Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (20)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Agricola
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Related Collections
Right arrow Crop Genetics
Right arrow Soybean
Crop Science 41:1337-1347 (2001)
© 2001 Crop Science Society of America

PLANT GENETIC RESOURCES

Genetic Diversity among Soybean Accessions from Three Countries Measured by RAPDs

Zenglu Lia and Randall L. Nelson*,b

a Dep. of Crop Sciences, 1101 W. Peabody Dr., Univ. of Illinois, Urbana, IL 61801
b USDA-Agricultural Research Service, Soybean/Maize Germplasm, Pathology, and Genetics Research Unit, Dep. of Crop Sciences, 1101 W. Peabody Dr., Univ. of Illinois, Urbana, IL 61801

* Corresponding author (rlnelson{at}uiuc.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Soybean [Glycine max (L.) Merr.] was domesticated in China but has a long history of cultivation on the Korean peninsula and in Japan. All three areas are considered important sources of soybean germplasm. The objectives of this study were to evaluate the genetic variation in soybean within and among China, S. Korea, and Japan by means of 120 accessions from eight Chinese and three S. Korean provinces, and three Japanese districts; and to relate genetic diversity patterns to geographical regions. Genetic relationships were estimated by 115 random amplified polymorphic DNA (RAPD) markers with simple matching coefficients expressed as Euclidean distances. Hierarchical and nonhierarchical cluster analyses as well as principal component analysis were used to define relationships among the genotypes. The results indicate that the mean genetic distance within China is much larger than that within Japan or S. Korea, but smaller than that between China and Japan or S. Korea. Cluster and principal component analyses almost completely separated the accessions from China from those of Japan and S. Korea, but could not distinguish between the accessions from Japan and S. Korea. These results are consistent with previous research using enzymes and morphological data to classify soybean germplasm from Asia. The groups formed by cluster analysis were mainly based on the frequencies of RAPD fragments among accessions and generally reflected the geographical regions of origin. No clear relationship was found between latitude and genetic diversity among accessions from these countries. Although the soybean accessions from Japan and S. Korea originally came from China, these data indicate that current accessions from Japan and S. Korea are genetically very distinct from those from China and more similar to each other.

Abbreviations: AMOVA, Analysis of molecular variance • HHH, Huang Huai Hai region in east central China • MG, maturity group • NE, northeast region in China • PCR, polymerase chain reaction • RAPD, random amplified polymorphic DNA • SMC, simple matching coefficient • UPGMA, unweighted pair group method using arithmetic average


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE USE OF PLANT INTRODUCTIONS for the development of soybean cultivars will be an important approach to create diversity in soybean breeding in the future (Sneller et al., 1997). Selecting useful diversity from the genetic resources available will be an enormous challenge. Knowledge of diversity patterns will allow breeders to better understand the evolutionary relationships among accessions, to sample germplasm in a more systematic fashion, and to develop strategies to incorporate useful diversity in their breeding programs (Bretting and Widrlechner, 1995). Traditionally, evaluating genetic diversity in soybean has been based on the differences in morphological and agronomic traits or pedigree information (Nelson et al., 1987, 1988; Juvik et al., 1989; Perry and McIntosh, 1991; Sneller, 1994; Gizlice et al., 1994; Bernard et al., 1998). Evaluation based on agronomic data is essential in applied soybean breeding; however, individual genotypes of soybean are only well adapted to certain regions, and the phenotypes are highly influenced by many environmental factors. Genotype x environment interactions greatly limit the range of soybean lines that can be directly compared with morphological data. Random amplified polymorphic DNA (RAPD) is a dominant marker that has been used for diversity studies in several crops (Haley et al., 1994; Mackill, 1995; Thompson et al., 1998; Thompson and Nelson 1998a, b). All previous reports showed that RAPDs were useful for the classification of genetic diversity. On the basis of the principal component analysis of RAPD data on 35 soybean lines, Thompson and Nelson (1998a) identified a set of RAPD primers with high polymorphism information content (PIC) scores that would be useful in surveying a broad spectrum of genetic diversity among soybean accessions.

On the basis of the seed protein data and historical, agronomic, and biogeographical literature, Hymowitz and Kaizuma (1981) suggested that the soybeans in S. Korea and Japan originally came from China, but exact routes of dissemination to either S. Korea or Japan are uncertain. Kihara (1969) reported that rice (Oryza sativa L.) was introduced to Japan on Kyushu Island around the 3rd century B.C.E. Some evidence indicates that rice came to Japan from Korea and other data supports a Chinese origin. Kihara (1969) indicated that soybean arrived in Japan about the same time as rice. Soybean is reported to have come to Korea around the 4th or 5th century B.C.E. (Kim, 1993). It is possible that soybean in Japan could have come from either Korea or China. Because soybean is sensitive to the changes in day length and temperature (Carlson and Lersten, 1987), it is postulated that germplasm is more likely to move east-west along similar latitudes than north-south across different latitudes (Hymowitz and Kaizuma, 1981). There is too little data on genetic variation to know whether general genetic relationships are actually related to latitude and the relatedness of primitive landraces at similar latitudes from S. Korea, Japan, and China are totally unknown. Although results have been reported for genetic relationships among some soybean accessions, evaluations have not included detailed origin information for each accession. A systematic analysis of soybean accessions from countries of ancient soybean cultivation would be helpful to understand the extent and distribution of genetic diversity in these countries. The objectives of this study were (i) to evaluate the genetic variation in soybean within and among China, S. Korea, and Japan using 120 accessions from eight Chinese and three S. Korean provinces, and three Japanese districts; and (ii) to relate genetic diversity patterns to geographical regions.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Genetic Materials
One hundred-twenty soybean accessions were selected from the USDA Soybean Germplasm Collection based on maturity group (MG) and origin information (Table 1). Only accessions considered landraces were selected to represent the diversity in each region except two improved breeding lines in MG 0 and 000 from Heilongjiang province in China. Eight Chinese provinces were included with eight accessions from each of Henan, Hebei, Shaanxi, and Ningxia provinces, and 10 accessions from each of Heilongjiang, Shandong, Jiangsu, and Shanxi provinces. Eight accessions were selected from each of three districts of Japan, Kyushu, Kanto, and Tohoku, and from each of three S. Korean provinces, Cheju, Cholla Nam, and Kangwon (Fig. 1). Shandong, Hebei, Henan, Shaanxi, and Shanxi provinces and part of Jiangsu are located in the Huang Huai Hai (HHH) region of east-central China where soybean is generally planted in a double cropping system. Ningxia province is located in northern China and Heilongjiang province is located in the northeast region of China where soybean is spring planted as a single crop. Cheju and Kyushu are islands in S. Korea and Japan, respectively. Because of their isolation, islands are often sources of unique germplasm. Because the HHH region of China has been considered as the center of origin (Hymowitz and Kaizuma, 1981), lines were selected from nearly all HHH provinces. To compare the genetic diversity patterns across countries at different latitudes, accessions were sampled from different areas in China, S. Korea, and Japan that have similar latitudes from 33 to 39°N. Accessions from Heilongjiang province in China, located at 44 to 53°N latitude were included to represent a very diverse region. Most of the accessions included are in the U.S. maturity groups (MG) III through VI, but seven accessions from Shandong, Jiangsu, and Shanxi of China are in MG I and II. All accessions selected from Heilongjiang province are in MG 000 through II.


View this table:
[in this window]
[in a new window]
  		Table 1. Soybean accessions surveyed with selected random amplified polymorphic DNA primers with origins and maturity groups.

 


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1. Geographical regions of China, Japan, and S. Korea from which the 120 selected soybean accessions originated.

 
RAPD Profiling
Leaf samples were taken from 10 greenhouse-grown plants for each accession at the unifoliolate leaf stage. The plant DNA was extracted by a modified protocol from Keim et al. (1988). Total genomic DNA concentration of each line was estimated with a Perkin-Elmer UV/VIS spectrometer (Perkin-Elmer Corporation, Norwalk, CT) and standardized to a uniform concentration (10 ng/µL) for the PCR reactions.

Random amplified polymorphic DNA reactions were prepared on the basis of the protocol described by Williams et al. (1990) and modified by Kresovich et al. (1994) and Thompson and Nelson (1998a). The reactions were performed on a Perkin-Elmer GeneAmp PCR System 9600 or 9700. A total of 35 10-base primers selected by Thompson and Nelson (1998a) from Operon Technologies (Almeda, CA) was used for PCR amplification. Amplified products were electrophoresed on 1% (w/v) agarose gels in 1x TBE buffer at 96 V for approximately 3 h. The gels were stained with ethidium bromide, viewed under ultraviolet light and photographed.

Statistical Analyses
Each amplified fragment was scored as present (1) or absent (0). Fragment size standards were used to assist in scoring the gels and all gels were compared to reference gels produced with the same primer to help ensure standard scoring across gels. To compare the polymorphic ratio across data sets, all polymorphic and monomorphic fragments were included in the analyses. A genetic similarity matrix of genotypes was calculated by means of simple matching coefficients (SMC), Sij = (a + d)/(a + b + c + d), where a = number of fragments in common between genotypes; d = number of fragments absent in both genotypes, and b and c = number of fragments not in common between two genotypes (Sokal and Michener, 1958). Euclidean distances, Dij = (1 - Sij)1/2, were calculated on the basis of the similarity coefficients between two genotypes by means of a program by Mumm and Dudley (1995). The MEANS procedure was used to compute the average genetic distances and standard deviations between and within countries (SAS Institute, 1989a) and the t test was used to test the difference of genetic distances between gene pools. Polymorphism information content (PIC) scores (Anderson et al., 1992) were calculated on the basis of the formula, PICi = 1 - {sum}P2ij (Weir, 1990, p. 124–134) where Pij is the frequency of the jth allele for Fragment i.

The distance matrix was submitted for cluster analysis by the hierarchical cluster analysis methods of unweighted pair group method using arithmetic average (UPGMA) (SAS Institute, 1989a) and Ward's minimum variance method (Ward, 1963). The TREE procedure was used to generate a dendrogram for both UPGMA and Ward's procedures (Thompson et al., 1998). In UPGMA, the distance between two clusters is the average distance between pairs of observations. In Ward's cluster method (Ward, 1963), the distance between two clusters is the analysis of variance (ANOVA) sum of squares between the two clusters summed over cluster members. The nonhierarchical cluster analysis procedure, VARCLUS (SAS Institute, 1989b), was also used in which the original fragment data were submitted as input to compute the covariance matrix. VARCLUS performs the disjoining clustering of variables based on a covariance matrix. The clusters are chosen to maximize the variation accounted for by the first principal component. To account for the most variation by the first eigenvalue, the maximum value for the second eigenvalue was set to 0.70. A principal component analysis (PCA) was done with a correlation matrix in PC-SAS using a distance matrix as input (SAS Institute, 1989b). The scatterplot was generated with only the first two principal components.

By means of the analysis of molecular variance (AMOVA) program (Schneider et al., 1997), components of variance attributable to differences among countries, among regional populations within countries, and among individuals within populations were estimated from the squared Euclidean distance matrix, as specified in AMOVA, calculated from SMC. The significance of variance components associated with the different possible levels of genetic structure was tested by a nonparametric permutation procedure (Excoffier et al., 1992). The number of permutations for significance testing was set at 100 for all analyses. The genetic distances between any two populations were represented by Fst, a value of F statistic analogues computed from AMOVA (Schneider et al., 1997). To demonstrate the relationship between regions, the inter-population distance matrix generated from AMOVA was used as input to perform a cluster analysis with UPGMA procedure (SAS Institute, 1989a).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
RAPD Profiles and Genetic Distances
Thirty-five selected primers yielded a total of 205 fragments among the 120 soybean accessions of which 115 fragments were polymorphic (56%). This ratio was lower than that found in selected G. max and G. soja Siebold & Zucc. lines from four Chinese provinces by Nelson and Li (1998). The size of fragments scored ranged from approximately 400 to 1600 base pairs. The average number of polymorphic fragments per primer was 3.3. Of the 205 RAPD fragments scored, eight fragments (4% of total) were not present in accessions from S. Korea, Japan or both (Table 2). Because fragments unique to a particular country were rare, patterns of divergence found among the gene pools were attributable largely to differences in fragment frequencies. The proportions of soybean accessions with the polymorphic RAPD fragments ranged from 2 to 98% with a mean frequency of 54% (Fig. 2). This even distribution of frequencies is advantageous for diversity studies.


View this table:
[in this window]
[in a new window]
  		Table 2. The number of soybean accessions that contained RAPD fragments that were not present in all countries and the frequencies of these fragments in the accessions from each country.

 


View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Distribution of the frequencies of polymorphic RAPD fragments among 120 soybean accessions from China, Japan, and S. Korea.

 
Polymorphism information content (PIC) scores represent the gene diversity for a specific locus. The higher the PIC score for a locus, the higher the probability that polymorphism will exist between two accessions at that locus. The range of PIC scores in this study was 0.03 to 0.50 with the mean of 0.32 and a standard deviation of 0.14. Because RAPD markers are usually dominant markers, 0.50 will be the highest PIC score for any fragment.

Genetic distances between genotypes ranged from 0.14 to 0.55 with a mean of 0.42. Two pairs of accessions, KO-CN4 and KO-CN6 from Cholla Nam, S. Korea and the accessions CH-HN5 and CN-HB6 from neighboring provinces Hebei and Henan in China, had the minimum genetic distance (0.14). Both pairs had different maturity groups and also differed in several commonly used descriptive traits. The maximum genetic distance (0.55) occurred between the pair of accessions CH-NX3 and KO-CN7 that came from Ningxia, China and Cholla Nam, S. Korea and were in MG III and V. Accessions with relatively similar maturity groups can be genetically very different. Among 200 pairs of accessions with the lowest genetic distances, the minimum distance occurred between lines from Cholla Nam, S. Korea, but only 4 of the 200 lowest distances came from this region. Most of the lowest genetic distances occurred between accessions from Japan and S. Korea. Many of these pairs were from different regions within these countries but some were from different countries. Only a few existed within China or between China and Japan or S. Korea. Among the 200 pairs of accessions with the highest distances, most occurred between the accessions from China and Japan or S. Korea as well as between accessions from different Chinese provinces. The accessions CH-SAX4, CH-NX3, CH-HN4, CH-HB2, and CH-HB3 from Shaanxi, Ningxia, Henan, and Hebei had the largest average genetic distances with other accessions (0.45–0.47). These genetic distances exceeded all of the mean genetic distances between countries.

The accessions from China had the largest within-country mean genetic distance (0.42). This mean was significantly different from the mean genetic distance within both Japan and S. Korea (Table 3). The accessions from Japan and S. Korea had similar but much smaller average genetic distances (0.37 and 0.37) (Table 3). All of the within- and between-country differences were compared with t tests (Table 3). For example, the intersection of the column labeled Japan with the row labeled China is comparing the average genetic distance within China with the average genetic distance within Japan. The average genetic distance between either China and Japan or China and S. Korea was much larger than that between Japan and S. Korea. The intersection of the column with the Japan-Korea heading with the row labeled Japan is comparing the average genetic distance between Japan and Korea with the average genetic distance within Japan. The average genetic distance between Japan and S. Korea was slightly greater than those within Japan and S. Korea, but it was much smaller than the average genetic distance within China (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Genetic distances for soybean germplasm within and among three countries calculated from RAPD fragments and significance test of inter-country genetic distances.

 
Principal Component Analysis
The principal component analysis (PCA) explains the variance-covariance structure through a few linear combinations of the original variables. With few exceptions, the accessions from Japan and S. Korea are completely separated from the Chinese accessions by plotting the first two principal components (Fig. 3). JP-TO3 and JP-KA4 from Japan were mixed with lines from Ningxia, China. KO-CH4 and KO-CN8 from S. Korea were placed very close to two lines from Shaanxi and one line from Heilongjiang, China. CH-HN7 from Henan province was the only Chinese line that was part of the Japan and S. Korea group. Lines from same provinces or regions generally clustered together. The first principal component accounted for 24% of the total variation and the second principal component accounted for 8% of the total variation. Although both components explained less than one third of the total variation, the results of the PCA are generally consistent with those obtained through clustering analyses.



View larger version (13K):
[in this window]
[in a new window]
  		Fig. 3. Scatterplot based on the first two principal components from a principal components analysis of RAPD fragment data demonstrating the genetic relationships among soybean accessions from China, Japan, and S. Korea. Accessions from Japan and S. Korea are identified only by country but accessions from China are identified by specific province. Observations are coded as follows: K = S. Korea; J = Japan; H = Heilongjiang; N = Henan; B = Hebei; A = Shaanxi; I = Shanxi; U = Jiangsu; S = Shandong; and X = Ningxia.

 
Cluster Analyses
The UPGMA procedure defined 14 clusters and two outliers and Ward's procedure separated the accessions into 15 clusters. The VARCLUS procedure defined 14 clusters that accounted for 72% of the total variation (Table 4). On the basis of the data from all three procedures, 27 consensus clusters were defined with two outliers (Table 4). Of the 27 consensus clusters, 17 clusters were consistently defined by all three procedures, nine clusters were determined by two of three procedures, and one cluster was defined with only the VARCLUS procedure. Sixteen clusters and two outliers involved only Chinese accessions of which 11 clusters were consistently defined by all three procedures. Six clusters included both Japanese and S. Korean accessions of which four clusters were consistently determined by the three procedures. One cluster defined by all three procedures consisted of only S. Korean accessions. There were three two-member clusters and one four-member cluster that contained both Chinese and Japanese or S. Korean accessions of which two clusters were consistently defined by all three procedures. There were one and half times more accessions from China than from both Japan and S. Korea, but the Chinese accessions formed more than two and half times more clusters.


View this table:
[in this window]
[in a new window]
  		Table 4. Group designations for 120 soybean accessions from three cluster procedures and assigned clusters based on all data.

 
Among the Chinese accessions, Clusters 1, 6, 7, 15, and 20 included only accessions from the same province (Jiangsu, Ningxia, or Heilongjiang) and Clusters 5, 8, 11, 13, 16, 23, and 27 were dominated by accessions from a single province including Shandong, Hebei, Shanxi, and Henan. All other members of these clusters except for Cluster 16 came from neighboring provinces. Cluster 16 contained six accessions from four provinces in the HHH region and one line from Heilongjiang. Cluster 12 consisted of six members with three each from Hebei and Shaanxi provinces. Clusters 3, 15, and 25 were two- or three-member clusters that included lines from the same region or neighboring provinces in China. CH-HL9 and CH-SX5 from Heilongjiang and Shaanxi provinces were not consistently defined by the three procedures and we classified them as outliers. CH-HL9 was one of the two improved breeding lines included in this study. The pedigree of this line included landraces from Jilin and Liaoning provinces, which are also in the northeast region of China.

Among those clusters containing no Chinese accessions, only Cluster 9 consisted of lines solely from S. Korea. The accessions in Cluster 4 were predominately from Cholla Nam and Cheju of S. Korea and Clusters 10, 18, and 14 were mostly lines from Kanto and Kyushu of Japan and Cheju of S. Korea, respectively. These three clusters as well as Clusters 2 and 24 contain lines from both Japan and S. Korea.

Clusters 19, 21, 22, and 26 contain both Chinese and Japanese or S. Korean accessions. Cluster 19 consistently defined by all three procedures contained one member, KO-KW8, from Kangwon of S. Korea and three members, CH-HN7, CH-JS5, and CH-SD4, from three neighboring provinces in China. Clusters 21, 22, and 26, each contained one line from Japan and one line from China.

AMOVA to Partition Genetic Variance among the Populations
The AMOVA program can estimate and partition total molecular marker variance among countries, among populations (regional groups) and within populations, and test the significance of partitioned variance components using a permutation testing procedure (Excoffier et al., 1992; Schneider et al., 1997). The differences among countries and among populations within countries were both significant but the greatest variation was found among individuals within populations (Table 5).


View this table:
[in this window]
[in a new window]
  		Table 5. Analysis of molecular variance results for the analysis of 120 soybean accessions from China, Japan, and S. Korea.

 
Population pairwise comparisons on the basis of Fst, values can be interpreted as standardized interpopulation distances between regional groups. The population pairwise distances ranged from 0.01 between the Tohoku and Kanto groups in Japan to 0.35 between the Kyushu, Japan and Ningxia, China groups (Table 6). The pairwise comparisons of regions between China and Japan ranged from 0.11 to 0.35. The same comparisons between China and S. Korea were similar with a range from 0.17 to 0.34 (Table 6). The pairwise comparisons between regions of S. Korea and Japan where generally much smaller and ranged from only 0.04 to 0.10. The populations from both Japan and S. Korea were genetically distinct from those of China but much more similar to each other. However, with the exceptions of Cheju vs. Tohoku, Cheju vs. Kyushu, and Kangwon vs. Tohoku, all other interpopulation distances between Japan and S. Korea were significantly different. Within China, all of interpopulation distances were significantly different except those combinations between Henan, Hebei, and Shaanxi provinces. These three provinces share common borders. Within Japan, none of the inter-population distances were significantly different. Although Cheju is an island in S. Korea, the accessions from Cheju were not genetically different from those from Cholla Nam.


View this table:
[in this window]
[in a new window]
  		Table 6. Regional pairwise comparisons based on standardized inter-population genetic distances for soybean accessions from China, Japan, and S. Korea.

 
To visualize the genetic relationship among regional populations clearly, the values of Fst from Table 6 were submitted to hierarchical clustering by UPGMA. The cluster analysis totally separated the Chinese lines from the Japanese and S. Korean lines (Fig. 4). The accessions from Henan, Hebei, and Shaanxi provinces of China were the most closely related and the accessions from Shanxi, Jiangsu, and Shandong also formed a tight cluster. The accessions from Heilongjiang and Ningxia were distinct from the others. These two provinces from different regions of China are the only Chinese provinces represented where soybean is generally a spring-planted crop. Accessions from all three regions of Japan were quite closely related. The accessions from Cholla Nam and Cheju in S. Korea formed a cluster, whereas, the accessions from Kangwon were the most different from those from any other region of S. Korea or Japan (Fig. 4).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 4. Dendrogram derived from the UPGMA procedure using genetic distances generated from the AMOVA program depicting the relationships among soybean populations from regions of China, Japan, and S. Korea. Genetic distances are estimated from RAPD markers.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, 56% of the scorable RAPD fragments were polymorphic. By comparison, in a survey of 38 soybean genotypes Keim et al. (1992) reported that 70% of RFLP probes detected polymorphisms and most loci had only two alleles. Lorenzen et al. (1995) reported that only 44% of screened RFLP loci detected polymorphisms among soybean cultivars, breeding lines, and ancestral lines. Skorupska et al. (1993) showed that 46% of the RFLP loci exhibited polymorphisms among 108 soybean genotypes. Thompson et al. (1998) found 34% of 833 amplified RAPD fragments were polymorphic among 35 soybean ancestors and PIs. The polymorphism ratio is mainly affected by the sequences of primers or probes and types and number of lines being evaluated (Keim et al., 1992). The ratio of polymorphic RAPD fragments observed in this study was relatively higher than previous reports. All of the RAPD primers used in this research were selected for high PIC scores in diverse germplasm (Thompson and Nelson, 1998b) and the soybean accessions were selected to represent the landrace diversity from diverse regions and were not representative of modern cultivars or enhanced germplasm lines that are more likely to have high genetic similarity.

The average PIC score of the primers used in this study was 0.32. Although PIC scores can determine the relative usefulness of probes or primers among accessions, they cannot distinguish between low and high fragment frequencies for the RAPDs. The average frequency of the polymorphic RAPD fragments among the accessions observed in this study was 0.54. Keim et al. (1992) found that polymorphism frequency was 0.55 per probe when only adapted lines were included, and 0.69 per probe when all types of genotypes were included. RAPD markers with a 0.5 average frequency will be the most informative in the study of genetic diversity.

The genetic distances between accessions from China and accessions from Japan or S. Korea were significantly different from those between accessions from Japan and S. Korea. All of the analytical procedures used clustered the accessions from Japan and S. Korea together and separated them from the lines from China. These results agree with the findings of Griffin and Palmer (1995) based on the variability of thirteen isoenzyme loci in 1,005 domesticated soybean accessions from Manchuria (present northeast China)-Siberia, India-South Central Asia, China, Japan, S. Korea, and Southeast Asia. They indicated that groups of G. max from Manchuria-Siberia and China were closely related and accessions from Japan and S. Korea were closely related but were distinct from those from China (Griffin and Palmer, 1995). However, they did not provide detailed origin information of soybean accessions from China, Japan, and S. Korea. The result of our study was also consistent with the report of Perry and McIntosh (1991) who evaluated 2250 soybean accessions from 78 countries based on the 17 morphological traits to determine variations among and within geographical regions. Canonical discriminant analysis and clustering of canonical means put the lines from S. Korea and Japan into one group and Chinese lines into a separate group. They also indicated that the clusters containing Chinese accessions were more diverse than other groups. The consistency of results based on three different classification tools provide strong evidence that accessions from Japan and S. Korea are genetically similar and are distinct from accessions from China.

Many cluster procedures have been reported for molecular marker data analysis (Lubbers et al., 1991; Grabau et al., 1992; Kresovich et al., 1994; Griffin and Palmer, 1995; Gizlice et al., 1996; Thompson et al., 1998) but most conclusions are usually based on one procedure only. If natural groupings exist among soybean accessions, different clustering procedures should provide similar results. In this study, we identified 14 groups with two outliers using the UPGMA procedure, 15 clusters with Ward's procedure, and 14 clusters using VARCLUS. Although none of the results from the individual procedures completely matched the consensus assignments of the clusters, 26 of 27 clusters were defined by at least two procedures. These data provide strong evidence that natural genetic groups exist among these accessions.

The accessions selected in this study came from similar latitudes (33 to 39°N lat) in the three countries except the lines from Heilongjiang province (44 to 53°N lat). There are consistent grouping in all of our analyses, but there is no clear relationship among accessions from similar latitudes. All the accessions from Japan and S. Korea that clustered with the Chinese lines in the consensus clusters also mixed with or were close to the Chinese lines in the PCA scatterplot (Fig. 3). Three accessions from S. Korea (KO-CH4, KO-CH5, and KO-CN8) that were very close to the Chinese lines in the PCA scatterplot but formed an independent cluster (Cluster 9).

The results from this study as well as from Griffin and Palmer (1995) and Perry and McIntosh (1991) indicate that the soybean genepool of Japan and S. Korea was probably derived from relatively few introductions from China. Fragment OPA-20450 occurred in 60% of the accessions from China but was absent from all accessions from Japan or S. Korea. The high genetic similarity between S. Korean and Japanese lines support the theory that they share a common origin. Since the soybean is reported on the Korean Peninsula earlier (Kim, 1993) than in Japan (Kihara, 1969), the soybean in Japan may have originally come from Korea. Within China the groups formed generally reflected the geographical origin of the accessions, but within Japan and S. Korea the clustering of accessions showed little relationship to origin. No cluster was identified that contained only accessions from Japan and only one cluster with only S. Korean accessions was formed.

The degree of genetic variation revealed by RAPDs within China was more extensive than that within Japan and S. Korea. Although the number of individual accessions sampled from each region was small and not fully representative of total available diversity within each country, these data do provide general genetic patterns. Accessions from S. Korea and Japan were sampled from three regions in the range of 33 to 39°N. latitude. The nonsignificance of all interpopulation distances within Japan indicated a high degree of uniformity within this gene pool. Among the three regions of S. Korea, the inter-population distances were significant only for Cheju vs. Kangwon, and Cholla Nam vs. Kangwon (Table 6). Although Cheju is an island approximately 3° latitude south of Cholla Nam, the interpopulation distance between the two regions was not significant. The interpopulation distances within Japan and S. Korea were much less than those between most provinces in China (Table 6).

Identifying a genetic structure within soybean germplasm is useful for establishing strategies for sampling and managing germplasm. On the basis of these results, Japan and S. Korea are secondary sources of soybean germplasm but are distinct from the Chinese gene pool. Crosses between the Chinese and Japanese or the Chinese and S. Korean gene pools could create more genetic variability than crosses between provinces in the Chinese gene pool.

As a center of origin, as many as 23 000 G. max and 6100 G. soja accessions have been collected from 31 and 25 provinces in China, respectively. At least 10 000 G. max and 1000 G. soja lines have been collected in Japan and S. Korea. Genetic patterns obtained from this survey of Chinese, Japanese, and S. Korean germplasm can help soybean breeders make better choices when selecting among the large numbers of accessions available.


   ACKNOWLEDGMENTS
 
Funding for this research was provided in part by the United Soybean Board and by the Illinois Soybean Program Operating Board.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the USDA or the University of Illinois and does not imply its approval to the exclusion of other products or vendors that may also be suitable.

Received for publication June 12, 2000.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
D. M. Nichols, W. Lianzheng, Y. Pei, K. D. Glover, and B. W. Diers
Variability among Chinese Glycine soja and Chinese and North American Soybean Genotypes
Crop Sci., May 31, 2007; 47(3): 1289 - 1298.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
D. L. Hyten, Q. Song, Y. Zhu, I.-Y. Choi, R. L. Nelson, J. M. Costa, J. E. Specht, R. C. Shoemaker, and P. B. Cregan
Impacts of genetic bottlenecks on soybean genome diversity
PNAS, November 7, 2006; 103(45): 16666 - 16671.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
J. Zhang, Y. Lu, R. G. Cantrell, and E. Hughs
Molecular Marker Diversity and Field Performance in Commercial Cotton Cultivars Evaluated in the Southwestern USA
Crop Sci., June 24, 2005; 45(4): 1483 - 1490.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
Y. Chen and R. L. Nelson
Relationship between Origin and Genetic Diversity in Chinese Soybean Germplasm
Crop Sci., June 24, 2005; 45(4): 1645 - 1652.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
Y. Chen and R. L. Nelson
Genetic Variation and Relationships among Cultivated, Wild, and Semiwild Soybean
Crop Sci., January 1, 2004; 44(1): 316 - 325.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (20)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Agricola
Right arrow Articles by Li, Z.
Right arrow Articles by Nelson, R. L.
Related Collections
Right arrow Crop Genetics
Right arrow Soybean

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome