Published online 31 May 2007
Published in Crop Sci 47:1289-1298 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
PLANT GENETIC RESOURCES
Variability among Chinese Glycine soja and Chinese and North American Soybean Genotypes
Devin M. Nicholsa,*,
Wang Lianzhengb,
Yanlong Peic,
Karl D. Gloverd and
Brian W. Diersa
a Dep. of Crop Sciences, Univ. of Illinois, Urbana, IL 61801-4798
b Chinese Academy of Agricultural Sciences, Beijing, China
c Dep. of Pathobiology, Univ. of Guelph, Guelph, ON, Canada
d Plant Science Dep., South Dakota State Univ., Brookings, SD 57007-2141
* Corresponding author (dmnichol{at}uiuc.edu).
 |
ABSTRACT
|
|---|
The narrow genetic base of elite soybean, Glycine max (L.) Merr., germplasm may impede further attempts to improve grain yield and other important agronomic characters. Germplasm collections of wild soybean, Glycine soja Siebold & Zucc., are a source of genetic variability for soybean breeding programs. The objectives of this research were to use genetic markers to characterize diversity among 60 G. soja accessions collected in China and to compare this diversity with 18 U.S. ancestral soybean genotypes, 12 Chinese G. max plant introductions (PIs), and 47 elite soybean lines from the northern USA. These accessions were genotyped with a set of 72 simple sequence repeat markers. The G. soja accessions were found to contain more alleles per locus (17) than the U.S. ancestral genotypes (5.8), the Chinese PIs (5.5), or the elite lines (4.5). Multivariate analyses were able to separate the G. max lines from the G. soja accessions and identify the most diverse subset of G. soja accessions. Multidimensional scaling separated G. soja accessions from high and low latitudes, while Ward's clustering method separated the G. soja accessions into distinct clusters that tended to include accessions from similar geographical regions. These data will be useful to breeders selecting G. soja accessions as parents in a breeding program and for establishing a core collection of G. soja to be used in future research.
Abbreviations: AFLP, amplified fragment length polymorphism • GD, genetic distance • MDS, multidimensional scaling • PCR, polymerase chain reaction • PI, plant introduction • RAPD, randomly amplified polymorphic DNA • SSR, simple sequence repeat.
|
INTRODUCTION
|
|---|
THE NARROW genetic base of modern soybean [Glycine max (L.) Merr.] cultivars in North America has been caused by a limited initial base and several decades of intensive breeding and selection. Gizlice et al. (1994) showed that >85% of the genes present in modern North American soybean cultivars could be traced to a collection of 18 ancestors and their initial progeny. Gizlice et al. (1994) studied only public cultivars released before 1988, but it has since been shown that private cultivars do not differ significantly from public cultivars (Sneller, 1994). Since genetic variability is necessary for genetic progress, this limitation of genetic diversity may impede further advances in soybean breeding unless new sources of genetic variability are introduced into breeding programs.
Several alternative gene pools are potential sources of genetic variability for North American soybean breeding programs. Released cultivars and advanced breeding lines from other soybean-producing regions of the world, including Japan, South Korea, and three distinct regions of China, have been shown to constitute distinct gene pools that differ from the North American pool (Li et al., 2001; Li and Nelson, 2001; Ude et al., 2003). Northeastern, northern, and southern China are geographically separated soybean-producing areas with different cultural practices and separate breeding programs, which is reflected in the diversity among these pools (Cui et al., 2000a, 2000b). Cui et al. (2000a) also showed that the genetic base of Chinese soybean breeding is much larger than that of the USA. The 18 most important ancestors of modern Chinese soybean cultivars constitute only 40% of the genetic base (Cui et al., 2000a), compared with 85% for the 18 most important U.S. ancestors (Gizlice et al., 1994).
Li and Nelson (2001) compared genetic diversity among 120 G. max accessions from China, South Korea, and Japan by using randomly amplified polymorphic DNA (RAPD) markers. Their study showed that Chinese germplasm contained a greater amount of genetic diversity than did South Korean or Japanese germplasm and that the Chinese gene pool was distinct from that of South Korea and Japan. Ude et al. (2003) used amplified fragment length polymorphism (AFLP) markers to study diversity patterns among North American soybean cultivars, North American soybean ancestors, Chinese cultivars, and Japanese cultivars. Their cluster analysis grouped the cultivars according to region of origin, suggesting that each of the regions represented a separate gene pool. Patterns of genetic differentiation demonstrated by all these studies suggest that introducing elite lines from China, Japan, and Korea into North American breeding programs would expand genetic diversity and may result in increased genetic gain.
Wild soybean accessions also have been studied to assess their usefulness for increasing genetic diversity of soybean. Glycine soja Siebold & Zucc., which grows wild throughout East Asia, is the progenitor of domestic soybean (Hymowitz and Bernard, 1991), and G. max and G. soja are generally interfertile. Several studies have shown that there is a much greater amount of genetic diversity within G. soja than within G. max. Maughan et al. (1995) tested 94 G. max and G. soja accessions with five simple sequence repeat (SSR) markers, which could be used for the efficient detection of species-specific polymorphisms. They observed 79 alleles total, with 43 more alleles in G. soja accessions than in G. max accessions. Maughan et al. (1996) later used AFLP markers to determine genetic relationships among 23 G. max and G. soja genotypes. The 15 AFLP markers tested produced 759 fragments among the 23 accessions. Of the 759 fragments identified, 274 were found to be polymorphic, with 37 fragments polymorphic only in the G. max accessions and 147 fragments polymorphic only in the G. soja accessions. Cluster and principal component analyses were able to separate the G. max and G. soja accessions. The G. max accessions clustered more closely than did the G. soja accessions, showing the relatively low genetic diversity present in G. max.
Li and Nelson (2002) studied genetic variation in G. max and G. soja and its geographical patterns by using RAPD markers. Eighty G. max and G. soja accessions from four Chinese provinces were included in the study. Twenty-three more polymorphic fragments were detected in G. soja than in G. max among the 172 polymorphic fragments scored. They reported that genetic distances between G. max and G. soja accessions were nearly double within-species distances. They also found, however, that the maximum genetic distance between an individual G. max and G. soja accession was approximately equal to the maximum genetic distance between two individual G. soja accessions. Cluster and principal component analyses were able to separate the G. max and G. soja accessions; however, no genetic association between G. max and G. soja accessions from the same Chinese province was found.
There are currently 20765 Glycine accessions available to soybean breeders in the USDA National Plant Germplasm System and tens of thousands of additional accessions in other national collections. Due to this enormous amount of germplasm available, it is useful to characterize the diversity within each of the above-mentioned classes. Such characterization allows breeders to more efficiently use the germplasm. It would be especially helpful to identify individual accessions or groups of accessions that are genetically the most divergent. Tanksley and McCouch (1997) hypothesized that the use of genetic profiles rather than physical appearance to select exotic germplasm to include in a breeding program increases the likelihood of finding novel and agronomically useful alleles.
Powell et al. (1996) compared the usefulness of restriction fragment length polymorphism, RAPD, AFLP, and SSR marker systems for germplasm analysis. Of these systems, SSRs were shown to have the highest expected heterozygosity, a measure of information content. Simple sequence repeat markers are also relatively easy to use since they are polymerase chain reaction (PCR)-based markers. The combination of high information content and ease of use makes SSR markers a good choice for germplasm analysis studies. The objectives of this research were to use SSR markers to characterize the diversity among 60 G. soja accessions collected in China and to compare that diversity with the diversity among 18 U.S. ancestral soybean genotypes, 12 Chinese G. max plant introductions (PIs), and 47 elite soybean lines from the northern USA to identify patterns of diversity and to identify divergent G. soja accessions that could be of use in breeding programs.
|
MATERIALS AND METHODS
|
|---|
Plant Material
We chose 96 accessions for SSR-marker testing and multivariate analysis. This included 60 G. soja accessions that were collected in China from an area ranging from 24°31' N to 48°35' N and 105°48' E to 134° E and at elevations ranging from 2.9 to 1400 m above sea level (Table 1). The 60 accessions included in the study were selected to represent the geographical distribution of accessions available in the collection of the Chinese Academy of Agricultural Sciences. In addition, two Chinese G. soja accessions from the USDA National Plant Germplasm System, PI 468398C and PI 522183A, representing the most distantly related pair of G. soja accessions of those studied by Li and Nelson (2002), were also included in the study.
View this table:
[in this window]
[in a new window]
|
Table 1. Glycine max (L.) Merr. and Glycine soja Siebold & Zucc. genotypes included in the multivariate diversity analyses with information on country of origin (G. max) and place of collection (G. soja).
|
|
The remaining accessions were all G. max and were selected to represent North American and Chinese soybean germplasm. The North American soybean germplasm was represented by 18 ancestors and their initial progeny that were determined to make up >85% of the genetic base of cultivated soybean in the USA by Gizlice et al. (1994) (labeled U1–U18 in Table 1). The cultivars HS93-4118 and IA3010 and the breeding lines C1979 and K1454, which have been used as parents during the last 5 yr in the University of Illinois soybean breeding program, were also tested. The Chinese soybean germplasm was represented by 12 G. max PIs from China selected from the USDA germplasm collection (labeled C1–C12 in Table 1). These 12 lines were selected on the basis of geographic origin, covering approximately the same geographical area as the G. soja accessions. Several of these lines were selected because they were determined to be the most important breeding material in their respective regions by Li et al. (2001). These lines included PI430595, PI468408A, PI602992, PI578497A, PI578493, PI561354, and PI578488B. The remaining lines were randomly selected from a list of all available PIs from the desired provinces.
The SSR-marker data collected previously on 47 elite lines and cultivars developed in the public or private sector in North America and used as parents in the University of Illinois soybean breeding program (Diers, unpublished data, 2002) were included in the number of alleles per locus analysis but not in the multivariate analyses. These lines were previously genotyped with the same set of molecular markers used in the study and described below. These data were not included in the multivariate analysis, however, because of the difficulty in aligning the marker fragment sizes with the data we collected from the diverse G. soja and G. max germplasm.
Molecular Marker Analysis
Each of the 96 accessions (Table 1) was genotyped with 72 SSR markers selected to be distributed across the entire soybean genome (Table 2). All of the SSR markers used were ATT trinucleotide repeat markers developed by P.B. Cregan (USDA-ARS, Beltsville, MD). Each of the markers was tested on 10 soybean genotypes and shown to produce only a single product in each, demonstrating that each marker corresponds to a single locus (Cregan et al., 1999). There were at least two marker loci mapping to each of the 20 soybean genetic linkage groups. The markers were labeled with three different fluorescent dyes to allow multiplexing.
View this table:
[in this window]
[in a new window]
|
Table 2. Number of alleles per simple sequence repeat marker locus in each Glycine max (L.) Merr. or Glycine soja Siebold & Zucc. germplasm group and total as calculated by the FSTAT program. Data from University of Illinois crossing block material (UIUC CXB) were from a previously collected data set.
|
|
Leaf tissue was collected in bulk from 10 greenhouse-grown plants of each line. The tissue was frozen at –4°C and lyophilized for 48 h. The DNA extraction was performed by using a version of the hexadecyltrimethylammonium bromide (CTAB) protocol as modified by Kabelka et al. (2006). Polymerase chain reactions were performed according to the conditions described by Cregan and Quigley (1997). The fluorescently labeled PCR products were analyzed with an ABI 377 DNA sequencer (Applied Biosystems, Foster City, CA). Four microliters of each sample were loaded on 4.8% acrylamide/bisacrylamide (20:1), 8 M urea, and 1x TBE gels (25 x 42 cm). The samples were then electrophoresed at a constant 200 W for 150 min.
Seed Weight
Seed weight of the G. soja accessions was estimated by weighing 20 seeds of each accession. Only 20 seeds were weighed because of insufficient seed availability. Seeds of all the accessions were not available from a single environment.
Data Analysis
Genescan Analysis Software (Applied Biosystems, 2000) and ABI PRISM Genotyper Software (Applied Biosystems, 2001) were used to determine the size of each PCR amplification fragment in base pairs. Each unique fragment identified for a given SSR marker was assigned a letter and then scored as present (1) or absent (0) in each accession. The program FSTAT (Goudet, 2001) was used to calculate the number of alleles per locus within each germplasm group and across all germplasm tested. Mean numbers of alleles per locus for each germplasm group were tested for significant differences by using t tests in SAS PROC TTEST (SAS Institute, 1989). The program NTSYSpc (Rohlf, 1992) was used to calculate the 96 x 96 pairwise similarity matrix with Jaccard's coefficient (Jaccard, 1908). This similarity measure was chosen because it does not count 0,0 matches between pairs of genotypes. The 0,0 matches occur when a fragment is absent in both accessions. Since SSR markers have multiple alleles at each locus, counting 0,0 matches would inflate the genetic similarity between individuals. The genetic distance (GD) between each pair of accessions was calculated as one minus Jaccard's similarity measure. The similarity matrix was inputted into PROC CLUSTER in SAS (SAS Institute, 1989) to perform cluster analysis with both Ward's minimum variance method (Ward, 1963) and UPGMA (unweighted paired group method using arithmetic averages) (Sneath and Sokal, 1973; Panchen, 1992) by specifying the WARDS and AVERAGE options. The SAS PROC TREE was used to generate dendrograms for both methods. Two clustering methods were used because there is no consensus as to which method best represents the true genetic relationships among accessions. Ward's and UPGMA are the most commonly used clustering algorithms for germplasm analysis. Ward's method uses the analysis of variance sum of squares summed across cluster members as the distance between clusters (Ward, 1963), while UPGMA uses the average distance between pairs of observations as the distance between clusters. Multidimensional scaling (MDS) was performed with SAS PROC MDS (SAS Institute, 1992) by using the Jaccard's similarity matrix from NTSYSpc as the input. SigmaPlot (Systat Software, 2004) was used to generate two-dimensional scatter plots of the MDS results.
Differences in the mean seed weight of the G. soja accessions, which were grouped with G. max lines in the multivariate analyses, and the mean of all of the G. soja lines studied were tested by using a t test in SAS PROC TTEST (SAS Institute, 1989).
|
RESULTS
|
|---|
The average number of alleles per locus across all of the germplasm tested was 18.6. This assumes that each SSR marker corresponds to a single locus, as demonstrated by Cregan et al. (1999). Within groups, the average number of alleles per locus was 17 in the G. soja collection, 5.8 in the U.S. ancestral genotypes, 5.5 in the Chinese G. max PIs, and 4.5 in the elite parents from the University of Illinois soybean breeding program (Table 2). The t tests showed that the number of alleles per locus found in the G. soja collection was significantly greater than the number in the three G. max groups (P < 0.0001). A t test also showed that the mean of 5.8 alleles per locus found in the 18 U.S. ancestral lines was significantly greater than the 4.5 alleles per locus found in the 47 elite lines used in the University of Illinois soybean breeding program (P < 0.0001). Generally, a continuum of alleles was observed, with each allele three base pairs longer than the previous allele, consistent with the "continuous ladder" trend observed by Maughan et al. (1995). They observed that the alleles of four SSR markers tested formed a continuous ladder, meaning that each allele was only one or two base pairs different in size from the next allele, depending on the length of the core repeat of each marker.
The average GD among the G. soja accessions in this study was 0.92, which is significantly (P < 0.0001) larger than the average GD of 0.84 observed among the G. max lines according to a t test. The largest GD observed between two G. soja accessions was 0.99 between HN7 and H8. This is larger than the GD of 0.95 between PI 522183A and PI 468398C, the most distantly related pair of G. soja accessions among those studied by Li and Nelson (2002). It is also larger than the greatest GD between two G. max lines in this study, which was 0.97 between ‘Ogden’ and ‘Arksoy’. The largest GD observed between a pair of G. max and G. soja accessions was the maximum value of 1. This value was observed in six cases, between Z8 and IA3010, GU1 and K1454, GU1 and PI 567323A, GU3 and PI 567323A, GU3 and ‘Lincoln’, and S4 and Lincoln. The smallest GD between two G. soja accessions was 0.32 between GU1 and GU3, while the smallest GD between two G. max lines was 0.17 between ‘AK(Harrow)’ and ‘Illini’.
Both Ward's (Fig. 1) and UPGMA (Fig. 2) clustering methods mostly separated the G. max and G. soja accessions in the study. Ward's method showed a more distinct separation of G. max and G. soja accessions and greater distance between the main G. soja and G. max clusters than did UPGMA. There were also fewer accessions of G. soja in the G. max clusters with Ward's method than with UPGMA. The difference in distances is due to the fact that the UPGMA method uses the average distance between pairs of observations in two clusters to calculate the distance between clusters and create plots, while in Ward's method the distance between two clusters is the ANOVA sum of squares between the clusters and semipartial correlations are used to create plots. Notably, both methods placed the G. soja accessions F2, L12, S11, HN7, J7, NX1, and S2 in the G. max cluster, and the four elite U.S. genotypes from the University of Illinois crossing block clustered together tightly with the U.S. ancestral lines Lincoln and ‘S100’, which together contributed 25.4% of the genes found in the North American gene pool (Gizlice et al., 1994).
View larger version (19K):
[in this window]
[in a new window]
|
Figure 1. Results of cluster analysis of 96 Glycine max (L.) Merr. and Glycine soja Siebold & Zucc. accessions based on 72 simple sequence repeat marker loci using Jaccard's coefficient and Ward's minimum variance clustering method. Accessions are designated according to their label from Table 1.
|
|
View larger version (36K):
[in this window]
[in a new window]
|
Figure 2. Results of cluster analysis of 96 Glycine max (L.) Merr. and Glycine soja Siebold & Zucc. accessions based on 72 simple sequence repeat marker loci using Jaccard's coefficient and unweighted paired group method using arithmetic averages clustering method. Accessions are designated according to their label from Table 1.
|
|
At a finer resolution, Ward's clustering (Fig. 1) separated the ancestors of northern U.S. cultivars from those of southern U.S. cultivars. The clustering of the ancestors in this study corresponds well with the results of Thompson et al. (1998) and Brown-Guedira et al. (2000), although some rearrangement has occurred. Several clusters of ancestors were consistent across the three studies, including the clustering of ‘Jackson’, Ogden, and ‘Roanoke’; that of Lincoln, Illini, S-100, and AK(Harrow); and that of Arksoy and ‘Ralsoy’. The Chinese G. max lines included in this study were mixed with the U.S. ancestral lines in the cluster analyses. This was expected, as many U.S. ancestral genotypes are from China (Gizlice et al., 1994; Thompson et al., 1998).
Ward's clustering (Fig. 1) also separated the G. soja accessions into four distinct clusters. One small cluster consisted of two accessions from Guizhuo Province in the southern soybean-growing region of China (Cui et al., 2000a), while another small cluster consisted of two accessions from Henan Province and one accession from a nearby region in Shandong Province, both in the northern soybean-growing region of China (Cui et al., 2000a). A third cluster, consisting of 13 accessions, contained five out of seven of the G. soja accessions from Shaanxi Province, which is located in the northern soybean-growing region of China (Cui et al., 2000a). The fourth G. soja cluster contained 37 accessions. This cluster contained accessions from many provinces covering a large geographical area. Although this cluster was diverse, the most similar individuals within the cluster were often from the same province or adjacent provinces.
Multidimensional scaling detected a clear distinction between the G. max and G. soja accessions across Dimension 1 (Fig. 3). In addition, MDS showed a trend toward separating the early-maturity group of G. max accessions from the late-maturity G. max group and the low-latitude G. soja accessions from the high-latitude G. soja along Dimension 2. Consistent with the other analyses, MDS placed several G. soja accessions, including F2, L12, S11, HN7, NX1, and JX5, near the G. max group.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 3. Two-dimension multidimensional scaling scatter plot showing patterns of diversity among the 96 Glycine soja Siebold & Zucc. and Glycine max (L.) Merr. accessions studied based on 72 simple sequence repeat marker loci. Accessions are designated according to their label from Table 1
|
|
A t test of the seed weight data showed that G. soja accessions that grouped with the G. max lines in the above analyses, including F2, HN7, J7, JX5, L12, NX1, S2, and S11, had significantly (P < 0.001) heavier seeds than did the average of the G. soja collection. The average seed weight of accessions in the G. soja collection was 15 mg seed–1 but the average seed weight of the above accessions was 49 mg seed–1. The average seed weight of G. max genotypes included in this study was approximately 150 to 200 mg seed–1.
|
DISCUSSION
|
|---|
There is much more genetic diversity contained in the G. soja collection than there is in either the U.S. or Chinese soybean germplasm based on the average number of SSR alleles per locus in each of these groups. The 60 G. soja lines were selected based on their origin throughout China, while the 18 ancestral G. max lines have been shown to make up 85% of the genetic base of cultivated soybean in the USA (Gizlice et al., 1994). The Chinese G. max accessions included in the study were selected on the basis of less information than was known about the North American ancestral lines, and it is difficult to determine what percentage of alleles present in cultivated Chinese soybean is contained in this sample. The greater number of alleles present in the G. soja accessions than G. max suggests that G. soja is a potential source for new alleles for use in soybean breeding programs. The effective utilization of G. soja collections, including the use of molecular marker data to select divergent accessions with unique and potentially useful alleles, could broaden the genetic base of cultivated soybean in the USA and increase the production potential of the crop.
There also were fewer alleles observed in the elite U.S. lines tested than in the U.S. ancestral lines. There are several possible explanations for this difference. One explanation is that additional generations of breeding and selection have continued to decrease the variability in soybean. A second explanation is that the elite lines tested in this study were all used in the University of Illinois breeding program and were adapted to that growing region, while the 18 ancestral genotypes tested originated from more diverse regions and therefore should represent a more diverse gene pool.
The larger average GD among paired G. soja accessions than found in their G. max counterparts also suggests that there is more diversity in G. soja than in G. max. The largest GD between two G. soja accessions from the group of 60 collected in China was greater than the GD between PI 522183A and PI 468398C, the two most diverse lines reported by Li and Nelson (2002). The average GD observed in our study was larger than that observed in previous studies (Maughan et al., 1996; Thompson et al., 1998; Brown-Guedira et al., 2000; Li and Nelson, 2002). This reflects the high degree of polymorphism at SSR loci, even among G. max lines where the largest pair-wise GD observed was 0.97. The high GDs were expected due to the high expected heterozygosity of SSR markers caused by the unique polymerase slippage mechanism that generates allelic diversity (Powell et al., 1996). The new allele formation rate of soybean SSR markers has been estimated by Diwan and Cregan (1997) to be approximately one new allele per 5000 meioses.
The cluster and multivariate analyses performed on the data were able to clearly separate G. max accessions from G. soja accessions. In all of the analyses, the G. max accessions clustered more tightly than did G. soja accessions. This reinforces the conclusion that there is much more genetic variability within the G. soja collection than within the G. max breeding pools of either the USA or China.
Ward's minimum variance clustering method separated the G. soja accessions into several distinct clusters. These clusters tended to define groups of accessions with similar geographic origins. This trend may prove useful when choosing subsets of accessions from germplasm collections.
The only additional recognizable pattern in the data was that MDS separated early-maturing G. max accessions from late-maturity groups and high-latitude G. soja accessions from low-latitude accessions. Within the high-latitude and low-latitude clusters, there was no evident regional clustering. For example, accessions collected in the same Chinese province did not tend to cluster together. These data suggest that soybean breeders interested in using a diverse subset of G. soja PIs in their programs could use latitude of origin as a criterion for selection.
There were several G. soja accessions that grouped with or near the G. max lines in multiple analyses. These accessions included F2, HN7, J7, JX5, L12, NX1, S2, and S11. Detailed phenotypic data for these accessions are not available, but an evaluation of seed weight shows that most have much heavier seeds than do the other G. soja accessions. Although seeds for all of the accessions were not available from a single environment, the difference in seed size was so great that it is unlikely that growing all of the accessions in a single environment would change the result. It is possible that both heavier seeds and clustering of these accessions with G. max accessions resulted from hybridization between G. soja and G. max. Other researchers have suggested that accessions with phenotypes intermediate between G. soja and G. max may be due to hybridization between the two (Hymowitz, 1970; Broich and Palmer, 1981). Since these accessions more closely resemble G. max than do other G. soja accessions, they may be more agronomically adapted and therefore require less effort to introgress genes from them into soybean cultivars; however, they may also not contain as many unique alleles as would the more diverse accessions.
There also were several G. soja accessions that were clearly separated from all other accessions in at least one analysis. These accessions include GX3, H9, HN1, HN2, J10, JX4, L4, S4, and S12. These most distantly related accessions are the most likely sources of rare alleles or rare combinations of alleles, which could be useful for soybean breeders (Tanksley and McCouch, 1997).
The two G. soja accessions from the USDA collection, PI468398C and PI522183A, clustered with the other G. soja accessions in all analyses. The two accessions were placed in separate clusters in both the Ward's and UPGMA cluster analyses and were placed on opposite sides of the G. soja cluster in the MDS analysis. These results are consistent with those of Li and Nelson (2002), which showed the two accessions to be quite divergent; however, PI468398C and PI522183A were not the most distantly related G. soja pair in this study.
The use of SSR markers in conjunction with multivariate statistical analyses is effective in characterizing relative amounts of genetic variability contained in various germplasm collections and its patterning. These analyses also are valuable tools for identifying the most diverse accessions in germplasm collections. We were successful in achieving our objectives of characterizing the diversity among G. soja accessions and identifying the most divergent accessions from among the collection. This means that while the choice of which G. soja accessions to use in a breeding program will remain difficult for breeders searching for unique alleles to improve agronomic traits in soybean, the use of SSR markers and multivariate analyses will provide some direction.
|
ACKNOWLEDGMENTS
|
|---|
This research was partially funded by the Illinois Soybean Association. We would like to thank the staff of the Illinois Genetic Marker Center for assistance in the generation of the molecular marker data.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication September 21, 2006.
|
REFERENCES
|
|---|
- Applied Biosystems. 2000. GeneScan analysis software. Applied Biosystems, Foster City, CA.
- Applied Biosystems. 2001. ABI PRISM genotyper software version 2.5. Applied Biosystems, Foster City, CA.
- Broich, S.L., and R.G. Palmer. 1981. Evolutionary studies of the soybean: The frequency and distribution of alleles among collections of Glycine max and G. soja of various origin. Euphytica 30:55–64.[CrossRef][ISI]
- Brown-Guedira, G.L., J.A. Thompson, R.L. Nelson, and M.L. Warburton. 2000. Evaluation of genetic diversity of soybean introductions and North American ancestors using RAPD and SSR markers. Crop Sci. 40:815–823.[Abstract/Free Full Text]
- Cregan, P.B., T. Jarvik, A.L. Bush, R.C. Shoemaker, K.G. Lark, A.L. Kahler, N. Kaya, T.T. VanToai, D.G. Lohnes, J. Chung, and J.E. Specht. 1999. An integrated genetic linkage map of the soybean genome. Crop Sci. 39:1464–1490.[Abstract/Free Full Text]
- Cregan, P.B., and C. Quigley. 1997. Simple sequence repeat DNA marker analysis. p. 173–185. In G. Caetano-Anolles and P.M. Gresshoff (ed.) DNA markers: Protocols, applications, and overview. John Wiley & Sons, New York.
- Cui, Z., T.E. Carter, Jr., and J.W. Burton. 2000a. Genetic base of 651 Chinese soybean cultivars released during 1923 to 1995. Crop Sci. 40:1470–1481.[Abstract/Free Full Text]
- Cui, Z., T.E. Carter, Jr., and J.W. Burton. 2000b. Genetic diversity patterns in Chinese soybean cultivars based on coefficient of parentage. Crop Sci. 40:1780–1793.[Abstract/Free Full Text]
- Diwan, N., and P.B. Cregan. 1997. Automated sizing of fluorescent-labeled simple sequence repeat (SSR) markers to assay genetic variation in soybean. Theor. Appl. Genet. 95:723–733.[CrossRef][ISI]
- Gizlice, Z., T.E. Carter, Jr., and J.W. Burton. 1994. Genetic base for North American public soybean cultivars released between 1947 and 1988. Crop Sci. 34:1143–1151.[Abstract/Free Full Text]
- Goudet, J. 2001. FSTAT, a program to estimate and test gene diversities and fixation indices (Version 2.9.3). Available at www2.unil.ch/popgen/softwares/fstat.htm (verified 7 Mar. 2007). Dep. of Ecology and Evolution, Lausanne Univ., Lausanne, Switzerland.
- Hymowitz, T. 1970. On the domestication of the soybean. Econ. Bot. 23:408–421.
- Hymowitz, T., and R.L. Bernard. 1991. Origin of soybean and germplasm introduction and development in North America. p. 147–164. In H.L. Shands and L.E. Wiesner (ed.) Use of plant introduction in cultivar development: Part 1. CSSA Spec. Publ. 17. CSSA, Madison, WI.
- Jaccard, P. 1908. Nouvelles recherches sur la distribution florale. Bull. Soc. Vaudoise Sci. Nat. 44:223–270.
- Kabelka, E.A., S.R. Carlson, and B.W. Diers. 2006. Glycine soja PI 468916 SCN resistance loci's associated effects on soybean seed yield and other agronomic traits. Crop Sci. 46:622–629.[Abstract/Free Full Text]
- Li, Z., and R.L. Nelson. 2001. Genetic diversity among soybean accessions from three countries measured by RAPDs. Crop Sci. 41:1337–1347.[Abstract/Free Full Text]
- Li, Z., and R.L. Nelson. 2002. RAPD marker diversity among cultivated and wild soybean accessions from four Chinese provinces. Crop Sci. 42:1737–1744.[Abstract/Free Full Text]
- Li, Z., L. Qiu, J.A. Thompson, M.M. Welsh, and R.L. Nelson. 2001. Molecular genetic analysis of U.S. and Chinese soybean ancestral lines. Crop Sci. 41:1330–1336.[Abstract/Free Full Text]
- Maughan, P.J., M.A. Saghai Maroof, and G.R. Buss. 1995. Microsatellite and amplified sequence length polymorphisms in cultivated and wild soybean. Genome 38:715–723.[Medline]
- Maughan, P.J., M.A. Saghai Maroof, G.R. Buss, and G.M. Huestis. 1996. Amplified fragment length polymorphism (AFLP) in soybean: Species diversity, inheritance, and near-isogenic line analysis. Theor. Appl. Genet. 93:392–401.[CrossRef][ISI]
- Panchen, A.L. 1992. Classification, evolution, and the nature of biology. Cambridge Univ. Press, Cambridge, UK.
- Powell, W., M. Morgante, C. Andre, M. Hanafey, J. Vogel, S. Tingey, and A. Rafalski. 1996. The comparison of RFLP, RAPD, AFLP, and SSR (microsatellite) markers for germplasm analysis. Mol. Breed. 2:225–238.[CrossRef]
- Rohlf, F.J. 1992. NTSYSpc (numerical taxonomy and multivariate analysis system), Version 1.70. Exeter Software, Setauket, NY.
- SAS Institute. 1989. SAS/STAT user's guide, Version 6. 4th ed. SAS Inst., Cary, NC.
- SAS Institute. 1992. Technical report P-229. SAS Inst., Cary, NC.
- Sneath, P.H.A., and R.R. Sokal. 1973. Numerical taxonomy. W.H. Freeman, San Francisco.
- Sneller, C.H. 1994. Pedigree analysis of elite soybean lines. Crop Sci. 34:1515–1522.[Abstract/Free Full Text]
- Systat Software. 2004. SigmaPlot 9.0. Systat Software, Point Richmond, CA.
- Tanksley, S.D., and S.R. McCouch. 1997. Seed banks and molecular maps: Unlocking genetic potential from the wild. Science 277:1063–1066.[Abstract/Free Full Text]
- Thompson, J.A., R.L. Nelson, and L.O. Vodkin. 1998. Identification of diverse soybean germplasm using RAPD markers. Crop Sci. 38:1348–1355.[Abstract/Free Full Text]
- Ude, G.N., W.J. Kenworthy, J.M. Costa, P.B. Cregan, and J. Alvernaz. 2003. Genetic diversity of soybean cultivars from China, Japan, North America, and North American ancestral lines determined by amplified fragment length polymorphism. Crop Sci. 43:1858–1867.[Abstract/Free Full Text]
- Ward, J.H., Jr. 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58:236–244.[CrossRef][ISI]
TOP
ABSTRACT
INTRODUCTION
