Oryza Rufipogon

Blackwell Science, LtdOxford, UKPSBPlant Species Biology0913-557XThe Society for the Study of Species Biology, 2005August 20052028392Original ArticleCOMMON WILD RICE BIODIVERSITYZ. SONG Et al. http://www.paper.edu.cn

Plant Species Biology (2005) 20, 83–92

Genetic diversity and conservation of common wild rice (Oryza ruﬁpogon) in China

ZHIPING SONG, BO LI, JIAKUAN CHEN and BAO-RONG LU Ministry of Education Key Laboratory for Biodiversity and Ecological Engineering, Institute of Biodiversity Science, Fudan University, Handan Road 220, Shanghai 200433, China

Abstract

Common wild rice (Oryza rufipogon Griff.), known as the ancestor of Asian cultivated rice (Oryza sativa L.), is the most important germplasm for rice improvement. The first male sterility gene was found in the wild rice, and introduced to the cultivated rice, which launched the fast development of the high-yielding hybrid rice. Other agronomically beneficial traits in the wild rice, such as rice tungro virus resistance, bacterial leaf blight (Xa21 gene) resistance and acid sulfate soil tolerance, have played important roles in rice breeding. China has the northernmost distribution area of wild rice possessing great genetic diversity. However, most of the populations of this species have disappeared in China over the last three decades, mainly caused by habitat loss, fragmentation and other human disturbances. Unfortunately, the decline of existing populations still continues. In the present study, we reviewed studies on genetic diversity and conservation of this wild rice in China, concentrating on population structure, pollen competition, pollen/ gene flow from cultivated rice to wild rice, and ecological restoration in relation to in situ conservation. The relatively high genetic diversity of populations of O. rufipogon in China suggests that there is great value for conservation. Considerable gene flow from cultivated rice to wild rice may alter the genetic structure of natural populations of O. rufipogon and eventually lead to its genetic erosion. Pollen competition between wild and cultivated rice has caused a low rate of crop-to-wild gene flow, but it does not completely prevent gene flow from the crop. Effective isolation measures should be undertaken in the regions where in situ conservation of O. rufipogon is carried out. Reintroduction is an important alternative for the in situ conservation of wild rice species. As wild rice is an important genetic resource, both in situ and ex situ conservation strategies are needed. Keywords: common wild rice, conservation, gene flow, genetic diversity, genetic resource, Oryza rufipogon, restoration. Received 7 January 2005; revision received 22 March 2005; accepted 5 April 2005

Introduction crop species, particularly those in the gene pool of wild relatives of crop species, will provide many more oppor- The continuous increase of the global population, the tunities. Through millions of years of evolution and reduction of farming land, the increasing shortage of genetic adaptation to environments, the wild relatives water and the loss of rural labor to the urban centers have accumulated abundant genetic diversity. Many traits profoundly challenge the world’s food supply. To meet are unique to the wild relatives and might be beneficial to the increasing demands for food supply, the human race the improvement of cultivated species. Serving as a vast has to significantly enhance crop productivity, for which genetic reservoir, wild relatives provide elite germplasm fuller exploitation and utilization of genetic resources in for improving crop varieties by transferring beneficial genes to the crops (Lu 1996, 1998). Correspondence: Bao-Rong Lu The Asian cultivated rice, Oryza sativa L., and African Email: [email protected] cultivated rice, Oryza glaberrima Steud., are classified in

中国科技论文在线 http://www.paper.edu.cn

84 Z. SONG ET AL.

the genus Oryza, which includes over 20 wild species that turbances to its habitats (Gao 2003). Conservation and are widely distributed in the pan-tropics and subtropics, related studies of O. rufipogon have been extensively car- particularly in Asian countries (Khush 1997). The wild ried out in China. Here we briefly summarized some of species in the genus Oryza and in the related genera in the the progress made over the past decades. tribe Oryzeae constitute an exceptionally valuable gene pool for rice improvement (Lu 1996; Bellon et al. 1998; A brief description of common wild rice Zhong et al. 2000). There are many successful examples of utilizing biodiversity in the rice gene pool, particularly in Common wild rice is here referred to as the perennial wild the wild Oryza gene pool for rice improvement. The most rice O. rufipogon, although this taxon is also thought to successful example of utilizing the wild Oryza species is include the annual wild O. nivara by some authors (e.g. hybrid rice, where the male sterility (MS) gene was intro- Chang 1976; Matsuo & Hoshikawa 1993) and was once duced from the perennial common wild rice (CWR, Oryza named as Oryza perennis by Oka (1988) because the plant rufipogon Griff.) found in Hainan Province (formerly has species-specific characteristics compared with other Hainan Island, a part of Guangdong Province), China, Oryza species. O. rufipogon has short to long rhizomes and subsequently the most essential MS system of hybrid from which tillers can emerge far from the main stalk. The rice was developed (Yuan 1993). Another prominent period for its vegetative growth is quite long (Matsuo & example is the varieties of grassy stunt virus-resistant Hoshikawa 1993). rice, where the virus-resistant gene was incorporated The ligule is long and split into two at its sharp tip. The from one accession of the annual CWR (Oryza nivara panicles are open and the spikelets are slender. The anther Sharma et Shastry) collected from India (Khush 1977). of O. rufipogon is usually long and is of a yellowish-white Recently, many disease-resistant and insect-resistant color. The stigma is colored from white to dark, but genes, high-yielding genes and abiotic stress-tolerant mainly pale purple; it appears out of the glume at flow- genes have also been found in wild Oryza species (Khush ering time (Fig. 1). The grain is usually slender with a long et al. 1990; Jena & Khush 1990; Brar et al. 1996; Xiao et al. awn (unhusked), is of a dark purple color at maturity and 1996). Some of these genes have successfully been trans- falls naturally. The color of the husked grain is red. The ferred to varieties of cultivated rice. Conserving the biodi- flowering time is basically the same as that of O. sativa, versity of the wild Oryza species is therefore essential for which is responsive to photoperiods. The cross-fertility is the world’s sustainable food supply and becomes increas- high in this species. ingly important for continued availability and sustainable Oryza rufipogon is widely distributed in the tropics and use of these valuable genetic resources. subtropics of Asia (Vaughan 1994). This species is The perennial CWR, Oryza rufipogon Griff., known as reported to occur in 113 counties of eight provinces in the ancestor of Asian cultivated rice (O. sativa L.), is the south China, including Guangdong, Guangxi, Hainan, most important germplasm for rice improvement (Oka Yunnan, Hunan, Jiangxi, Fujian and Taiwan (Taiwan 1988). The collected samples of this wild rice species have O. rufipogon populations disappeared in 1978; Kiang et al. been extensively used by scientists and breeders from 1979; Fig. 2). The range of CWR in China stretches from agricultural research stations and universities for breed- 18∞09¢N to 28∞14¢N and 100∞40¢E to 121∞15¢E (Pang & ing and research. The well-known Chinese rice variety Chen 2002). ‘Zhongshan no. 1’, which is tolerant of cold tempera- tures and other abiotic stresses, was bred by Profes- Exploration and collection of common wild rice sor Ding Ying in 1931 through wide hybridization with in China O. rufipogon, except for the famous hybrid that incorporated the MS trait identified in O. rufipogon in the early The exploration and collection of wild Oryza species in 1970s (Yuan et al. 1989; Yuan 1993). Other agronomically China can be traced back to as early as 1917 when Dr E. beneficial traits, such as high-yielding, rice tungro virus D. Merrill and colleagues first found the perennial CWR resistance, elongation ability and tolerance of acid sulfate (O. rufipogon) at Lofu Mountain and the Shilong Plain in soil, found in the wild rice are of great potential for rice Guangdong Province (Wu 1990). It was Professor Ding breeding (Xiao et al. 1996; Bellon et al. 1998). China is the Ying who initiated the systematic exploration and collec- northern boundary of O. rufipogon’s natural range, where tion of wild Oryza species in China. In 1926, he found great genetic diversity has been found in its populations O. rufipogon and collected its samples in many more sites, (Wu 1990; Zhou 1995; Wang & Sun 1996; Gao 1997; Ge such as those in Guangzhou, Heiyang, Zhengcheng, et al. 1999; Song et al. 2003a). However, this species has Qingyuan and Sashui in Guangdong Province, on Hainan been under serious threats in China over the past decades Island, and in the Xijiang River Basin in Guangxi Prov- because of the changes in farming systems, economic ince. These findings and collections had significantly development, rapid urbanization and other human dis- enriched the knowledge of Chinese wild Oryza species

中国科技论文在线 http://www.paper.edu.cn

COMMON WILD RICE BIODIVERSITY 85

spikelet

stigma

anther

awn

Fig. 1 A photo showing part of an opening panicle of Oryza ruﬁ- pogon.

After 1949, the exploration and collection of wild Oryza germplasm in China was further emphasized and came to a new era. In the 1960s, many Chinese scientists from various agricultural research stations and universities organized several expedition trips, which covered larger areas of the wild Oryza species’ natural range. Three wild Oryza species, O. rufipogon, Oryza granulata and Oryza officinalis, were found and collected in many more sites in south China. In addition, the weedy type of wild rice identified as Oryza sativa f. spontanea was also collected in many sites, particularly in the areas where O. rufipogon and O. sativa coexisted (Wu 1990). This revealed the abun- dance of the wild Oryza species in the southern part of China at that time and the frequent introgression between O. rufipogon and O. sativa in the areas where the two species coexisted. However, all activities for the exploration and collection of wild Oryza species were mostly limited to some parts in Guangdong, Guangxi, Yunnan and Hainan Prov- inces before the 1970s. In order to understand the general distribution patterns of the wild Oryza species in China and to collect genetic resources of wild Oryza species more Fig. 2 Geographical distribution and in situ conservation sites of extensively, a nation-wide project to collect and explore Oryza rufipogon in China. (...) Wild rice distribution area; () in- wild Oryza species was conducted in China during 1978– situ conservation site. 1982. The Chinese Academy of Agricultural Sciences, provincial academies of agricultural sciences and agricultural and drew attention to its study and utilization. Many of research stations or extension services stations were these wild Oryza collections have been kept in the major actively and cooperatively involved in the surveys. The herbaria of China (e.g. Kunmin Botanical Institute and exploration teams surveyed and examined the natural Beijing Botanical Institute; Wu 1990; Yin 1993). habitats, distribution patterns, morphological characters

中国科技论文在线 http://www.paper.edu.cn

86 Z. SONG ET AL.

and genetic variation of the wild Oryza species and their assays, adding to that at a morphological level. Cai et al. biotypes, as well as their responses to insect and disease (1996) have found that Chinese CWR populations hold attacks. As a consequence, a great number of tillers of wild relatively richer allozyme variability than those from rice, seed samples (3238) and herbarium specimens (790) Thailand, Indonesia and India. Their study also indicates were collected. Plants originating from the collected tillers that CWR populations from Guangxi present higher and seed samples were characterized under the experi- genetic variation than those from Jiangxi and Yunnan mental conditions based on standard descriptors. Several Provinces. Gao (1997) has more extensively examined the exploration and collection missions at different scales genetic diversity of Chinese wild rices using the allozyme were undertaken in the late 1980s and early 1990s. method, covering three Oryza species—O. rufipogon, O. officinalis and O. granulata. His findings show that the CWR populations from Guangxi and Guangdong have Genetic variation in Chinese populations of higher genetic diversity than those from other regions of common wild rice China, whereas Yunnan populations have lower genetic Based on surveys on wild rice resources, Chinese scien- diversity. Lower genetic diversity and less differentiation tists have a sound understanding of the genetic variability as well as a high deficiency of heterozygotes exist in the of CWR in China at different levels involving morpholog- marginal populations compared with the central popula- ical, cytological, allozyme and DNA levels. In general, tions, and also lie in small populations compared with CWR was found to have accumulated abundant genetic large populations (Gao et al. 2000a,b, 2001, 2002a,b). Ge variation during the long-term evolutionary process. The et al. (1999) have compared RAPD variation of Chinese first comprehensive document recording CWR diversity CWR with that from Brazil, indicating that there is greater is the book titled, List of Chinese Rice Resources (Institute of genetic diversity in the Chinese CWR populations. The Crop Germplasm Resources of the Chinese Academy of results of SSR assays by Song et al. (2003a, 2004a), Zhou Agricultural Science 1991), which included 3733 acces- et al. (2003) and Gao (2004) all reveal high genetic diver- sions of CWR collected across China. The distribution and sity in Chinese CWR populations, which vary among dif- botanical characters of each CWR accession, including ferent regions—the Guangdong and Guangxi populations sampling locality; growth form; flowering period; stem have significantly higher variation than others in China and leaf color; and characteristics of the spikelet, awn, (Table 1). stigma, anther, grain and seed, were carefully described In general, high genetic diversity is found within Chi- in the book, which is the first-hand resource for research- nese populations with relatively low genetic differentia- ing, utilizing and conserving genetic diversity of Chinese tion. South China is a putative center of genetic diversity CWR. The second edition of this book was published in in China. A gradual reduction of genetic variability occurs 1998, and recorded nearly 6000 accessions of Chinese from its center to the margins and in small populations CWR (Pang & Chen 2002), indicating rich wild rice with a decline in population numbers and area in this resources in China. species (Gao et al. 2000c). The CWR populations intro- With the rapid development of the molecular marker gressed with cultivated rice present relatively higher method, the genetic variation and structure of CWR pop- genetic diversity than the well-isolated populations. ulations were gradually revealed by allozymes, restriction These studies suggest that the abundant genetic diversity fragment length polymorphism, random amplified poly- in Chinese CWR might be caused by the following factors: morphic DNA (RAPD) and simple sequence repeat (SSR) (i) China is a putative genetic diversity center of CWR; (ii)

Table 1 A comparison of SSR diversity of common wild rice between different provinces in China based on the results of Gao (2004), Song et al. (2003a, 2004a) and Zhou et al. (2003)

Ae He Province Gao Song et al. Zhou et al. Gao Song et al. Zhou et al.

Fujian 2.796 ND ND 0.627 ND ND Guangdong 4.040 2.773 4.000 0.708 0.609 0.488 Guangxi 3.289 ND 4.100 0.639 ND 0.502 Hainan 3.170 ND 2.930 0.627 ND 0.358 Hunan 2.826 1.915 ND 0.628 0.467 ND Jiangxi 2.107 2.086 1.957 0.493 0.471 0.274 Yunnan 2.991 ND ND 0.547 ND ND Mean 3.407 2.269 3.350 0.640 0.480 0.413

Ae, number of effective alleles; He, expected heterozygosity; ND, no data obtained.

中国科技论文在线 http://www.paper.edu.cn

COMMON WILD RICE BIODIVERSITY 87

diversity of habitats and distribution in its natural range suggest that this species will naturally become extinct, in China; and (iii) introgression of cultivated rice (Wang which means that human disturbance might have driven & Sun 1996; Song et al. 2003a, 2004a). CWR to become endangered or extinct. The impact of human activities on CWR mainly include land transfor- mation, rice production, grazing and mowing, which Threats to common wild rice and its habitats have led to habitat destruction, or at best, fragmentation. in China Most of the localities of CWR have been transformed to Despite its abundant genetic diversity and relatively wide cultivated rice fields, fish ponds, residential areas, facto- distribution in China, Chinese CWR populations are ries, railroads and highways, with the rapid growth of a declining so rapidly that it was listed in the Chinese Red market economy in China since ‘the reform and opening Data Book of Plant Species: Rare and Endangering Plants (Fu up policy’ began in the 1980s (Gao 2003). For example, the & Jin 1992). It has been confirmed that CWR (O. rufipogon) Nan-Kun railroad project (Nanning City to Kunming has disappeared in the only two localities in the Taiwan City) resulted in the decline in the CWR population size Province, Xinzhu and Taoyun (Kiang et al. 1979), in which and drove several populations to extinction (Pang & Chen it had been previously found. Based on Gao’s (1997, 2003) 1998). The single population in Fujian Province located in data, drastic genetic erosion is occurring in China’s CWR Zhangpu County was buried completely by the Zhang- populations. For instance, the species used to be widely Shan highway (Zhangzhou City to Shantou City). Two distributed in Jinhong County (Yunnan Province), where CWR populations distributed in Zhongshan County at least 26 populations were identified. Unfortunately, (Guangxi Chuang Municipality) disappeared in 2003, only three of them still exist, and the surviving popula- resulting from fish cultivation (Z. Song, unpubl. obs., tions are declining very rapidly (Gao et al. 2001). Similarly, 2003). The expansion of rice fields to CWR habitats over 80% of CWR populations have disappeared and the resulted in the disappearance of six subpopulations in remaining populations are also under serious threat of Dongxiang County (Jiangxi Province; Song et al. 2003a). extinction in Guangdong and Hainan Provinces, where Habitat destruction and decrease in quality both affect CWR was found at 1182 sites of varying size. The largest the development and stability of CWR populations, even CWR population in China, located in Maliaotang of Guix- though human activities did not directly drive CWR to ian County (Guangxi Province), is also on the verge of extinction, for example, irrigating projects, fish cultivation extinction (Gao et al. 1998). In Dongxiang County (Jiangxi or planting lotus as a vegetable in wetlands where CWR Province), the northernmost boundary of the species, six once grew. These activities all lead to the fragmentation of the nine recorded subpopulations have recently disap- of CWR habitats that further split large populations into peared (Gao et al. 2001; Song et al. 2003a). The survey by small populations. It is well known that small populations Pang & Chen (1998) has shown a general decline in the usually have a number of ecological and genetic conse- size of Guangxi CWR populations, several of which have quences, for example, relatively strong inbreeding depres- become extinct. Our recent field surveys conducted from sion and genetic drift that both quicken the process of 1999 to 2004 in south China also show that CWR is reach- genetic erosion. A number of studies have shown that ing a serious endangered status. For example, the single genetic diversity level is correlated to CWR population population located in Zhangpu County (Fujian Province) size, that is, large populations usually have higher genetic has disappeared and two CWR populations distributed in variability than small populations (Xie et al. 2001; Song Zhongshan County (Guangxi Chuang Municipality) dis- et al. 2003a). For instance, the genetic diversity of the appeared in 2003 (Z. Song, unpubl. obs., 2003). Fortu- Dongxiang populations in Jiangxi Province decreased nately, wide concerns about the threatened status of CWR gradually with greater deviation from the Hardy–Weiberg have recently been increasing in China, which has greatly expectation and high deficiency of heterozygotes in the promoted CWR conservation and related research. period from 1980 to 1994 (Gao 1997; Gao et al. 2001). The same trend has also been found in CWR populations in other countries. For example, Akimoto et al. (1999) have Causes of the endangered status of common observed that habitat fragmentation during the period of wild rice 1984–1995 resulted in a rapid decline of CWR genetic There are two major reasons leading to the loss of biodi- diversity in Thailand, and concomitantly an increase in versity—the natural processes of biological extinction and introgression from cultivated rice in the investigated human disturbance. CWR has adapted to its natural envi- populations. ronment through evolution over millions of years. As an Mowing and/or grazing are a frequent phenomenon in important rice germplasm, CWR has been studied from areas where CWR is distributed. These activities have the gene, individual, population and community levels. been revealed to seriously reduce sexual reproductive These studies have shown that there is no evidence to success and recruitment of seed banks, as well as the

88 Z. SONG ET AL.

competitive ability of CWR, consequently leading to pop- servation center for all crop genetic resources in China, ulation eclipse (Zhou 1995). which has a capacity of more than 400 000 accessions. Up In addition, the invasions of cultivated rice into CWR to the end of 1997, a total of 310 000 accessions of germ- habitats, including growing near CWR, play important plasm resources had been stored in the gene bank. To roles in the decline in CWR populations indirectly by gene date, NCGB preserves more than 5000 CWR samples, introgression, even though these activities do not directly over 90% of which were collected in China. The Gene- lead to habitat destruction or loss. Introgression between Bank of Chinese National Rice Research Institute (Hang- cultivated rice and CWR has been suggested to occur zhou), and gene banks and/or seed banks of provincial frequently when the two species grow within the same and local agricultural agencies are also keeping a large vicinity because CWR has mixed mating systems with a number of CWR samples. In addition, two National Wild relatively high outcrossing rate and high outcrossing Rice Nurseries in the 1980s were established in Guang- compatibility and a considerable rate of gene exchange zhou City in Guangdong Province and Nanning City in with cultivated rice under natural conditions. It is Guangxi Province in China. They keep more than 8000 believed that hybrids of the two species have the potential accessions of CWR samples cultivated as a living stock. to persist under natural conditions (Oka & Morishima 1967; Oka 1988; Morishima & Barbier 1990; Morishima In situ conservation of common wild rice in China et al. 1998; Gao et al. 2000a; Song et al. 2002, 2003a,b, 2004a,b; Lu et al. 2003). Introgression may also lead to a The central and provincial governments have also paid change in CWR population genetic diversity and struc- increasing attention to the conservation of wild rice spe- ture as mentioned above. In extreme cases, introgression cies. For example, several in situ conservation sites have even causes the extinction of local CWR populations, as been established as it is widely realized that wild rice observed in Taiwan (Kiang et al. 1979; Morishima 1998). relatives are important genetic resources, but their habi- Therefore, isolation with sufficient distance from culti- tats—aquatics and wetlands—are very fragile and threat- vated rice fields should be set aside in the actions of in ened. The earliest in situ conservation site for CWR in situ conservation of CWR (Song et al. 2004c). China is Dongxiang in Jiangxi Province, which was established in 1986 by the Rice Research Institute of the Jiangxi Academy of Agricultural Science. The second is in Chal- Conservation actions of common wild rice ing in Hunan Province, which was built in 1994 by the in China Chaling County and Wuhan University. Subsequently, the Depending on the objectives and scopes of the activity, Zengcheng CWR in situ conservation site was set up by there are two basic approaches to germplasm conserva- the Zengcheng County (Guangdong Province) in 1996. tion—ex situ and in situ conservation. Ex situ conservation The first national action relating to CWR in situ conserva- is an approach that entails the actual removal of genetic tion was initiated in 2001 by the Ministry of Agriculture resources (seeds, pollen, sperm, individual organisms) of the People’s Republic of China. This action included from the original habitats or natural environments. How- five in situ conservation sites that were distributed in dif- ever, in relation to evolution, ex situ conservation is static; ferent provinces and used different methods of conserva- thus, it may reduce the adaptive potential of the wild tion (Table 2; Zhang & Yang 2004). To date, a total of 10 in species and their populations in the future because seed situ conservation sites for CWR have been established in samples are isolated from the environments where micro- China: Dongxiang (Jiangxi Province), Chaling (Hunan evolution occurred (Bellon et al. 1998; Morishima 1998). In Province), Zhangpu (Fujian Province), Yuanjian (Yunnan situ conservation is a method that is used to preserve the integrity of genetic resources by conserving them within Table 2 Background of in situ conservation sites of common wild the evolutionary dynamic ecosystems of their original rice established in 2001 by the Ministry of Agriculture of the habitats or natural environments. In contrast to the ex situ People’s Republic of China (adapted from Zhang & Yang 2004) approach, this method is intended for preserving wild species or populations in a dynamic way using a continu- Conserved ing evolutionary process (Morishima 1998; Lu 1999). Site area (ha) Method applied Dongxiang, Jiangxi 6.70 Wall Ex situ conservation of common wild rice in China Pengshan in Gaozhou, 42.10 Fence Guangdong The national and provincial governments of China have Heya in Gaozhou, 22.60 Fence made more effort in the ex situ conservation of wild ger- Guangdong Yuanjiang, Yunan 0.08 Plant fence mplasm, including wild rices. The National Crop Gene Wuxuan, Guangxi 16.70 Farmer’s participatory Bank (NCGB, Beijing), built in 1986, is the long-term con-

COMMON WILD RICE BIODIVERSITY 89

Province), Gaozhou (two sites), Zhengcheng (Guangdong Province), Yulin (Guangxi Chuang Municipality), Wuxuan (Guangxi Chuang Municipality) and Qionghai (Hainan Province; Fig. 2). More in situ conservation sites for Chinese CWR are planned to be set up in cooperation with national and provincial conservation programs in the future.

Case studies of common wild rice conservation in China 1996 1997 1998 With a fuller understanding of the roles of CWR in rice improvement and the endangered status of wild rice species, conservation of wild rice has been attracting more and more attention. In particular, the extinction of CWR in Taoyuan in Taiwan Province called for urgent attention towards the species in China (Kiang et al. 1979). In order to restore this population, Oka reintroduced CWR to Taoyun using seeds and tillers that were initially collected N from Taoyun and preserved in the Gene-bank of Japan, 0 5 10m and monitored the establishment and dynamics of the reintroduced population (Oka 1991, 1992a,b). This was 1999 2000 2001 the first study designed to restore the lost populations of CWR, although Taoyun’s CWR finally disappeared, Fig. 3 Reintroduction sites of the Oryza rufipogon population in which might have been caused by competitive exclusion 1993 and the dynamics of spatial distribution during 1997–2001 by a co-occurring plant species Leersia hexandra. in Huli Marsh, Chaling County (Hunan Province), China (Liu Another relatively thorough conservation study has et al. 2003). been conducted for Chaling CWR populations in Hunan Province. The Chaling CWR populations are distributed (2003a) applied the SSR method to estimate the genetic in the Huli wetland with an area of about 2 ha. As a result variation of the standing CWR populations in Dongxiang of the changes in the water level, some of the Chaling and Chaling, and revealed that the reintroduced popula- O. rufipogon populations were already extinct before 1993. tions in Chaling hold approximately 90% of the genetic Zhou (1995) reintroduced O. rufipogon using samples col- variation of the original populations. These results show lected in 1983 and preserved as clonal strains in the wild that the attempts to conserve the Chaling populations is rice conservation plots of Hunan Academy of Agricultural effective, and the experiences from this CWR site can be Sciences in Changsha City to restore the extinct popula- potentially applicable to other conservation programs tion (Zhou 1995). Since then, a detailed investigation of practiced in other sites of CWR in China. population dynamics and community structure has been Li et al. (1999) reported an interesting attempt to con- carried out for Chaling CWR, which has lasted for serve CWR. In the program, he constituted an artificial 10 years. During this period, Chaling CWR was well CWR community in Yujiang County (Jiangxi Province) restored; and its populations expanded rapidly. The num- with specific reference to the original structure and spe- ber of O. rufipogon individuals increased steadily from cies composition of the CWR community that previously 1993 to 2001. The CWR patch number, patch area, mean existed in Doxiang County. A 4-year investigation of an patch size and the largest patch size increased over this ex situ community in an artificial wetland has demon- period (Fig. 3). The number of CWR patches increased strated that CWR can regenerate well by both clonal from 27 to more than 1000, most of which have an area of growth and sexual reproduction, but seedling recruit- greater than 4 m2. The total area of CWR populations in ment plays a major role in population increase. CWR Chaling reaches more than 45 000 m2, which is much populations showed a rapid increase, both in number and larger than the initial area of 3000 m2 (Liu et al. 2001, 2003; in distribution area, suggesting that the CWR popula- Z. Song, unpubl. obs., 2003). These results demonstrate tions can adapt well to new habitats, thus ex situ conser- that the measure (i.e. reintroduction) taken for Chaling vation of the endangered species can be successful. CWR populations has been effective for wild population Additionally, control of the water level is suggested to be restoration and in situ conservation. This view is also sup- critical to the maintenance of the ex situ populations (Li ported by the result of our molecular analysis. Song et al. et al. 1999). This approach can be used as an effective

90 Z. SONG ET AL.

alternative to wild rice conservation when its original cultivated rice to CWR occurs at a considerable rate

habitats are destroyed. (Song et al. 2003b). The performance comparison of F1 Based on the assumption that the gene flow of culti- hybrids with their parental species (a cultivated rice vated rice (germplasm introgression) plays a significant variety [O. sativa L.] and CWR) showed that the hybrids role in shaping the genetic structure of Chinese CWR present inferiority at the stage of sexual reproduction, a populations (Wang & Sun 1996; Gao 1997; Ge et al. 1999), slightly higher hybrid vigor at the growth stage and bet- we recently designed a series of experiments to address ter tillering ability, but no significant differences in com- the possibility of introgression from cultivated rice to posite fitness across the whole life history compared CWR. The aspects we examined included the viability with wild parental species, suggesting the possibility of and longevity of rice pollen, the pattern and range of rice persistence in crop-wild rice hybrids under natural con- pollen dispersal, the effects of interspecific pollen compe- ditions (Song et al. 2004b). In addition, the genetic analy- tition on crop-to-wild gene flow, the rate and range of ses of introgressed CWR populations in comparison crop-to-wild gene flow, the performance of crop-wild rice with well-isolated populations suggest that introgressed hybrids, as well as genetic variation of putative intro- populations present relatively higher genetic variation, gressed CWR populations (Song 2001). The study of pol- and that introgression has impacts on the genetic struc- len competition showed that foreign pollen (of cultivated ture of natural CWR populations and population differ- rice) is disadvantageous, even in the absence of pollen entiation (Song et al. 2001a, 2003a; Z. Song et al. unpubl. competition. Although conspecific pollen (of CWR) is data, 2004). Our case studies provide the basic informa- often more successful than foreign pollen, hybridization tion for in situ conservation of CWR. The implication for is still possible following the deposition of pollen mix- in situ conservation is that an effective isolation buffer tures, especially when foreign pollen arrives earlier than zone needs to be established to prevent pollen invasions. conspecific pollen. This suggests that pollen competition Furthermore, we have established an experimental sys- between wild and cultivated rice can slow the rate at tem both for in situ conservation of CWR and for the which crop genes move to wild populations, but even if biosafety assessment of transgene escape for wind- it were ubiquitous it would not prevent gene flow from pollinated crops. the crop (Song et al. 2002). In conclusion, the conserved biodiversity of wild rice Data obtained from pollen traps for six designed pop- germplasm has provided not only valuable resources for ulations (as pollen sources) at different intervals showed rice breeding, but also important materials for scientific that the dispersal of rice pollen decreases with an research. Biodiversity conservation of the wild Oryza spe- increase in the distance from pollen sources, and that cies has become increasingly important for the sustainable rice pollen flow is significantly influenced by weather use of wild Oryza genetic resources for rice improvement conditions, particularly by wind direction and speed. and development as well as for the generation of scientific For a mean wind speed of 2.52 m/s in a downwind information in China. The collection and conservation of direction, the observed distance of rice pollen dispersal germplasms of wild Oryza species is becoming more and is 38.4 m, indicating that rice pollen grains normally dis- more important because wild Oryza species have played perse at a relatively small range. However, the maxi- an increasingly important role in the improvement of rice mum distance of rice pollen flow can be as far as 110 m, varieties, particularly by providing genes for resistance to, using regression analysis of pollen flow and wind speed, or tolerance of, biotic and abiotic stresses. The accelerated when the wind speed reaches 10 m/s in the study by extinction of wild Oryza populations or the significant Song et al. (2004c). The frequency of pollen flow is posi- decline in population size in many countries as a result of tively correlated to pollen source size within a given regional economic development makes the task of conser- range, suggesting that pollen flow occurs effectively at a vation more important and urgent. A well-organized and considerable rate in the rice field with sufficiently large established international cooperative network is therefore pollen sources (Song et al. 2004c). The results of a crucial for the effective conservation of wild Oryza species designed experiment for crop-to-wild gene flow indi- worldwide, particularly under the implementation of the cated that the natural gene flow frequencies from culti- ‘Convention on Biological Diversity’, in which the indi- vated rice to CWR is approximately 3% when they vidual country has the property right to its biodiversity adjacently occur in fields, and that the maximum or genetic resources. More effective and practical mecha- observed distance of gene flow reaches a distance of nisms need to be developed for ex situ conservation of more than 40 m (43.2 m). The occurrence of the crop-to- wild Oryza species. Conservation of the wild Oryza spe- wild gene flow is also significantly associated with wind cies by an in situ approach also deserves more attention directions, and the frequencies of gene flow decrease and input to allow evolutionary dynamics to play its role, significantly with an increase in the distance from and to guarantee the availability of the valuable wild the pollen sources, demonstrating that gene flow from Oryza germplasm.

COMMON WILD RICE BIODIVERSITY 91

Acknowledgments and allozyme loci. Theoretical and Applied Genetics 106: 173– 180. The support of The National Nature Science Foundation Gao L. Z., Ge S. & Hong D. Y. (2002b) Low levels of allozyme of China (grant no. 30300019), Shanghai Commission diversity and conservation genetics of common wild rice of Science and Technology (grant no. 02JC14022 and Oryza rufipogon Griff. from Yunnan, China. Euphytica 124: 03dz19309), and 211 Project (project name: Biodiversity 273–281. and Regional Ecosafety) is kindly acknowledged. Gao L. Z., Wei C. L. & Yang Q. J. (2001) Intra-population genetic structure of Oryza rufipogon Griff. in Yunnan, China. Journal of Plant Research 114: 107–113. References Gao L. Z., Zhou Y., Ge S. & Hong D. Y. (1998) Germplasm resources of common wild rice Oryza rufipogon Griff. and its Akimoto M., Shimamoto Y. & Morishima H. (1999) The extinc- conservation strategies in Guangxi Province, China. Acta Agri- tion of genetic resources of Asian wild rice, Oryza rufipogon cultura Sinica 31: 32–39. (In Chinese with English abstract.) Griff.: A case study in Thailand. Genetics Research and Crop Ge S., Oliveira G. C. X., Schaal B. A., Gao L. Z. & Hong D. Y. Evolution 46: 419–425. (1999) RAPD variation within and between natural popula- Bellon M. R., Brar D. S., Lu B. R. & Pham J. L. (1998) Rice genetic tions of the wild rice Oryza rufipogon from China and Brazil. resources. In: Dwoling N. G., Greenfield S. M. & Fischer K. S. Heredity 82: 638–644. (eds). Sustainability of Rice in the Global Food System, Chapter Institute of Crop Germplasm Resources of the Chinese Academy 16. Pacific Basin Study Center and International Rice of Agricultural Science. ed (1991) List of Chinese Rice Research Institute, Manila, pp. 251–283. Resources. Chinese Agriculture Press, Beijing. Brar D. S., Dalmacio R., Elloran R., Aggarwal R., Angeles R. & Jena K. K. & Khush G. S. (1990) Introgression of genes from O. Khush G. S. (1996) Gene transfer and molecular characteriza- officinalis Well. ex Watt. to cultivated rice, O. sativa. Theoretical tion of introgression from wild Oryza species into rice. In: and Applied Genetics 80: 737–745. Khush G. S. (ed.). Rice Genetics III. International Rice Research Khush G. S. (1977) Disease and insect resistance in rice. Advances Institute, Manila, pp. 477–486. in Agronomics 29: 265–361. Cai H. W., Wang X. K. & Morishima H. (1996) Genetic diversity Khush G. S. (1997) Origin, dispersal, cultivation and variation of of Chinese wild rice populations. In: Wang X. K. & Sun C. Q. rice. Plant Molecular Biology 35: 25–34. (eds). Origin and Differentiation of Chinese Cultivation Rice. Khush G. S., Bacalangco E. & Ogawa T. (1990) A new gene for China Agricultural University Press, Beijing, pp. 154–156. resistance to bacterial blight from O. longistaminata. Rice Chang T. T. (1976) The origin, evolution, cultivation, dissemina- Genetics Newsletter 7: 121–122. tion and diversification of Asian and African rices. Euphytica Kiang Y. T., Antonovics D. & Wu L. (1979) The extinction of wild 25: 425–441. rice (Oryza perennis formasana) in Taiwan. Journal of Asian Fu L.G. & Jin J.M. eds (1992) Chinese Red Data Book of Plant Species: Ecology 1: 1–9. Rare and Endangering Plants. Science Press, Beijing. Li J., Chen J. K., Zhou J. & He G. Q. (1999) Dynamics and eval- Gao L. Z. (1997) A study on genetic variation of three wild rices uation of Oryza rufipogon and Ranalisma rostratum population (Oryza spp.) in China and their conservation biology. PhD in ex situ conservation of communities. Acta Phytoecologica Dissertation. Institute of Botany, Chinese Academy of Sci- Sinica 23: 275–282. (In Chinese with English abstract.) ences, Beijing. (In Chinese with English abstract.) Liu G. H., Wang H. Y., Zhou J. & Guo Y. H. (2001) Spatial Gao L. Z. (2003) The conservation of rice biodiversity in China: distribution and niche analysis of dominant species at the Significance, genetic erosion, ethnobotany and prospect. Oryza rufipogon reserve in Chaling, Hunan. Acta Phytoecologica Genetic Resources and Crop Evolution 50: 17–32. Sinica 25: 178–185. (In Chinese with English abstract.) Gao L. Z. (2004) Population structure and conservation genetics Liu G. H., Zhou J., Huang D. S. & Li W. (2003) Spatial and of wild rice Oryza rufipogon (Poaceae): A region-wide perspec- temporal dynamics of a restored population of Oryza rufi- tive from microsatellite variation. Molecular Ecology 13: 1009– pogon in Huli Marsh, South China. Restoration Ecology 12: 456– 1024. 463. Gao L. Z., Hong D. Y. & Ge S. (2000a) Allozyme variation and Lu B. R. (1996) Diversity of the rice genepool and its sustainable population genetic structure of common wild rice Oryza rufi- utilization. In: Zhang A. & Wu S. (eds). Proceedings of the pogon Griff. in China. Theoretical and Applied Genetics 101: 494– International Symposium on Florestic Characteristics and Diversity 502. of East Asian Plants; Kunming City, China. China Higher Edu- Gao L. Z., Chen W., Jiang W. Z., Ge S., Hong D. Y. & Wang X. cation Press, Beijing and Springer-Verlag, Berlin, pp. 454–460. Q. (2000b) Genetic erosion in northern marginal population Lu B. R. (1998) Diversity of rice genetic resources and its utiliza- of the common wild rice Oryza rufipogon Griff., and its con- tion and conservation. Chinese Biodiversity 6: 63–72. (In Chi- servation, revealed by the change of population genetic struc- nese with English abstract.) ture. Hereditas 133: 47–53. Lu B. R. (1999) Need to wild rice species in Nepal. International Gao L. Z., Ge S., Hong D. Y., Chen W., Jiang W. Z. & Wang X. K. Rice Research Notes 24: 41. (2000c) Genetic erosion in northern marginal population of Lu B. R., Song Z. P. & Chen J. K. (2003) Can transgenic rice cause common wild rice Oryza rufipogon Griff. and its conservation, ecological risks through transgene escape? Progress in Natural revealed by allozyme analysis. Hereditas 133: 47–53. Science 13: 17–24. Gao L. Z., Schaal B. A., Zhang C. H., Jia J. Z. & Dong Y. S. (2002a) Matsuo T. & Hoshikawa K. (1993) Science of the Rice Plant, Vol. 1. Assessment of population genetic structure of common wild Morphology. Food and Agriculture Policy Research Center, rice Oryza rufipogon Griff. detected by microsatellite DNA Tokyo.

92 Z. SONG ET AL.

Morishima H. (1998) Conservation and genetic characterization Song Z. P., Lu B. R. & Chen J. K. (2004c) Pollen flow of cultivated of plant genetic resources. In: Seko H., Vaughan D. A., Okuno rice measured under experimental conditions. Biodiversity and K., Shirata K. & Ebana K. (eds). Plant Genetic Resources: Char- Conservation 13: 579–590. acterization and Evaluation. Research Council Secretariat of the Song Z. P., Xu X., Wang B., Chen J. K. & Lu B. R. (2003a) Genetic Ministry of Agriculture, Forestry and Fisheries and National diversity in the northernmost Oryza rufipogon populations Institute of Agrobiological Resources, Tsukuba Japan, pp. 31– estimated by SSR markers. Theoretical and Applied Genetics 107: 42. 1492–1499. Morishima H. & Barbier P. (1990) Mating system and genetic Song Z. P., Lu B. R., Zhu Y. G. & Chen J. K. (2003b) Gene flow structure of natural populations in wild rice, Oryza rufipogon. from cultivated rice to the wild species Oryza rufipogon under Plant Species Biology 5: 31–39. experimental field conditions. New Phytologist 157: 657–665. Oka H. I. (1988) Origin of Cultivated Rice. Japan Scientific Societies Vaughan D. A. (1994) The Wild Relatives of Rice: A Genetic Press, Tokyo. Resources Handbook. International Rice Research Institute, Oka H. I. (1991) Ecology of wild rice planted in Taiwan: I. Manila. Sequential distribution of species and their interactions in Wang X. K. & Sun C. Q. (1996) Origin and Differentiation of Chinese weed communities. Botany Bulletin of the Academia Sincica 32: Cultivation Rice. China Agricultural University Press, Beijing. 287–293. (In Chinese.) Oka H. I. (1992a) Ecology of wild rice planted in Taiwan: II. Wu M. S. (1990) Proceedings of Wild Rice Resources Researches. Comparison of two populations with different genotypes. China Science and Technology Press, Beijing. (In Chinese.) Botany Bulletin of the Academia Sincica 33: 75–84. Xiao J. H., Grandillo S., Ahn S. N. et al. (1996) Genes from wild Oka H. I. (1992b) Ecology of wild rice planted in Taiwan: III. rice improve yield. Nature 384: 223–224. Differences in regenerating strategies among genetic stocks. Xie Z. W., Lu Y. Q., Ge S., Hong D. Y. & Li F. Z. (2001) Clonality Botany Bulletin of the Academia Sincica 33: 133–140. in wild rice (Oryza rufipogon, Poaceae) and its implications for Oka H. I. & Morishima H. (1967) Variation in the breeding sys- conservation management. American Journal of Botany 88: tems of a wild rice, Oryza perennis. Evolution 21: 249–258. 1058–1064. Pang H. H. & Chen C. B. (1998) Investigation of common wild Ying C. S. (1993) Germplasm Resources in China. Chinese Agricul- rice (Oryza rufipogon Griff.) along the Nan-Kun railroad. Crop tural Technology and Science Press, Beijing. (In Chinese.) Variety and Resources 4: 16–18. (In Chinese with English Yuan L. P. (1993) Advantages of and constraints to the use of abstract.) hybrid rice varieties. In: Wilson K. J. (ed.). International Work- Pang H. H. & Chen C. B. (2002) Wild Rice Resources in China. shop on Apomixis in Rice. Hunan Hybrid Rice Research Center, Guangxi Science and Biotechnology Press, Nanning. (In Changsha, China. Chinese.) Yuan L. P., Virmani S. S. & Mao C. X. (1989) Hybrid rice: Achieve- Song Z. P. (2001) Gene exchange between Oryza rufipogon and ments and further outlook. In: Progress in Irrigated Rice cultivated rice O. sativa under controlled conditions. PhD Dis- Research. International Rice Research Institute, Manila, pp. sertation. Wuhan University, China. (In Chinese with English 219–223. abstract.) Zhang W. X. & Yang Q. W. (2004) Collection, evaluation and Song Z. P., Lu B. R., Zhu Y. G. & Chen J. K. (2001a) Changes of conservation of wild rice resources in China. In: Proceedings genetic structure in common wild rice (Oryza rufipogon) CL- of the First National Conference on Wild Rice in China: Studies populations as referred by RAPD markers. Journal of Genetics and Applications of Wild rice in China. Meteorology Press, and Molecular Biology 12: 78–84. Beijing, pp. 19–25. (In Chinese with English abstract.) Song Z. P., Lu B. R., Zhu Y. G. & Chen J. K. (2002) Pollen com- Zhong D. B., Luo L. J. & Ying C. S. (2000) Advances on transfer- petition between cultivated and wild rice species (Oryza sativa ring elite gene from wild rice species into cultivated rice. and O. rufipogon). New Phytologist 153: 289–296. Chinese Journal of Rice Science 14: 103–106. (In Chinese with Song Z. P., Rong J., Zhu W. & Lu B. R. (2004a) SSR marker detected English abstract.) genetic variations of common wild rice (Oryza rufipogon Griff.) Zhou J. (1995) Studies on conservation biology of three northern in Guangdong Province. In: Proceedings of the First National populations of common wild rice (Oryza rufipogon). PhD Dis- Conference on Wild Rice in China: Studies and Applications of Wild sertation. Wuhan University, China. (In Chinese with English Rice in China; September 2003, Nanchang. Meteorology Press, abstract.) Beijing, pp. 19–25. (In Chinese with English abstract.) Zhou H. F., Xie Z. W. & Ge S. (2003) Microsatellite analysis of Song Z. P., Lu B. R., Wang B. & Chen J. K. (2004b) Fitness esti- genetic diversity and population genetic structure of a wild

mation through performance comparison of F1 hybrids with rice (Oryza ruﬁpogon Griff.) in China. Theoretical and Applied their parental species Oryza ruﬁpogon and O. sativa. Annuals Genetics 107: 332–339. of Botany 93: 311–316.