Comparative Studies on Intra-varietal Heterogeneity between Rice Landraces and Improved Varieties

Yunnan province is the Asian cultivated rice primary and diversity center, some rice landraces have been planted continuously for a long term. These landraces are all important research samples for effective conservation and sustainable development of crop germplasm resources. In this study, twenty-four pairs of SSR markers were used to compare the intra-varietal heterogeneity among 3 rice landraces that have been planted in Yuanyang with a long history and 3 improved varieties that have been eliminated in Yuanyang. A total of 103 SSR alleles have been found. The alleles revealed in rice landraces are much more than that in improved varieties and some SSR alleles have disappeared in improved varieties. The results showed that the intra-varietal heterogeneity in rice landraces is richer than that in improved varieties; therefore, rice landraces have high adaptability with lighter selective pressures.


Background
Yunnan province is the origin and diversity center for Asian cultivated rices (Chang, 1976;Zeng et al., 2009). There are remarkably diverse sets of rice landraces with rare traits. For example, some product, such as Xiaoxiugu, Magu, Huanuo etc., are tolerant to cold stress; others have 300 to 500 grains per panicle, such as Sanpangdouqishiluo, gongju 73, Banli1 etc. (Porceddu et al., 1988). Some of them have durable resistance to rice blast disease after continuously planted in same region Xiao et al., 2001). For the past few years, we have found that many of rice landraces have been continuously planted for over hundreds years , such as Baijiaolaojing, Hongjiaolaojing, Yuelianggu and so on.
Instead, most of the genetic improvement or hybrid varieties have weak genetic background. Previous researches have been confirmed the problem in China is more serious. Wei et al (2010) have found that there were large promotions of Aizizhan, Nantehao, Nongken58 and Yinfangzhu by the pedigree of Chinese improved rice varieties. The numbers of allozymes allele in improved varieties were 73.1% and 58.3% loci have lost various number of alleles (Wei et al., 2003). Hua et al (2007) have assayed that the genetic diversity of improved rice has been decreased from 1950s by utilizing 36 SSR primers. The spread of modern crop varieties lead to a concern about species, genetic and ecosystem diversity erosion and decline to pose a grave threat to biodiversity protection (Hector et al., 1999). It is likely that the potential risk of narrow genetic diversity which we have realized may be the tip of the iceberg.
Twenty-four pairs of SSR markers were used to compare the intra-varietal heterogeneity among 3 rice landraces that had been planted in Yuanyang with a long history and 3 improved varieties that had been eliminated in Yuanyang. We aimed to analyze why the rice landraces can be successful in planting in Yuanyang with a long history but improved varieties fail in this paper.

Comparison of two kinds of alleles between rice landraces and improved varieties
The results indicated that there are various alleles among the populations detected by 24 pairs of primers (Table 1). The alleles in rice landraces in Yuanyang are much more than that in improved varieties that have been eliminated in Yuanyang. Some alleles were disappeared in improved varieties. Total 6 alleles were detected by RM280, 6 alleles in Baijiaolaojing, 4 alleles in Yuelianggu, for example, RM280 primers have detected 6 alleles in samples. In the Yuanyang population, 6 alleles, 4 alleles and 5 alleles have been detected in Baijiaolaojing, Yuelianggu and Hongjiaolaojing population, separately, whilst 1 allele, 2 alleles, and 1 allele have been detected in Chujing26, Chujing27, and Hexi22 population, respectively. The results obviously indicated that the alleles were missing in improved varieties. Similar conditions could be found in other primers (Table 1). A total of 103 SSR alleles were revealed, the 73 alleles, 83 alleles, 76 alleles, 29 alleles, 32 alleles, 37 alleles have been found in Baijiaolaojing, Yuelianggu, Hongjiaolaojing, Chujing26, Chujing27, and Hexi22 population, respectively. The alleles in rice landraces were twice or 3 times more than that in improved varieties. The average number of alleles was 4.7 per locus. 2.7 alleles, 3.8 alleles, 3.9 alleles, 0.8 alleles, 1.2 alleles, and 1.7 alleles have been found in Baijiaolaojing, Yuelianggu, Hongjiaolaojing, Chujing26, Chujing27, Hexi22 population, separately, which has indicated the same information ( Table1; Table 2). Meanwhile, our study also showed that the number of alleles was different among the rice landraces. The results indicated that there were various differences in alleles among landraces populations, while the phenomenon did not exist in improved varieties (Table 1).

Comparison of genetic similarity between rice landraces and improved varieties
An UPGMA dendrogram based on the cluster analysis of genetic similarity showed a significant genetic variation among the rice varieties, with the similarity coefficients under 0.40. The similarity coefficients was 0.7005, 0.9731, 0.6950, 0.9593, 0.9996, 0.8989 in Baijiaolaojing, Yuelianggu, Hongjiaolaojing, Chujing26, Chujing27 and Hexi22 population, respectively ( Figure 1). The results obviously indicated that the genetic similarity of improved varieties was very high, while the genetic similarity of rice landraces was low ( Figure 1). The genetic diversity and structure in rice landraces of Yuanyang were relatively complex. The genetic diversity and structure in improved varieties were uniform with rare exceptions and the similarity coefficients were almost 1.0000. The genetic background was various among farmer's household. This situation was a suggestion of the complexity in genetic structure of rice landraces in Yuanyang (Figure 1).

Comparison of genetic differentiation between rice landraces and improved varieties
In terms of being expected heterozygosity, there were great differences between the rice landraces in Yuanyang and eliminated improved varieties, so did the even expected heterozygosity (Table 3). The Nei's coefficient of differentiation was different among Baijiaolaojing, Yuelianggu and Hongjiaolaojing population with great expected heterozygosity. In Baijiaolaojing population there were 34% and 66% genetic differentiation from intro-farmers and interfarmers, separately, in the Hongjiaolaojing population, there were 69% and 31% genetic differentiation comes from intro-farmers and interfarmers, respectively, in the Yuelianggu population, there were 90% and 10% genetic differentiation from intro-farmers and inter-farmers, separately. The Nei's coefficient of differentiation was also different among Chujing26, Chujing27 and Hexi22 population with small expected heterozygosity. In the Chujing26 population, there were 34% and 66% genetic differentiation from intro-farmers and inter-farmers, separately; in the Chujing27 population, there were 96% and 4% genetic differentiation from intro-farmers and inter-farmers, respectively, as well as those of 50% and 50% genetic differentiation in Yuelianggu population from introfarmers and inter-farmers, separately; This situation suggested that the intra-varietal heterogeneity in rice landraces was complex, but in improved varieties was more homozygous.  Given the Nei's coefficient and phylogenetic tree, the results showed that the G ST of rice landraces in Yuanyang was different with low genetic similarity, for improved varieties that had been washed out after promoted short-term in Yuanyang, the Nei's coefficient of differentiation was various, but with higher genetic similarity. This situation suggests that the intra-varietal heterogeneity in rice landraces is complex among farmer household.

Discussions
The genetic resources of rice is rich in Yunnan province, some high-quality rice landraces have long cultivated history Lv et al., 2009). This paper indicated that there were more alleles in rice landraces than that in eliminated improved at most of the 24 SSR loci. Some alleles were disappeared at most of the 24 SSR loci, whilst, the numbers of the alleles was high at a few SSR loci. Thus, a part of the alleles was missing in the major inbred rice varieties based on the SSR data (Hua et al., 2007), on the contrary, compared with the landraces, the genetic diversity of modern rice varieties was not lower (Yang et al., 1998). However, these two conclusions were not contradictory. Selecting the different SSR loci, the results may be contradictory. Our results demonstrated that the genetic diversity of modern rice varieties was lower than rice landraces based on the data of genetic diversity and genetic similarity.
Through the genetic variation comparision, our research indicated that the rice landraces had rich intra-varietal heterogeneity, especially in the farmer household. This character provided an infinite potential for the future utilization. The lack of genetic resources of parental (Yang et al., 2007;He et al., 2006) resulted in little improvement of the production, quality and resistant of hybrid varieties (Zhang et al., 2008). Wild rice was the valuable resources for rice breeding. Although the rice landraces resources are rich all over the world, they could not be efficiently used, especially in rice breeding. The most important reason was that the research was weak. Thus, our primary task was to protect rice landraces as effective as wild rice (Yang et al., 2005(Yang et al., , 2007Li et al., 2007;Wang et al., 2009). More researches needed to be done in order to preserve the alternative valuable resources for parental backbone selection, heterosis utilization and new genetic resources exploitation.

Materials
A total of 30 rice (Oryza sativa L.) accessions were used in this study for the SSR analysis, including 5 accessions of Baijiaolaojing, Yuelianggu, Hongjiaolaojing, Chujing26 and Chujing27, were collected 60 randomly accessions in 2005, respectively, and randomly selected 5 accessions. Selecting 50 seedlings at about the three-leaf stage to extract DNA, and the 1500 DNA samples were store at -20℃.

DNA extraction and PCR assay
DNA samples were extracted from leaf tissues of a single seedling at about the three-leaf stage, using the CTAB method to describe by Song et al (2003). A total number of 24 SSR primer pairs were selected to analyze genetic polymorphisms in the rice varieties (Table 1). The PCR reactions were performed in a Mastercyder Gradiet PCR (Eppendorf 5333) programmed following the description based on . Adenaturation period of 2 min at 94℃ was followed by 36 cycles of 40 s at 94℃, 30 s at 55℃, and 40 s at 72℃, and then 10 min at 72℃ for the final extension. Reactions were carried out in a volume of 20 µL containing 1×buffer，0.2 mmol/L each of dATP, dCTP, dGTP and dTTP, 1 mmol/L of SSR primer, 50 ng of genomic DNA and 1 unit of Taq polymerase (TaKaRa Inc.). The PCR products were separated in 6% polyacrylamide denaturing gels ).

Data analysis
The amplified SSR DNA bands representing different alleles were scored as different genotypes. Banding patterns identified in the rice varieties IR36, IR64, and Nihonbare available at the RiceGenes Database were used as reference materials to help score different alleles in all the rice samples. As a result, the co-dominant SSR banding patterns were scored as AA, BB or CC (for homozygote) and AB or BC (for heterozygote) genotypes corresponding to the alleles identified in the Rice Genes Database. The average number of alleles, effective number of alleles, total number of alleles, number of polymorphic loci, the percentage of polymorphic loci, observed heterozygosity, expected heterozygosity, Shannon's information index and so on between accessions were quantified for genetic diversity assessment. The genetic diversity index was analysis with POPGENE (Yeh et al., 1999). Relationships of the rice varieties were estimated based on the similarity coefficient using the UPGMA clustering method. The UPGMA tree was constructed using the NTsyspc program ver. 20.2a (Rohlf, 1997).

Authors' Contributions
DG collected and analyzed the data, drafted and modified the manuscript; RZM analyzed part of data; YYZ directed the whole project and modified the manuscript. All authors have read and approved the final manuscript.