|
|
||||||||
Agriculture and Agri-Food Canada, Saskatoon Research Centre, 107 Science Place, Saskatoon, SK, S7H 0X2
* Corresponding author (coulmanb{at}em.agr.ca)
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
There are three crested wheatgrass species of commercial importance in North American rangelands. Fairway crested wheatgrass (A. cristatum) occurs in diploid, tetraploid, and hexaploid form. Standard crested wheatgrass (A. desertorum) and Siberian crested wheatgrass (A. fragile) are both tetraploid.
The relationship between A. cristatum and A. desertorum has been subject to extensive discussion. Knowles (1955) considered A. desertorum to be either an autoploid of A. cristatum, or an amphiploid of A. cristatum and some other closely related diploid. A number of papers written in the 1960s (Dewey, 1961; Tai and Dewey, 1966; Dewey and Pendse, 1967, 1969; Dewey, 1969) concluded, on the basis of meiotic chromosome pairing, that A. desertorum was an autotetraploid of A. cristatum, or a segmental allotetraploid. Schulz-Schaeffer et al. (1963) considered Dewey's (1961) interpretation of polyhaploid chromosome pairing to be incorrect and that A. desertorum is a segmental allotetraploid, and not an autotetraploid. Dewey and Pendse (1967) stated that all crested wheatgrasses had the same basic genome which they called the "C" genome, and this was subsequently changed to the "P" genome.
Taylor and McCoy (1973), using chromatographic and karyotypic analysis, concluded that A. desertorum originated from amphidiploids between Agropyron imbricatum (M. B.) Roem et Schult. and Agropyron pectiniforme Roem et Schult. Both A. imbricatum and A. pectiniforme are very similar morphologically to each other and to A. cristatum, differentiated by level of pubescence on, and spacing between the spikelets (Nevskii, 1934), and both have been reclassified as subspecies of A. cristatum (Tzvelev, 1976).
Dewey (1981) reported on the existence of a second diploid species, Agropyron mongolicum Keng., which is indigenous to China. Agropyron mongolicum is differentiated from A. cristatum by its narrow, linear spikes. All other diploid accessions that had been collected until then had broad spikes, which is characteristic of A. cristatum. Hsiao et al. (1986) reported that A. cristatum and A. mongolicum have similar genome length and differ only by small structural rearrangements on some chromosomes. Hsiao et al. (1989) noted that interspecific hybrids of A. cristatum and A. mongolicum had a similar spike morphology to A. desertorum. Asay et al. (1992) concluded, through a multivariate morphological analysis, that A. desertorum was the result of a hybrid between A. cristatum and A. mongolicum, followed by a chromosome doubling. It was also postulated that A. fragile was an autoploid of A. mongolicum.
On the basis of a study of genome length in the perennial Triticeae, Vogel et al. (1999) concluded that the genome length of A. desertorum was more consistent with an autoploid of A. cristatum than that of an allotetraploid between A. cristatum and A. mongolicum.
There are variations in ploidy levels within species and interspecific hybrids are common. The cultivars Kirk (tetraploid) and Douglas (hexaploid) are both described as A. cristatum. Hybrids between species of different ploidy levels have been used for genetic improvement of crested wheatgrass (Asay et al., 1986). The cultivar Hycrest was developed from a hybrid between colchicine-doubled A. cristatum and A. desertorum (Asay et al., 1985).
Molecular markers are an efficient method for determining genetic relationships because they are not affected by the environmental or epistatic interactions that may affect morphological traits (Schut et al., 1997). AFLP is a polymerase chain reaction-based fingerprinting technique which relies on the selective amplification of restriction fragments (Vos et al., 1995). It has the ability to detect large numbers of polymorphisms with single primer pairs, without any prior knowledge of the target genome. Results are highly repeatable, as there are very stringent reaction conditions for primer annealing. AFLPs have been used to determine diversity within species such as Vigna angularis (Willd.) (Yee et al., 1999) and Camellia sinensis (L.) O. Kuntze (Paul et al., 1997). In studies involving the phylogeny among Lens spp., results obtained from AFLPs were similar to those for random amplified polymorphic DNA (RAPD), but were of much higher resolution (Sharma et al., 1996).
AMOVA is a statistical analysis, that apportions genetic variance into within- and among-population variance (Excoffier et al., 1992). It functions by converting interindividual distance matrices into an analysis of variance (Huff et al., 1993). This tool allows for the separation of populations even when the majority of genetic variance is expressed within populations, such as in outcrossing species. The phi-value (
st), calculated by AMOVA, is proportional to the amount of the total variation accounted for by the variation between two populations and is an indication of the interpopulation distance between the two populations (Huff, 1997).
The objectives of this study were to: (i) examine the phylogeny of selected populations from the genus Agropyron by means of AFLPs; (ii) examine the partitioning of within and among population variances of selected common crested wheatgrass cultivars; and (iii) compare cluster analyses on the basis of
st and Euclidian distance from population bulks.
| MATERIALS AND METHODS |
|---|
|
|
|---|
|
AFLP Analysis
The AFLP procedure was performed as described by Vos et al. (1995) with the AFLP Analysis System 1 provided by Life Technologies, Burlington, ON, Canada. To summarize, a 250-ng genomic DNA sample was digested by EcoRI and MseI restriction enzymes. Adapters were ligated to the digested DNA with T4 DNA ligase.
The adapter sequences were:
Preamplification reaction included one EcoRI primer (5'-GACTGCGTACCAATTCA) and one MseI primer (5'-GATGAGTCCTGAGTAAC) with one selective nucleotide (underlined), 1.5 mM MgCl2, Taq DNA polymerase and 230 µM of each dNTP. Twenty cycles were performed at 94°C for 30 s, 56°C for 60 s, and then 72°C for 60 s.
Four EcoRI primers (5'-GACTGCGTACCAATTC+ AGG,ACG, AAG, and ACT) were end labeled with [
33P] (2000 ci/mmol) in the presence of T4 DNA kinase. The EcoRI primers were each combined separately with eight MseI primers (5'-GATGAGTCCTGAGTAA+ CAG, CAT, CTC, CTT, CTG, CTA, CGC, CAC) in selective amplification, which included three selective nucleotides. The reaction included Taq DNA polymerase, 1.5 mM MgCl2 and 230 µM of each dNTP. Twenty-three cycles were performed at 94°C for 30 s, 56°C for 30 s, and then 72°C for 60 s. After the reaction, formamide dye was added at 0.5 volume to the samples. The samples were then separated by electrophoresis on a 5% (w/v) polyacrylamide gel for 2.5 h at 90 W. The gel was dried onto Wattman 3-mm paper and exposed to Kodak Biomax film at -80°C for 1 to 7 d, depending on signal intensity.
Data Analysis
Primer screening was conducted before genotyping of individuals and bulks. The goal was to select primer combinations that would yield polymorphic banding patterns that could be scored without ambiguity. Of the 32 primer combinations screened, eight were selected to genotype the bulk populations and five were chosen for the analysis of individual plants. From 11 to 29 polymorphic bands were derived from each primer combination. A total of 114 polymorphic bands were used for the bulk population analysis, and 67 for the analysis of individual plants. These polymorphic bands were transformed into two binary matrices, one for bulks and one for individuals.
Population Bulks
Pairwise Euclidean distance matrices between all possible bulked populations were computed by NTSYS-PC 2.01 (Rohlf, 1997). Dendrograms were constructed by the unweighted pair-group method of arithmetic averages (UPGMA) cluster analysis on the basis of the distance matrix. Four dendrograms were constructed with 57, 85, 97, and all 114 markers to compare genetic distances and clustering among various random subsets of the available markers.
AMOVA Analysis
Six populations were used for the AMOVA (Table 1). Euclidean squared distances were estimated between all 90 individuals and genetic variations were estimated by the AMOVA component in the ARLEQUIN 2.000 software (Schneider et al., 2000). The total variance was partitioned into variance between individuals within populations, and variance among populations. Homogeneity of the variances between pairs of populations was tested by the Bartlett's test. Within population mean squares (AMOVA sum of squares divided by n-1) were calculated to examine the different levels of within population variance (Fischer and Matthies, 1998). The phi-statistic (
st), calculated by AMOVA, was used as an interpopulation genetic distance measurement as described by Huff (1997). The UPGMA clustering procedure from NTSYS-PC 2.01 was used to construct a dendrogram on the basis of the interpopulation distances matrix.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
Agropyron mongolicum clustered separately from all other populations. All crested wheatgrass species share the same genome (Dewey and Pendse, 1967), but Hsiao et al. (1986) stated that there were some structural rearrangements between the P genome of A. cristatum and A. mongolicum. In the interests of clarity, in the following discussion, the P genome of A. cristatum will be referred to as "Pc" and the P genome found in A. mongolicum as "Pm".
Nordan, an A. desertorum cultivar, clustered closer to A. cristatum than to A. mongolicum, but the relationship to either species was not close. These molecular data corroborate morphological and cytological evidence that A. desertorum is an alloploid of A. cristatum and A. mongolicum (Hsiao et al., 1986, 1989; Asay et al., 1992). This would indicate that A. desertorum has a genome make up of "Pc Pc Pm Pm".
Douglas, a hexaploid classified as A. cristatum, clustered closest to Nordan. A possible explanation for this is presented by Shultz-Schaeffer et al. (1963), who compared karyotypes from diploid and hexaploid A. cristatum and tetraploid A. desertorum. They found that hexaploids had two genomes which were similar to the one present in A. cristatum, and that the third genome was similar to the second genome of A. desertorum. Using the genome labels indicated above, Douglas would have a genome makeup of "Pc Pc Pc Pc Pm Pm". This would likely make it more closely resemble A. cristatum morphologically (i.e., predominance of Pc), but have an AFLP profile more similar to A. desertorum (i.e., presence of both Pc and Pm). Svitashev et al. (1996) was unable to differentiate hexaploid and tetraploid Elymus L. species with similar genomic constitutions, using repetitive DNA sequences.
Hycrest, CD-II, Kirk and S-8959 clustered somewhat closer to A. cristatum than to A. desertorum. Hycrest and CD-II were the most closely clustered of all populations, likely because CD-II is a synthetic cultivar derived from Hycrest (Asay et al., 1997). The cross between a colchicine-doubled A. cristatum (Pc Pc Pc Pc) with A. desertorum (Pc Pc Pm Pm) would result in progeny with a genome makeup of "Pc Pc Pc Pm". After successive generations this genome makeup could be altered by factors such as degree of homology between the Pc and Pm genomes, and selection. Asay et al. (1986) examined 20 F5 individuals from the hybrid population during meiosis and stated that they were behaving as autotetraploids. The clustering of Hycrest and CD-II with A. cristatum could indicate that the genome of these cultivars is "Pc Pc Pc Pc".
Kirk and S-8959 clustered together. Kirk, classified as A. cristatum, was more distantly grouped to the A. cristatum cluster than were the two hybrid cultivars, Hycrest and CD-II. The original accession from which Kirk was selected was grown in close proximity to A. desertorum accessions, potentially allowing for outcrossing and incorporation of germplasm from the latter (R.P. Knowles, unpublished). S-8959, a population of unknown species, clustered with Kirk, and may have a similar background.
Ephraim grouped loosely with Nordan and Douglas even though it is classified as A. cristatum. Ephraim is the only rhizomatous cultivar represented here, although Fairway individuals will sometimes produce short rhizomes. Tzvelev (1976) describes a species, Agropyron cimmericum, for which the main differentiating feature from A. cristatum is the production of long branched rhizomes. The spike morphology of A. cimmericum is quite similar to the A. cristatum type. The inclusion of a known A. cimmericum population would be required to confirm if Ephraim was a population of this species.
The cultivar Vavilov was included as a representative of A. fragile to determine if it is an autoploid of A. mongolicum, as has previously been suggested (Asay et al., 1992). In the varietal description of Vavilov, it is indicated that outcrossing occurred between the A. fragile population used to form this cultivar and A. desertorum, resulting in a mixed or hybrid population (Asay et al., 1995a). This is demonstrated by the clustering of Vavilov with the Nordan cluster. For more conclusive results on the phylogeny of A. fragile, a pure population would have to be used.
AMOVA
The majority of the AFLP variation (88%) observed among the 90 individuals from the six populations was accounted for by within population variance (Table 2)
. The remaining variation that was attributed to among population variance is relatively small, but is statistically significant (P < 0.001). The higher percentage of within population variance is expected in Agropyron spp. since they are all outcrossing species.
|
st values observed were not random. Individual plants of each line generally clustered together (data not shown) with some overlap with plants of neighboring lines in the dendrogram. This is an indication of the ability of the AMOVA to detect differences between closely related lines, even in the face of high within-population variances. Huff et al. (1993) found similar results using RAPD markers to apportion variance in buffalo grass [Buchloë dactyloides (Nutt.) Engelm.] into among individuals within populations, among populations within regions, and among regions. They were able to differentiate both among populations within regions and among regions, even though there was significant within population and within region variation.
|
|
Within-line mean squares for the tetraploid lines were not greater than those of the diploid lines. Considering that the tetraploids have twice as many chromosomes, it may be expected that there would be a larger number of polymorphisms. In a study conducted by Xu et al. (1991) on tall fescue (Festuca arundinacea Schreb.), it was found that the percentage of polymorphic RFLP bands increased with increasing ploidy level. In contrast, Stammers et al. (1995), in a phylogenic study of Lolium and Festuca spp., found that increased ploidy level and nuclear DNA content did not increase the numbers of RAPD polymorphisms detected. In the present study, the lack of increased polymorphism may be due to the similarity of the genomes making up the tetraploid lines.
Comparison of AMOVA and Bulk Clustering
The dendrogram formed from a cluster analysis of
st values, from the analysis of individual plants within six populations (Fig. 2)
, is similar to a dendrogram of Euclidean distances among the same six populations constructed on the basis of data from population bulks (Fig. 3)
. Although the respective distances between populations are proportionally different, the clustering is identical. This corroborates the data and conclusions which were drawn from the analysis of the population bulks.
|
|
| CONCLUSIONS |
|---|
|
|
|---|
Within-population variance accounted for 88% of the total AFLP variation in the six Agropyron populations assayed, while among-population variance accounted for only 12% (Table 2). There were significant differences between all six populations assayed (Table 3) and different within-population variances may have been due to initial population size and relative selection pressures (Table 4).
The phylogenic relationships drawn from the individual genotype (AMOVA) analysis are the same as those drawn from the bulk analysis. This indicates that bulking strategies may be preferred in similar studies as less effort is required.
The present study examined all of the crested wheatgrass cultivars commercially available in North America; however, for each of the species, there were only one or a few lines. Future studies which could add greater depth to these results would include the examination of a larger sample of Agropyron germplasm, especially more accessions of A. desertorum, A. fragile, and A. mongolicum. It would be of particular interest to include accessions collected from the natural ranges of these species, where they may not have had the opportunity to hybridize. Another area of investigation would be to attempt to find markers which are specific to the Pc and Pm genome and use these markers to examine further the genomic make up of A. desertorum, A. fragile, and hexaploid lines.
| ACKNOWLEDGMENTS |
|---|
Received for publication July 17, 2001.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. J. Ariss and G. J. Vandemark Assessment of Genetic Diversity among Nondormant and Semidormant Alfalfa Populations Using Sequence-Related Amplified Polymorphisms Crop Sci., November 7, 2007; 47(6): 2274 - 2284. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. B. Jensen, S. R. Larson, B. L. Waldron, and K. H. Asay Cytogenetic and Molecular Characterization of Hybrids between 6x, 4x, and 2x Ploidy Levels in Crested Wheatgrass Crop Sci., December 2, 2005; 46(1): 105 - 112. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Rajasekar, S.-h. Fei, and N. E. Christians Analysis of Genetic Diversity in Rough Bluegrass Determined by RAPD Markers Crop Sci., December 2, 2005; 46(1): 162 - 167. [Abstract] [Full Text] [PDF] |
||||
![]() |
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] |
||||
![]() |
Y. S. N. Ferdinandez, B. E. Coulman, and Y.-B. Fu Detecting Genetic Changes over Two Generations of Seed Increase in an Awned Slender Wheatgrass Population Using AFLP Markers Crop Sci., May 6, 2005; 45(3): 1064 - 1068. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y.-B. Fu, A. T. Phan, B. Coulman, and K. W. Richards Genetic Diversity in Natural Populations and Corresponding Seed Collections of Little Bluestem as Revealed by AFLP Markers Crop Sci., November 1, 2004; 44(6): 2254 - 2260. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. S. N. Ferdinandez and B. E. Coulman Genetic Relationships among Smooth Bromegrass Cultivars of Different Ecotypes Detected by AFLP Markers Crop Sci., January 1, 2004; 44(1): 241 - 247. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y.-B. Fu, Y. S. N. Ferdinandez, A. T. Phan, B. Coulman, and K. W. Richards AFLP Variation in Four Blue Grama Seed Sources Crop Sci., January 1, 2004; 44(1): 283 - 288. [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 | |||