| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Plant Breeding, Cornell Univ., Ithaca, NY 14853-1901
b RiceTec, Inc., P.O. Box 1305, Alvin, TX 77512
* Corresponding author (SRM4{at}cornell.edu)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: PIC, polymorphism information content • RFLP, restriction fragment length polymorphism • RS, random selection • SAF, shared allele frequency • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In a study of pedigree relationships among 140 U.S. rice accessions, Dilday (1990) concluded that all parental germplasm in public cultivars used widely in the southern USA could be traced back to 22 plant introductions from the early 1900s, and those used in California could be traced to 23 introductions. With the use of DNA profiles, these pedigree relationships can be examined in greater detail, providing insight as to which alleles and portions of the genome have been inherited from different ancestral stocks. DNA markers can also be used to develop fingerprints that can help characterize the genetic uniqueness of each accession in a germplasm collection or in a population. Further, the identity, frequency, and putative origin of specific marker alleles and allele combinations can be clearly described. With more than 80 000 rice germplasm accessions available in national and international collections (Chang, 1984; Jackson, 1997), it is well known that many represent nearly identical samples of the same cultivar, while others embody rare alleles or highly unusual allele combinations, offering breeders and geneticists a rich source of genetic diversity for crop improvement. Effective management and utilization of these resources depends to a large extent on appropriate characterization of the material represented in the collection. Molecular marker characterization can help to define the genetic architecture of germplasm resources (Brown and Kresovich, 1996; Lu et al., 2005) and to identify alleles that are associated with key phenotypic traits.
In this study, DNA marker analysis was used to quantify the allelic diversity of 125 U.S. rice cultivars and to compare that diversity with variation detected in 111 diverse accessions collected from elsewhere around the world. The overall objective was to develop a robust methodology for choosing a subset of accessions that embodied a large proportion of the genetic variation of the entire collection on the basis of shared allele frequencies, the presence of rare alleles, unusual allele frequency patterns, and regionally defined patterns of variation. This methodology may be useful in the development of future core collections.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
The 236 rice accessions were genotyped by means of 173 previously mapped molecular markers, including 60 SSRs and 113 RFLPs. These markers were well distributed on the 12 rice chromosomes, providing coverage of the whole rice genome at approximately 10- to 12-cM intervals on the basis of previously published maps (Causse et al., 1994; Chen et al., 1997; Temnykh et al., 2000, 2002; McCouch et al., 2002).
Sixty-four (56.6%) of the 113 RFLP markers were selected from the anchor marker set which was developed for comparative mapping across cereal species (van Deynze et al., 1998). Southern blots and hybridizations were made as described by McCouch et al. (1988). One of five major restriction enzymes (HindIII, EcoRI, EcoRV, DraI, ScaI, XbaI) was randomly chosen for hybridization. Priority was given to the combination used in previous mapping studies, but as interspecific parents were used to construct the mapping population, alternative probe–enzyme combinations were often more informative for this study. If the first chosen enzyme did not detect allelic diversity, another enzyme was chosen for further analysis until genetic diversity was identified by a specific RFLP probe. Only one probe–enzyme combination was used in the data analysis.
The 60 SSRs used in this study were selected from those reported in Panaud et al. (1996) and Chen et al. (1997). Primer sequences, PCR conditions, and silver staining procedures were as described in Panaud et al. (1996) and Chen et al. (1997). When a unique allele was detected in only a single germplasm accession, PCR was repeated to confirm that the allele was not an artifact.
Molecular weight determination was done for RFLPs by measuring migration distances of marker alleles on X-ray film in relation to known molecular weight standards. For SSRs, migration distances were estimated in relation to a 1-kb ladder that served as a size standard on silver stained gels.
Statistical Analysis
The PIC value, described by Botstein et al. (1980), was used to refer to the relative value of each marker with respect to the amount of polymorphism exhibited. The PIC value was estimated by
 |
For each marker locus, the total number of alleles and allele sizes, plus the least and most frequent alleles were determined. Genetic similarity between any pair of cultivars was calculated on the basis of the number of shared alleles and the shared allele frequency (SAF) across all the marker loci. SAF is estimated as the number of loci at which a pair of cultivars has the same alleles divided by the total number of loci, which is statistically equivalent to Bowcock et al. (1994)'s statistic, Ps, the number of shared alleles summed over loci/(2 x number of loci compared). The more similar the two cultivars, the higher the SAF they have. In a germplasm collection consisting of n cultivars, each cultivar can be compared with the rest of the cultivars (n – 1) so that n – 1 SAFs can be obtained. An average SAF can be then obtained for each cultivar from the n – 1 SAFs. The lower the average SAF a cultivar has, the larger the genetic difference there is between this cultivar and the rest of n – 1 cultivars. Theoretically, SAF is negatively related to the informativeness of the markers.
Rice accessions were clustered into two major groups, which corresponded to the indica and japonica subspecies, by STATISTICA (StatSoft, Inc., Tulsa, OK) software, the unweighted pair-group average (UPGA) linkage rule, and SAFs across 100 RFLP marker loci as similarity indices.
For the efficient construction of a more diverse core collection of rice germplasm, accessions were selected on the basis of two independant criteria: (i) the frequency of unique RFLP and SSR alleles and (ii) SAFs among accessions. First the frequency of unique alleles–accession was calculated. Next, SAFs at RFLP, SSR and both (RFLP + SSR) marker loci were calculated for each cultivar. The objective was to identify a subset of accessions that corresponded to approximately 10% of the collection and represented 
70% of the total number of alleles. Accessions containing one or more unique alleles and low SAFs were evaluated as candidates for the core collection. Subsets of accessions representing 5 to 50% of the U.S. or world collections were selected on the basis of random sampling with replacement to validate this subset criterion. Two hundred permutations per subset were evaluated for number of alleles in each group and this number was compared with the total number of alleles identified in the collection from which the subgroups were sampled.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Number of Alleles
The distribution of RFLP and SSR allele frequencies in the world and U.S. collections is presented in Fig. 1A, 1B
and summarized in Table 1. A total of 274 alleles were detected at the 100 RFLP loci evaluated on 236 rice cultivars. This represented an average of 2.7 alleles per locus, with a range of two to eight alleles detected by a single probe/enzyme combination. Microsatellite markers detected a total of 714 alleles at the 60 loci, representing an average of 11.9 SSR alleles per locus. The number of alleles detected by single markers varied from two (e.g., RM34) to 34 (RM257). Six to 16 alleles were detected at most loci. Over 20 alleles were detected at each of six loci: 34 alleles at RM257 (chromosome 9), 29 at RM204 (chromosome 6), 24 at RM247 (chromosome 12) and RM20A (chromosome 12), 23 at RM259 (chromosome 1) and 21 at RM228 (chromosome 10).
|
Two RFLP alleles were detected exclusively in U.S. germplasm. One was a 25-kb XbaI allele at locus CDO202 on chromosome 5, which was detected in only one cultivar, M7. The other was a 5.5-kb ScaI allele at RZ740 on chromosome 4. This allele was observed in five U.S. cultivars, Caloro, Dawn, Bonnet 73, Bond, and Nato. Pedigree analysis suggested that Bond received this allele from its parent, Dawn. However, these two RFLP alleles could not be traced farther back to any of their shared ancestors. At SSR loci, a total of 30 (4.2%) alleles were found exclusively in the U.S. germplasm. These alleles may be useful as diagnostic markers for rice cultivars bred in the USA.
Allele Frequency
RFLP.
Specific alleles at RFLP marker loci were detected with very different frequencies. A total of 21 unique alleles were detected at 19 RFLP loci (Table 1 and Table 2). Typical examples were the 25-kb allele at CDO202 on chromosome 5, observed in M7 (mentioned above), and a 7-kb allele at RZ141 on chromosome 11, observed in DV85. In both cases, only two alleles were detected at each locus, with one of the alleles found in only one accession and the other in all 235 other accessions. Many uncommon alleles occurred in less than 10% of the accessions in the entire collection, and of the 100 RFLP loci analyzed, 53 harbored at least one of these uncommon alleles. There was no apparent clustering along the chromosome or obvious distribution pattern for these low frequency alleles. Common alleles found in 90% of accessions represent the other side of the mirror. If only two alleles occur at a locus and one is very rare, the other allele must be very common; thus, the same 19 RFLP loci with unique alleles also contain common alleles.
|
PIC Value
The relative informativeness of each marker can be evaluated on the basis of its PIC value. The average PIC value was almost twice as high for SSR (0.66) as for RFLP (0.36) markers for the entire collection (Fig. 1C, 1D). Average PIC values for the 100 RFLP markers were twice as high for the world as for the U.S. collection, with a PIC of 0.40 and a range of 0 (CDO202) to 0.80 (RZ424X) for the world collection, and a PIC of 0.21 and a range of 0 (CDO686, CDO36, RZ141, CDO127A) to 0.68 (RZ424X) for the U.S. collection. Average PIC values for the 60 SSR markers were 0.74 for the world collection, with a range of 0.17 (RM256) to 0.92 (RM204), and 0.50 for the U.S. cultivars, with a range of 0.02 (RM211 and RM256) to 0.88 (RM38). Lower PIC values estimated for the U.S. germplasm reflected the lower level of genetic diversity embodied in these accessions compared to the world collection.
Genotypic Diversity among Rice Accessions
Unique Alleles
Table 1 provides a list of accessions containing unique alleles identified on the basis of the entire collection. Fifteen (6.4%) of the rice accessions had unique alleles for at least one RFLP locus, of which 11 had a unique allele at only one marker locus. A total of 21 unique RFLP alleles were found; four existed in Basmati 122 at loci RZ590, RG213, RZ537, and BCD808, and two unique alleles were found in each of the three accessions, DV85, BJ 1, and Zhai-Ye-Qing 8. Only one U.S. rice cultivar, M7, had a unique RFLP allele (Table 1). Eighty-one (34.3%) rice accessions had unique SSR alleles, of which 48 had a unique allele at only one marker locus. A total of 153 unique SSR alleles were identified. The cultivars with the highest number of unique alleles all belonged to the world collection. Kasalath and SLO17 had unique alleles at seven and six SSR loci, respectively. Jojutla and DV85 each had unique alleles at five loci. Eight of the accessions (13-0926, K-65, Basmati 122, BJ1, Bulu Dalam, BS125, Zhai-Ye-Qing 8, and Wase Aikoku 3) had unique alleles at four loci. In the U.S. collection, unique alleles were found in 20 cultivars, but all contained only one with the exception of Caloro, an early California cultivar released in 1920 (three unique alleles) and Delitus, a Louisiana cultivar developed in 1918 (two unique alleles).
Shared-Allele Frequency (SAF)
Table 1 lists the average SAF for each cultivar at RFLP and SSR marker loci when compared with the remaining 235 cultivars. The most similar accessions shared alleles at all marker loci while the least similar accessions shared alleles at zero to four marker loci. For example, no genetic polymorphism could be detected at any RFLP or SSR locus between Calrose and Calrose 76. These cultivars are isolines, with Calrose 76 representing a variant derived from Calrose via chemical mutagenesis (Rutger et al., 1977). U.S. cultivars M-5, M-301, M-103, S-201, Calrose, Calrose 76, CS-M3, and Calmochi-202 all trace back to a common ancestor, Caloro, and all shared the same panel of alleles at all 100 RFLP loci, confirming their close pedigree relationships. However, SSR markers detected polymorphism at 1.8 to 29.1% loci in these eight cultivars and can be used to distinguish them.
For each cultivar, genetic similarities (SAFs) were averaged over all comparisons. For RFLP markers, these averaged similarities ranged from 42.0% (DV85) to 74.3% (B3812A) for the entire collection, while for SSR markers, these values ranged from 11.8% (Taducan) to 44.6% (Calady 40) (Table 1). When the world and U.S. collections were evaluated separately (data not shown), RFLP SAFs for the world collection ranged from 48.4% (DV85) to 64.5% (Cha-Mi-Li), while SSR SAFs ranged from 13.2% (DV85) to 32.1% (W6154). This analysis demonstrated that DV85 shared the fewest alleles with all other samples in the world collection and can be considered the most genetically distinct sample in this dataset. For the U.S. collection, RFLP SAFs ranged from 38.3 (Della) to 85.2% (B3812A) while SSR SAFs ranged from 8.1% (Rexmont) to 59.2% (Toro), suggesting that Della and Rexmont are the most genetically distinct varieties in this set.
Heterozygosity in Germplasm Accessions
Using both RFLP and SSR markers, heterozygosity was detected in 120 (50.8%) of the 236 rice accessions and the number of heterozygous loci detected in a single rice accession ranged from 0 to 39 (24.4%) (Table 1). These heterozygous allele patterns could indicate either seed mixtures or true heterozygosity remaining in these cultivars despite efforts to purify all accessions before starting this work. Seventy-four (31.4%) accessions were found to be heterozygous for at least one of the RFLP loci while 83 (35.2%) of the 236 rice accessions were found to be heterozygous for at least one SSR locus. The number of heterozygous loci detected in a single rice accession ranged from 0 to 22 (22%) for RFLP and from 0 to 17 (28.3%) for SSR markers. The numbers of heterozygous RFLP and SSR loci were highly correlated (r = 0.76**, n = 120). Twelve rice accessions were found to be heterozygous at more than 10% of SSR loci, while only two of them (Early Wataribune and Cha-Mi-Li) also were found to be heterozygous at over 10% of RFLP loci. In general, traditional rice cultivars had a higher level of heterozygosity, as previously reported by Olufowote et al. (1997). The seven most heterozygous–heterogeneous accessions detected by both RFLP and SSR markers included B3812A, Bellemont, Bulu Dalam, Early Wataribune, Cha-Mi-Li, Er-Bai-Li, and Tian-Luo-Huang, the first two of which are from the U.S. collection and the last three are traditional Chinese rice cultivars. Older varieties have previously been shown to be more heterogeneous (Olufowote et al., 1997) than most modern counterparts and thus the purified samples used in this study represent a minimum estimate of diversity present in the original accessions.
Missing and Underrepresented Alleles in U.S. Cultivars
To identify distinctive alleles, allele combinations and allele frequency patterns that could be considered characteristic of U.S. cultivars, we compared the frequencies of alleles at all loci between the U.S. and the world collections. Loci that showed the greatest changes in allele frequency between the world and U.S. collections were identified as either missing or underrepresented alleles (Table 2).
Missing Alleles.
Seven alleles at two RFLP and five SSR loci were represented at frequencies of over 20% (20.4–33.6%) in the world collection, but were entirely lacking in the U.S. cultivars (hereafter termed missing alleles). On the contrary, there was no allele that was frequent in the U.S. collection but was lacking in the world collection, although some unique alleles were found in the U.S. cultivars. For the seven alleles that were lacking in the U.S. collection, six were observed in 62 rice cultivars of the world collection, most of which were tropical indicas such as those from IRRI (IR8, IR20, IR24, IR36, and IR64) and China (Zhai-ye-Qing 8, Zao-Er-Lu 14, W6154, and Wen-Xuan-Qing). Two varieties, Barkat (from India) and IR19661-364-1-2-3 (from IRRI), together contain all seven of these alleles. While indica cultivars are generally not well adapted to the USA, cultivars such as IR8, TN1, SLO17, and H-4, together containing six of these alleles, had been used as parents in U.S. breeding programs (Dilday, 1990). This suggests that negative selection limited the transmission of those alleles into the U.S. gene pool. A similar phenomenon is observed with respect to cv. LA110, a modern U.S. cultivar intended for use as brewers' rice (McIlrath et al., 1979), which was a selection from a cross between two Asian parents (TN1 and H-4). These parents contained two of the alleles mentioned above, a 6-kb allele from H-4 at BCD808 and a 131bp-allele from TN1 at RM207, but neither was transmitted to LA110.
Underrepresented alleles are those that occur in very low frequency (
5%) in the U.S. collection but are frequent (20%) in the world collection. On the basis of these criteria, the alleles that were underrepresented in the U.S. collection included 64 alleles from 62 marker loci (35 RFLP and 27 SSR) distributed on all rice chromosomes, with two to nine loci on each chromosome (Table 2). Eight alleles, from two RFLP and six SSR loci, were found to be frequent in the U.S. collection (22.6–46.6%) but underrepresented in the world collection with allele frequencies less than 5%. Six of these alleles were from the loci at which other alleles were lacking or underrepresented in the U.S. collection.
To address the possibility that differences in allele frequencies between the U.S. and world collections may be associated with the subspecies identity of the germplasm in question, the world collection was divided into two groups, the "world japonica group" consisting of 44 japonica varieties, and the "world indica group" consisting of 67 indica varieties. These groupings were in good agreement with previously published reports using isozymes (Glaszmann, 1987), RFLPs (Wang and Tanksley, 1989), and SSRs (Ni et al., 2002; A. Garris, Cornell University, pers. communication), except that the three varieties, Caloro, Della, and Rexmont, that retained most of the alleles underrepresented in the U.S. collection, were classified into the indica group.
When allele frequencies at the 35 RFLP and 29 SSR loci that were underrepresented in the U.S. collection were compared between the U.S. collection and each of the two subspecies groups within the world collection, a highly positive correlation was observed between the individuals in the U.S. collection and the 44 japonica cultivars in the world japonica group (r = 0.89), while a highly negative correlation was observed between the U.S. collection and the world indica group (r = –0.6). This result highlights the deep population structure in O. sativa such that significant differences in allele frequencies were detected between the two subspecies at 43% of the genetic loci evaluated in this study. It also illustrates the clear japonica identity of U.S. rice germplasm.
Within the U.S. collection, rare alleles (occurring in <5% of accessions) were also detected, and 39 cultivars were found to contain at least one rare allele. Most of these alleles were concentrated in two to four U.S. cultivars. For example, a 2.0-kb allele in EcoRV digested DNA at RZ500 on chromosome 10 was identified in four cultivars (Della, Rexmont, Palmyra, and A301), while a 123-bp allele at the microsatellite locus, RM205 on chromosome 9, was found in two of the same cultivars (Della and Rexmont). Considering all 64 underrepresented alleles, Della alone retained 46 (71.9%) of them, and Rexmont retained 44 (68.8%). Because 33 of these alleles were shared by both Della and Rexmont, a total of 57 (89.1%) were harbored by these two cultivars alone. Three other cultivars hosting relatively large numbers of low frequency alleles included Caloro (21 alleles), LA110 (20), and Palmyra (15). If shared alleles are not counted twice, Della, Rexmont, and Caloro contained 61 (95.3%) of the 64 underrepresented alleles. This means that six well chosen cultivars would hold all underrepresented alleles.
Construction of a Diverse Collection of Rice Germplasm
A subset of rice accessions was selected on the basis of SAFs and number of alleles represented that embodied most of the genetic diversity identified in the 236 accessions. Because RFLP markers had fewer alleles per locus than SSR markers, the SAFs at RFLP loci were much higher than those at SSR loci.
Using SAF to select the top 30% (71) of genetically diverse cultivars, we were able to represent over 95% of RFLP alleles existing in the entire (world and U.S.) collection (Table 1; Subset A). A similar proportion of cultivars was required to represent 95% of RFLP alleles for the world collection. However, only the top 10% of U.S. rice accessions were needed to represent 95% of the RFLP alleles in that collection (Fig. 2A) . To represent the same proportion of SSR alleles, over 50% of genetically diverse cultivars had to be selected from each of the varietal collections (2B), although the number of alleles detected increased more gradually for the world collection.
|
To minimize the size of the collection, 56 rice accessions that shared over 50% alleles at RFLP loci or over 20% alleles at SSR loci, or that had unique alleles at fewer than three loci were excluded from the 87-cultivar set, resulting in a new subgroup with 31 rice accessions (13.1%) (Table 1; Subset B). This new group represented 94.9% (260/274) of RFLP alleles and 74.4% (531/714) of SSR alleles and included two U.S. cultivars, Caloro and Della. To obtain the same level of representativeness, more than 71 (30%) of rice accessions would have to be selected if the selection were based on the total SAF alone. Therefore, to construct a core collection of rice germplasm resources using as few accessions as possible but representing as much genetic diversity as possible, the cultivars having the highest percentage of unique alleles and the lowest SAF should be selected. Using these criteria, a subcollection of 31 diverse O. sativa germplasm was identified. When the same proportions (5–30%) of cultivars were randomly sampled, the number of marker alleles detected also increased with the number of cultivars sampled. However, compared with selection based on shared allele frequencies, randomly chosen cultivars embodied a lower level of allelic diversity as illustrated in Fig. 2A, 2B.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Construction of a Subcollection of Diverse Germplasm: Possibilities and Challenges
Several different methods have been used to construct core collections, aiming to represent most of the genetic diversity with the fewest number of accessions possible (Crossa et al., 1995; Hamon et al., 1995; Schoen and Brown, 1995; van Hintum, 1995). On the basis of phenotypic evaluation of economically important traits and the use of both isozymes and DNA markers, studies of genetic diversity aimed at developing core collections have been reported for several plant species including peanut (Arachis hypogaea L.) (Holbrook et al., 1993), annual Medicago species (Diwan et al., 1994), common bean (Phaseolus vulgaris L.) (Tohme et al., 1995; Miklas et al., 1999; Zeven et al., 1999), potato (Solanum tuberosum L.) (Huaman et al., 2000), and barley (Hordeum vulgare L.) (Knupffer and van Hintum, 1995; Liu et al., 1999). The objective of these studies was to select approximately 10% of the germplasm accessions to represent at least 70% of the genetic variability (e.g., Brown, 1989a, 1989b). In this study, we used RFLP and SSR markers to identify a collection based on calculations of SAF and the number of unique alleles. We determined that 13% of the 236 rice accessions (31 accessions) could be selected to represent 95% of the RFLP alleles and 74% of the SSR alleles. This resource may serve as a source of unique alleles for genetic studies and for broadening the genetic base of U.S. rice cultivars. While many undesirable traits are undoubtedly associated with the use of exotic cultivars, the use of molecular markers to selectively introgress valuable transgressive genetic variation from diverse crosses offers a powerful approach for applications in rice improvement (Tanksley and McCouch, 1997; Xiao et al., 1998; Moncada et al., 2001; Brondani et al., 2002; Nguyen et al., 2003; Thomson et al., 2003; Septiningsih et al., 2003a, 2003b).
Construction of core collections based on richness of marker alleles provides a powerful way to obtain high allele representativeness from a large number of germplasm accessions. Marker allele-based selection alone would theoretically exclude most commercial cultivars from core collections if they are considered along with exotic and/or diverse germplasm accessions, because commercial cultivars usually have a relatively narrow genetic basis and most of the existing alleles can be traced to, and represented by, their ancestral accessions. To construct more representative core collections for breeding purposes, alleles and allele combinations at genetic loci known to be associated with traits of agronomic importance, should be taken into account. As more information becomes available about genes underlying agronomic traits, core collections can be developed using gene and allele specific markers, rather than neutral markers, to assemble collections worthy of preservation.
Potential Applications of the Released Data on Germplasm and Molecular Markers
Rice geneticists may use the data released in this study as a guide in choosing crosses for genetic studies. For instance, it provides preliminary polymorphism data for 236 x 235/2 = 27730 possible cross combinations, including thousands of indica/indica, indica/japonica, and japonica/japonica combinations, expanding on previous studies by Ni et al. (2002); Coburn et al. (2002); Cho et al. (2000) and Blair et al. (2002). As gene structure-function relationships are clarified with greater precision, it will be increasingly possible to focus attention on functionally relevant genetic diversity, i.e., polymorphisms detected within the active sites of structural genes or within key promoter regions. This will make it productive to screen large germplasm collections for functional nucleotide polymorphisms, targeting the search for alleles that are both phenotypically relevant and have high breeding value. Along with such targeted approaches, studies using neutral markers will continue to provide geneticists and breeders with important information about the evolution, breeding history and overall architecture of crop genetic resources.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication April 25, 2003.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Related articles in Crop Science:
This article has been cited by other articles:
|
|
 |
|
  K. K. Jena and D. J. Mackill Molecular Markers and Their Use in Marker-Assisted Selection in Rice Crop Sci., July 1, 2008; 48(4): 1266 - 1276. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. R. Kottapalli, M. D. Burow, G. Burow, J. Burke, and N. Puppala Molecular Characterization of the U.S. Peanut Mini Core Collection Using Microsatellite Markers Crop Sci., July 30, 2007; 47(4): 1718 - 1727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. E. Giarrocco, M. A. Marassi, and G. L. Salerno Assessment of the Genetic Diversity in Argentine Rice Cultivars with SSR Markers Crop Sci., March 1, 2007; 47(2): 853 - 858. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
