. Genetic comparisons of chromosome 6 between wild and cultivated rice. Jpn. J. Breed. 42:561-572.
41. Sano Y, Morishima H, Oka HI. 1980. Intermediate perennial-annual populations of Oryza perennis found in Thailand and their evolutionary significance. Bot. Mag. Tokyo 93:291-305.
42. Sano Y, Morishima H. 1982. Variation in resource allocation and adaptive strategy of a wild rice, Oryza perennis Moench. Bot. Gaz. 143:518-523.
43. Sano Y, Yi HX, Shao QQ, Iyama S. 1989. Ribosomal DNA spacer-length variations in a wild rice population from Dongxiang, China. In: Breeding research: the key to the survival of the earth. Proceedings of the 6th International SABRAO Congress, Tokyo, Japan. p 493-496.
44. Sano Y, Sano R. 1990. Variation of the intergenic spacer region of ribosomal DNA in cultivated and wild rice species. Genome 33:209-218.
45. Sato YI. 1996. Origin of rice and rice cultivation based on DNA analysis. Tokyo (Japan): NHK Books. 227 p. (In Japanese.)
46. Sato YI, Tang SX, Yang LU, Tang LH. 1991. Wild rice seeds found in an oldest rice remain. Rice Genet. Newsl. 8:76-78.
47. Sato YI, Tang LH, Nakamura I. 1995. Amplification of DNA fragments from charred rice grains by polymerase chain reaction. Rice Genet. Newsl. 12:260-261.
48. Second G. 1982. Origin of the genetic diversity of cultivated rice: study of the polymorphism scored at 40 isozyme loci. Jpn. J. Genet. 57:25-57.
49. Second G. 1985. Evolutionary relationships in the Sativa group of Oryza based on isozyme data. Genet. Sel. Evol. 17:89-114.
50. Second G. 1991. Molecular markers in rice systematics and the evaluation of genetic resources. In: Bajaj YPS, editor. Biotechnology in agriculture and forestry. Vol 14. Rice. New York (USA): Springer. p 468-494.
51. Second G, Wang ZY, 1992. Mitochondrial DNA RFLP in genus Oryza and cultivated rice. Genet. Res. Crop Evol. 39:125-140.
52. Sun C, Wang XK, Yoshimura A, Iwata N. 1996a. Genetic differentiation of nuclear genome in common wild rice and cultivated rice. In: Wang XK, Sun C, editors. Origin and differentiation of Chinese cultivated rice. Beijing (China): China Agricultural University Press. p 120-133. (In Chinese with English summary.)
53. Sun C, Wang XK, Yoshimura A, Iwata N. 1996b. Genetic differentiation of mitochondrial DNA in common wild rice and cultivated rice. In: Wang XK, Sun C, editors. Origin and differentiation of Chinese cultivated rice. Beijing (China): China Agricultural University Press. p 134-139. (In Chinese with English summary.)

54. Sun C, Yoshimura A, Zichao L, Haiying Z, Wang XK, Iwata N. 1996c. Genetic differentiation of chloroplast genome in O. rufipogon Griff. and O. sativa L. In: Wang XK, Sun C, editors. Origin and differentiation of Chinese cultivated rice. Beijing (China): China Agricultural University Press. p 140-146. (In Chinese with English summary.)

55. Wang C, Udatsu T, Tang L, Zhou J, Zheng Y, Sasaki A, Yanagisawa K, Fujiwara H. 1998. Cultivar group of rice cultivated in Caoxieshan site (B.P. 6000-present) determined by the morphology of plant opals and its historical change. Breed. Sci. 48:387-394. (In Japanese with English summary.)
56. Wang XK, Cai HW, Cheng KS. 1992. The discovery of an Est locus related to the origin, evolution and classification of Asian rice. Chin. Rice Res. Newsl. 7:1-2.
57. Wang ZY, Tanksley SD. 1989. Restriction fragment length polymorphism in Oryza sativa L. Genome 32:1113-1118.
58. Wang ZY, Second G, Tanksley SD. 1992. Polymorphism and phylogenetic relationships among species in the genus Oryza as determined by analysis of nuclear RFLPs. Theor. Appl. Genet. 83:565-581.
59. Xiong LZ, Liu KD, Dai KK, Xu CG, Zhang Q. 1999. Identification of genetic factors controlling domestication-related traits of rice using an F₂ population of a cross between Oryza sativa and O. rufipogon. Theor. Appl. Genet. 98:243-251.