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 (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Agricola
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Related Collections
Right arrow Turfgrass Management
Right arrow Turfgrass
Published in Crop Sci. 43:2162-2171 (2003).
© 2003 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

TURFGRASS SCIENCE

AFLP Analyses of Genetic Diversity in Bentgrass

G. V. Vergara and S. S. Bughrara*

Department of Crop and Soil Science, Michigan State University, Room 286 Plant and Soil Sciences Building, East Lansing, MI, USA 48824

* Corresponding author (bughrara{at}msu.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Bentgrasses (Agrostis spp.) are widely occurring temperate grasses with more than 220 species that represent a vast resource for genetic improvement of turfgrass cultivars. Bentgrasses are normally outcrossing species and exhibit many ploidy levels. Difficulties in morphological characterization, which are largely subjected to environmental influences, have resulted in many synonymous species and uncertainties in phylogenetic relationships. To study the genetic diversity and relationships between bentgrass species, 40 accessions from the USDA's germplasm collection representing 14 species of Agrostis from twenty countries were investigated by fluorescence-labeled amplified fragment length polymorphism (AFLP) analysis. Four hundred AFLP markers from five chosen primer combinations were used to differentiate between bentgrass accessions of a bulk of 25 genotypes per accession. Genetic similarities between accessions ranged from 0.62 to 0.98 showing no duplication in the collection and a high level of diversity in Agrostis. Both principal component analysis and Unweighted Pair Group Method with Arithmetic Mean (UPGMA) dendrogram clearly distinguished seven groups. Genetic relationships between diploids and other polyploids were revealed in the cluster groupings.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
BENTGRASS, Agrostis spp., derivation from Greek: grass, forage, is distributed throughout the world and belongs to the Poaceae family (Watson and Dallwitz, 1992). It is a perennial or annual outcrossing polyploid (2n = 14, 28, 42 etc.) and is widely used for putting greens, golf courses, parks, and forage. Some species were used for erosion control and revegetation programs in disturbed areas where grazing, mining, and flooding had occurred (Brown and Johnston, 1979; Cornely et al., 1983; Winterhalder, 1990). Most bentgrass species are widely adapted to temperate climates and occur in a variety of habitats. Variation in Agrostis species denotes potential for genetic improvement of the species for use as turf.

Much of the work in classification of Agrostis is predominantly based on morphological and cytological features (Bjorkman, 1960). Linnaeus grouped Agrostis on the basis of the presence or absence of awns, panicle shape and color, and orientation of culms and roots. The basic chromosome number of Agrostis is x = 7 and differences in ploidy level often determine species boundaries. Diploid species (2n = 2x = 14) are A. alpina Leyss., A. elegans Walt., A. canina L., A. flaccida Hack., and A. nebulosa Boiss. & Reut. Most bentgrass species such as A. capillaris L., A. castellana Boiss. & Reut., A. palustris Huds., and A. stolonifera L. are tetraploids with 2n = 4x = 28 (Brede and Sellman, 2001). Species such as A. castellana, A. gigantea R., A. alba L., A. exarata Trin., A. clavata Trin., A. diegoensis Vasey, and A. hallii Vasey are known to exist in tetraploid and hexaploid forms (2n = 6x = 42). In a wide collection of A. gigantea, Jones (1955a) only found hexaploids. Because of the outcrossing nature of bentgrass, ploidy levels need to be identified. Bonos et al. (1999) used laser flow cytometry as a rapid option to validate the number of chromosomes of certain species of Agrostis. Current significant species in the USA are velvet bentgrass, A. canina; colonial bentgrass, A. capillaris or A. tenuis L.; dryland or highland bentgrass, A. castellana; creeping bentgrass, A. stolonifera; red top bentgrass, A. gigantea; and Idaho bentgrass, A. idahoensis Nash. Creeping bentgrass is the premier turfgrass species for closely mowed golf course putting greens (Funk, 1998).

Intra- and interspecific hybridization is possible in Agrostis; however, very few genetic studies have explained the contribution of diploids to polyploids and their relationship. Hybridization experiments by Davies (1953), using germplasm from UK, showed that hybrids of A. stolonifera x A. gigantea were among the easiest to produce and indicated some degree of homology between the two polyploid species. Bivalent formation during meiosis in hybrids between A. capillaris x A. vinealis Schreb. established A. capillaris to be a segmental allotetraploid and from hybrids of A. stolonifera x A. canina that A. stolonifera was a strict allotetraploid (Jones, 1955a). The use of interspecific hybrids of A. tenuis, A. stolonifera, and A. gigantea and their offspring helped decode the genome constitution in Agrostis. Jones (1955b) suggested that if A. tenuis or A. capillaris was A1A1A2A2 and A. stolonifera was A2A2A3A3, then A. gigantea would be A1A1A2A2A3A3. The A2 genomes of the two species were not confirmed to be absolutely identical. Creeping bentgrass, A. stolonifera was also found to hybridize with ticklegrass, A. scabra Willd. and spike bentgrass, A. exarata (Welsh et al., 1987). A. scabra was found to hybridize with A. trinii Turcz. (Probatova and Kharkevich, 1983). Warnke et al. (1998) used isozyme analysis to study allotetraploid inheritance in creeping bentgrass and found strong genetic evidence for disomic rather than tetrasomic inheritance.

Little is known about bentgrass genetic relationships and there is much confusion regarding assignment to groups and classification. Early effort to use morphological characters in classification resulted in many synonymous species. A. stolonifera was listed as being synonymous with A. alba var. palustris Huds., A. alba var. stolonifera (L.) Sm., A. stolonifera var. compacta Hart., A. stolonifera var. palustris (Huds.) Farw., and A. maritima Lam. (Biota of North America Program, BONAP Poaceae Listing; Plant Gene Resources of Canada (GRIN-CA). Agrostis gigantea is synonymous with three other species (BONAP Poaceae Listing) and may include subspecies on the basis of rhizome and shoot development and the morphology of the spikelet (Dihoru, 1980). Species A. trinii was found synonymous with A. vinealis spp. trinii (Tzvelev, 1971) and A. flacidda spp. trinii (Koyama, 1987), while Soreng et al. (2002) taxonomy listed A. vinealis as being synonymous with A. canina spp. vinealis (Turcz.) Hult. Kurchenko (1979) listed A. trinii and A. canina both from Russia as different species on the basis of morphoanalytical characters and behavior in their natural environments. Differences in phenotypic expressions are oftentimes the result of environmental fluctuations. DNA fingerprinting is considered a more stable and reliable technique to explore genetic diversity and relationships.

A wide variety of DNA marker technologies have been used to differentiate bentgrass cultivars and species such as isozyme (Yamamoto and Duich, 1994; Warnke et al., 1997), random amplified polymorphic DNA (RAPD) (Golembiewski et al., 1997, Scheef et al., 2003) and restriction fragment length polymorphism (RFLP) (Caceres et al., 2000). These techniques are limited by the low levels of polymorphism at the intra- and interspecific levels. A more powerful tool in fingerprinting for biodiversity studies over other PCR based techniques is by amplified fragment length polymorphism (AFLP) analysis (Zabeau and Vos, 1993; Vos et al. 1995). AFLPs have been successfully used to study genetic diversity in crops like rice (Oryza sativa L.) (Zhu et al., 1998), cotton (Gossypium spp.) (Abdalla et al., 2001) and common bean (Phaseolus vulgaris L.) (Tohme et al., 1996). The high frequency of identifiable polymorphism is useful for distinguishing among genotypes, detecting linkages and for mapping loci in turfgrasses. Zhang et al. (1999) used AFLPs to differentiate bermudagrass (Cynodon spp.) genotypes and determine genetic relationships among genotypes. In addition, they showed that the use of fluorescence-based detection of AFLPs has improved both fragment scoring and data handling. Ebina et al. (1999) constructed an AFLP-based genome map of zoysiagrass (Zoysia spp.) and developed a linkage map of QTLs associated with some major traits. AFLPs have not been used to study bentgrass. Understanding bentgrass diversity would facilitate the efficient use of germplasm accessions to combine the favorable agronomic and disease resistance traits to produce superior cultivars.

The objectives of this research were to study the genetic diversity between forty Plant Introduction (PI) accessions comprising 14 species of Agrostis from 20 countries and determine the genetic relationships among the species based on AFLP profiles.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Plant Materials and DNA Extraction
Bulked leaf samples from 25 plants each of 40 Plant Introduction (PI) accessions (source: USDA, Regional Plant Introduction Station, Pullman, WA) of Agrostis species from different countries were used (Table 1, Fig. 1). Warnke et al. (1997) compared different number of samples per bulk from the same accession in bentgrass and suggested the least variation using 25 plants per bulk for analysis to minimize against sampling errors. Tissues were ground in liquid nitrogen (Extraction buffer: Tris-HCl, SDS, NaCl) and precipitated using chloroform, sodium acetate and ethanol. All samples were treated with RNase and twice reprecipitated. Extracted DNA was stored in 1% TE buffer. DNA quality was checked by running 5 µL of the undigested samples in 1% agarose gel containing TBE buffer and compared to EcoR1 digested samples. DNA quantification was performed using DYNA Quant 200 Fluorometer (Pharmacia Biotech, San Francisco, CA).


View this table:
[in this window]
[in a new window]
  		Table 1. List of plant introductions (PI), species, number of chromosomes and geographic origin of bentgrass (Agrostis spp.) accessions.

 


View larger version (52K):
[in this window]
[in a new window]
  		Fig. 1. World regional sources of 40 accessions of Agrostis species. Legend: CA = Canada; CH = Switzerland; DE = Germany; DK = Denmark; ES = Spain; ET = Ethiopia; IR = Iran; IT = Italy; MO = Mongolia; NL = Netherlands; PT = Portugal; RO = Romania; RU = Russia and former USSR; SE = Sweden; TR = Turkey; UA = Ukraine; US = America; UY = Uruguay; ZA = South Africa.

 
AFLP Analysis
Approximately 150 to 200 ng of DNA was used to do AFLP analysis for each accession. Digestion was conducted with two restriction enzymes, EcoRI, a six base pair cutter and MseI, a four base pair cutter. The AFLP procedure used in this study was as described by Vos et al. (1995) with modifications. Preamplification was done on PTC-100 thermal cycler (MJ Research, Waltham, MA) using 72°C 2 min, 30 cycles of 94°C 30 s, 60°C 30 s, 72°C 1 min, followed by elongation at 72°C 5 min and 4°C hold. PCR products from initial ligation of adapters were checked on 1.5% (w/v) TBE agarose gel. Combinations of fluorescent(*) dye labeled E and M primers each with three selective nucleotides at the 3' ends were used (Table 2). The following E* primers were used: E*-ACA and E*-AGC. The following M primers were tested: M-CAT, M-CAG, M-CGG, M-CTT, and M-CCT. The 15-µL selective amplification mixture consisted of 15 pmol E*-primer, 75 pmol M-primer, 2 mM dNTP, 1.5 mM PCR buffer, 37.5 mM MgCl2, 1.0 unit Taq Polymerase, and 1 µL preamplified product in deionized distilled water. The PCR cycling sequence used for selective amplification was as recommended by Vos et al. (1995). Products from selective amplification were checked initially on 1.5% (w/v) agarose and diluted four times with 0.1 TE buffer. For separation on acrylamide gels, samples consisted of 0.5 µL of the amplified product and 0.5 µL loading dye. The samples were denatured at 95°C for 5 min before loading and ran on a 6% (w/v) Long Ranger polyacrylamide gel with 0.7x TBE buffer with a LI-COR DNA Analyzer A200 (LI-COR Inc., Lincoln, NE) at a constant 800 V for 6 h at 50°C.


View this table:
[in this window]
[in a new window]
  		Table 2. Number of polymorphic bands obtained from different primer combinations. Adapter and primer sequences were adapted from Vos et al. (1995).

 
Chromosome Analysis
Root tips of species with unknown number of chromosomes were collected and fixed in 3:1 ethanol (95%, v/v), acetic acid (Farmer's fixative). The root tip was rinsed and placed in 3 M HCl for 10 to 20 min to soften. On a glass slide, the meristematic region was cut and squashed with a blunt end of the needle and a drop of acetocarmine dye was added. Slides were viewed under a phase contrast microscope to determine the number of chromosomes from a several mitotic cells. Chromosome numbers were determined for PI195197 and PI299461, A. lachnantha Nees.; PI230236, A. munroana Aitch. & Hemsl.; PI283174, A. transcaspica Litv.; PI477045, A. hygrometrica Litv.; and PI362190, A. mongolica Roshev.

Data Analyses
Gels were visualized by means of Gene ImagIR 4.0 (Scanalytics, Inc. VA). Each informative polymorphic band was scored manually as 1 for presence and 0 for absence. Analyses were done using Numerical Taxonomy and Multivariate Analysis System, NTSYS v.2.1 (Rohlf, 1993). Genetic similarities based on Jaccard's coefficients (Jaccard, 1908) were calculated among all possible pairs using the SIMQUAL option and ordered in a similarity matrix. The similarity matrix was run on Sequential, Agglomerative, Hierarchical and Nested clustering, SAHN (Sneath and Sokal, 1973) using Unweighted Pair Group Method with Arithmetic Mean (UPGMA) as an option (Sokal and Michener, 1958). Cophenetic correlation was calculated to measure goodness of fit. Principal component analysis (PCA) was run using SYSTAT to identify the number of groups based on Eigen vectors. The TREE module of NTSYS v.2.1 was used to produce the dendrogram and cluster groupings (Rohlf, 2000).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
AFLP Fingerprinting and Polymorphism Level in Bentgrass Accessions
Selective amplifications for bentgrasses were made using different primer combinations from the final products of EcoRI/MseI digestion, adaptor ligation, and preamplification. From the initial eight selective primer combinations, five combinations were chosen for clarity, repetitiveness in duplicated gel runs, and high levels of polymorphisms. Four hundred polymorphic markers were scored from the five chosen selective primer combinations. The average number of polymorphic bands across the accessions ranged from 100 to 180 markers per primer combination and 20 to 145 polymorphic bands per individual lane. Data in Table 2 showed the selected primers and number of polymorphic markers scored. The high level of polymorphism revealed the wide genetic diversity between the Agrostis accessions. Monte et al. (1993) found that if extensive diversity exists among taxa or genus, fewer probes or primer combinations showing high polymorphism may be sufficient to distinguish genotypes. None of the bentgrass accessions shared an identical DNA fingerprint. AFLP profiles between accessions indicated that the collection does not contain duplications. The high level of polymorphism has facilitated analysis of the genetic diversity among bentgrass genotypes.

The most robust primer combination with the highest number of polymorphism was E-ACA/M-CAG, whereas the lowest number was observed with E-AGC/M-CGG. Approximately 12% of the markers were specific for individual Agrostis species. Specific AFLP markers were found for A. lachnantha, A. vinealis, A. scabra, A. munroana, A. stolonifera/A. palustris, A. transcaspica and A. hygrometrica, whereas no specific markers were detected for A. canina, A. capillaris, A. castellana, A. mongolica and A. gigantea. Data suggest the possibility of developing probes from specific markers to discriminate effectively between some species in the future.

Diversity between Accessions and Species
Gaps in phylogenetic information of Agrostis have resulted from too few hybridization studies, difficulty in obtaining old botanical records, and inconsistencies in morphological characterizations, i.e., synonyms, subspecies, and re-classification of some species. New phenotypes with different number of chromosomes and distinct characteristics have been given new species names (Probatova and Karkevich, 1983). Another possible source of uncertainty results from spontaneous natural hybrids where lineage and ploidy levels are not known. In this study, assessment of number of chromosomes of the 14 species of Agrostis from the literature and conducted chromosome analyses confirmed the wide range in ploidy levels (2n = 14, 21, 42) (Table 1). Physical and cytological examinations of the different species showed that ploidy level may not be indicative of plant size in bentgrass. In the germplasm collection, diploid A. canina was much smaller compared to another diploid A. transcaspica, which in turn had wider and longer leaves and was larger than a triploid, A. munroana. Investigations at the molecular level may provide a clearer understanding of the diversity and relationships in bentgrasses.

In this research, four hundred AFLP markers from five selective primer combinations were used to compare 40 bentgrass accessions. Pairwise similarity coefficients (sc) were computed based on shared and unique amplification products using UPGMA (Table 3). On the basis of similarity coefficients, the closest pair would be PI235440 (Switzerland) and PI235541 (Sweden) at 0.98. Both accessions belonged to A. palustris (creeping bentgrass), had the same number of chromosomes, and were found in Europe. The difference between the two PI lines may be due to minimal within-species allelic variation probably resulting from recombination events. Data in Table 3 also showed that the most dissimilar pair would be PI230235, A. stolonifera from Iran and PI477045, A. hygrometrica from Uruguay. The similarity coefficient was only 0.58 indicating large variability in the genomic constitution. Iran and Uruguay are found geographically in distant hemispheres that separated A. stolonifera and A. hygrometrica and AFLP analysis suggested the two species had the least homology, genetic exchange, or introduction among the accessions studied.


View this table:
[in this window]
[in a new window]
  		Table 3. Genetic similarity coefficients for 40 bentgrass (Agrostis spp.) accessions from data of five primer combinations using fluorescence-labeled AFLP technique.

 
A single dendrogram was generated from the UPGMA cluster analysis with one possible tie found between the closest pair (Fig. 2). A cophenetic-value (ultrametric) matrix was generated from the coefficients of SAHN's cluster analysis of the distance matrix. The cophenetic correlation was calculated (r = 0.96) as a measure of goodness of fit and the results were plotted in a phenogram (Fig. 3). Using SYSTAT, we used a rotated PCA with the bands as observations to determine the number of factors or groups on the basis of eigenvalues greater than one. Seven groups were extracted, which explained 72% of the total variance. The dendrogram showed a similarity coefficient of 0.73 for these seven groups.



View larger version (57K):
[in this window]
[in a new window]
  		Fig. 2. UPGMA dendrogram of 40 accessions of 14 Agrostis spp. from 20 different countries. PCA analysis distinguishes seven groups based on Eigen values > 1.0.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Plot analysis of cophenetic correlation and similarity coefficient as a measure of goodness of fit of the similarity indices. r = 0.95951 = normalized Mantel statistic Z; Approximate Mantel t-test: t = 11.9767; P(Z < obs. Z: P = 1.0000).

 
Group 1 consisted of two accessions of A. canina (Netherlands and Iran, sc = 0.90) and A. vinealis (Russian Federation, sc = 0.76 to A. canina). Though geographically distant in origin, the two A. canina diploids were morphologically similar and AFLP data supported their grouping. In Fig. 1, A. vinealis grouped closest to A. canina than the other 12 species in the study and confirmed their genetic and physical similarity. Species in Group 1 showed very fine, short erect leaves and low growth, but A. vinealis was rhizomatous and confirmed earlier descriptions (Hubbard, 1984; Funk, 1998; Brilman, 2001). In catalogs, A. vinealis was listed as being synonymous with eight other species names: A. canina spp. montana Hartman, A. stricta J.F. Gmel (BONAP Poaceae Listing), A. syreistschikowii P.A.S., A rubra L., A. ericetorum P.B, A. tenuifolia M.B., A. coarctata E., and A. pusilla D. by the Royal Botanic Garden Edinburgh. Jones (1955a) found that A. canina spp. montana (A. vinealis) also known as brown velvet bentgrass, was an autotetraploid form of A. canina and difficult to distinguish in field specimens. Davies (1953) correlated the morphological difference with ecological preference. Velvet bentgrass occurred in natural wet and damp soils while brown velvet bentgrass was found in heaths and upland ground. Molecular characterization by AFLP confirmed the relationship between A. canina and A. vinealis. A. canina was morphologically distinguished from other species of Agrostis physically by having a longer, more pointed ligule and shorter palea than A. capillaris or A. palustris (Hubbard, 1984). Support that A. canina may be grouped separately from the other species was also shown by AFLP data.

Group 2 consisted only of PI234681, A. scabra (Canada), and AFLP analysis showed that the similarity with A. canina was only 0.74. Morphologically, A. scabra was also a nonrhizomatous bunch type grass like A. canina but the two species differed in ploidy levels. Dissimilarity may come from the hexaploid nature and more diverse genetic composition of the species in Group 2. Known as ticklegrass or hairgrass, A. scabra Willd., a delicate species was listed as being synonymous with winter bentgrass (A. hyemalis Tuck.) but differed in flowering time (Lawrence, 1986).

Group 3 consisted of three subgroups. Subgroup A consisted of the seven accessions of A. capillaris (colonial bentgrass) from seven countries. This indicated that the seven allotetraploid accessions were similar and/or may have originated from common diploid ancestors. Visual classification between colonial and highland bentgrass was difficult. Hubbard (1984) separated them by flower color, ligule size and growth habit. Two accessions of A. capillaris (USA) cv. Highland and cv. Exeter were reclassified as A. castellana (Hubbard, 1984; Shildrick, 1976; Brilman, 2001). Steiner and Lupold (1978) supported the suggestion that A. capillaris cv. Highland should belong to A. castellana on the basis of the percentage distribution of the form of palea apex, the presence and angle of basal hairs and awn types. In this study, AFLP analysis of PI469217 Highland and PI578528 Exeter showed that though they contained a small number of specific bands present only in A. castellana, they clustered more with the A. capillaris group. There is no information whether the plant introductions were seeded in open or isolated places during seed increase. Cross contamination at any time may add to the taxonomic variation. Yamamoto and Duich (1994) could not distinguish Highland and Exeter from the other 10 colonial bentgrasses using phosphoglucoisomerase, glutamate oxalotransaminase, peroxidase, and topoisomerase isozymes. Difference at the Pox-2 locus was observed (Yamamoto and Duich 1994) but peroxidase isozymes are often unstable and dependent on stress conditions. Subgroup B consisted of all PI accessions of A. castellana from Spain and Portugal with no distinction as to groupings based on origin. The two subgroups, A and B, may share common diploid species progenitors and may be related with A. capillaris from Germany and Italy.

Subgroup C consisted of PI598462 A. trinii (Russian Federation) and PI443051 A. gigantea (USA). Classification of A. trinii Turcz. has been unclear and listed as being synonymous to A. canina spp. trinii (Turcz.) Hulten (Soreng et al., 2002). Cluster analysis however showed A. trinii to be dissimilar to A. canina (sc = 0.70 to 0.73). Skolovskaya (1938) described A. trinii in Siberia and Orient as having both diploid and tetraploid forms. The diploid form of A. trinii may have been recognized as subspecies of A. canina (Brilman, 2002). PI598462 A. trinii was known to be a tetraploid and appeared different phenotypically from A. vinealis, an autotetraploid form of A. canina. A. trinii (Russian Federation) could be an allopolyploid with one chromosome set different from A. canina and A. vinealis, thus forming separate groups. The other member of this subgroup, A. gigantea (US) surprisingly was not grouped with the A. gigantea (Turkey). Hexaploid A. gigantea genomic constitution of A1A1A2A2A3A3 may have consisted of A1A1 from A. canina, with the A2A2A3A3 probably from A. stolonifera or A1A1A2A2 from A. capillaris (Jones, 1955b). AFLP data supported both possibilities. The two PI lines of A. gigantea may possibly have different chromosome sets but share the common A1A1 genome. Based on the clustering, PI443051 (USA) may have the A1A1A2A2 from A. capillaris while PI383584 (Turkey) may have the A2A2A3A3 from A. stolonifera. Agrostis gigantea (Turkey) may have a second genomic constitution originating from another species. In this study, A. transcaspica (Former USSR) was found to be diploid and clustered closest with the A. gigantea (Turkey). This indicated the possibility that A. transcaspica was closely related to this species and may be the source for the A3A3 genome in A. gigantea (Turkey). This relationship would be interesting to examine in future interspecific hybridization and cytogenetic studies.

Group 4 consisted mainly of the creeping bentgrass (A. palustris and A. stolonifera), A. gigantea (Turkey), A. mongolica, and A. transcaspica. There has been considerable taxonomic confusion regarding creeping bentgrass and whether they should also be A. stolonifera with subspecies palustris, spp. stolonifera or spp. gigantea. The genetic similarity coefficient between PI accessions in this group ranged from 0.82 to 0.95. Genetic dissimilarity computed from 1 - sc x 100% (Zhang et al., 1999) ranged from 5 to 18%. Turf breeders believe A. palustris (USA) have originated from and frequently outcrossed with materials from Europe. AFLP data supported this idea and results showed that USA creeping bentgrasses may share some genetic similarity with those from Switzerland and Sweden. The most divergent in A. stolonifera would be from Turkey, Iran, and the former USSR as opposed to those originating from other parts of Europe. No separation of groups for A. palustris and A. stolonifera was observed on the basis of AFLP, but they differed slightly from A. gigantea. The difference may be due to the third chromosome set of hexaploid A. gigantea. Caceres et al. (2000) used RFLP markers to distinguish four creeping bentgrass A. stolonifera cultivars from A. capillaris Highland. AFLP data supported that PI accessions of A. stolonifera from different parts of the globe were in one group and differed from A. capillaris (Group 3, Subgroup A). Species A. stolonifera also shared slight similarities with A. mongolica (sc = 0.76 to 0.87) and A. transcaspica (sc = 0.73 to 0.80). PI362190 A. mongolica plant type was also stoloniferous and chromosome analysis showed that the bentgrass from Eastern Europe was also a tetraploid. UPGMA analysis grouped A. mongolica (Mongolia) closest to A. stolonifera (Russia). Because Mongolia and Russia are geographically adjacent, these countries may have similar environmental conditions favorable to both species. The fourth species in Group 4, which comprised mostly of stoloniferous bentgrasses, was A. transcaspica, PI283174 (Former USSR). Physical examination showed A. transcaspica also to be creeping but differed in leaf characteristics. The species has wider (6–18 mm) dark green, thick leaves with pointed tips. Soreng et al. (2002) has listed A. transcaspica Litv. as being synonymous to A. stolonifera spp. transcaspica (Litv.) Tzvelev. AFLP clustering and similarity coefficients indicated that A. transcaspica, a diploid may have contributed to the tetraploid or hexaploid creeping bentgrass genome.

Group 5 consisted of A. lachnantha N. (Ethiopia and South Africa). The African bentgrasses were morphologically distinct from the other bentgrass species. PI195917 (Ethiopia) bentgrasses were shorter (10.2–25 cm) than PI299461 (South Africa, >25 cm). The latter also has fewer leaves, harder stalks, and produced very tall flowering panicles (>61 cm) in the greenhouse but were highly sterile. The chromosome number reported here differed from the listing of the Index to Plant Chromosome number (IPCN) with gametophytic count (n = 21) and sporophytic 2n = 28. Chromosome analysis of 2n = 21 may suggest intra- or intercrossing variation during seed increase. The two PI accessions of A. lachnantha has sc = 0.94 on the basis of AFLP analysis. Eleven specific AFLP markers were found that could distinguish A. lachnantha species from the other 13 species.

Group 6 consisted only of PI230236 A. munroana (Iran). Background information about A. munroana Aitch. & Hemsl. was minimal and was earlier referred to as Calamagrostis munroana (Aitch. & Hemsl.) Boiss. in 1884 (Soreng et al., 2002). Chromosome analysis of plants of PI230236 showed 2n = 21 which confirmed earlier reports (Gohil and Koul, 1986; Mouinuddin et al., 1994). Physical analysis of the mature triploid plant showed a short plant stature (5.1–15.2cm) with fine (2–3 mm width), flat, soft, normal green leaves. Species A. munroana was observed to be a bunch type bentgrass and early flowering with the florets openly branched. A. munroana plants were morphologically distinct from triploids of A. lachnantha. Their triploid genome constitutions may be largely unlike and AFLP results showed sc = 0.61, thus forming separate groups.

Group 7 was the most genetically distant from the thirteen other species studied and included A. hygrometrica (Uruguay). Seven specific AFLP markers were found for this species. A. hygrometrica Nees. has nine synonyms in genus Agrostis or Bromidium (Soreng et al., 2002). Plant materials from PI477045 were low growing, bunch-type grasses with hard, lengthy flowering panicles. AFLP analysis showed that bentgrass germplasm from Uruguay (South America) formed a separate group from other bentgrasses of Europe, Asia or North America. Distinct ecological conditions among continents and unique germplasm pools from which the A. hygrometrica may intercross would differentiate the accessions.

Assessment of genetic diversity in germplasm collections from several geographic locations by means of AFLP markers has been conducted for Morus germplasm (Sharma et al., 2000) and the Triticeae tribe (Monte et al., 1993). Positive correlations were found for cluster groupings and geographic distances. Results in this study indicated that geographically adjacent countries like Spain and Portugal have bentgrass accessions which also clustered together as in Group 3, Subgroup B and bentgrass accessions from distant locations (Iran vs. Uruguay) were in separate groups with low similarity coefficients. Possibilities of genetic introductions may have occurred with migration, selection and breeding among the colonial, highland and creeping bentgrasses from Europe and USA. Local environmental adaptation may play a significant role in Agrostis diversity.

AFLP analysis revealed its usefulness for assessing germplasm collection for possible duplications. It may also indicate where incorrect species determinations would be in the GRIN system or germplasm collection. The 400 polymorphic markers from five chosen primer combinations showed the high level of diversity between the Agrostis germplasm and distinguished the seven groups. Differences in ploidy levels did not give ambiguous results in scoring as highly polymorphic, repetitive and specific bands were found. The high cophenetic correlation showed the goodness of fit of the similarity indices. The dendrogram showed the relationships between A. canina with A. vinealis but not with A. trinii. Cluster analysis also showed that two hexaploid A. gigantea from different geographic sources (USA and Turkey) were not grouped together, possibly because of different chromosome sets. A possible diploid progenitor would be A. transcaspica. The percentage genetic dissimilarity among creeping bentgrasses indicated considerable potential for the improvement of turf. Turfgrass breeders may develop superior cultivars either by crosses with germplasm accessions from the same species or among varying species. Important traits from other Agrostis species can be introduced to cultivated bentgrasses and AFLP analysis would be a useful tool to monitor introgression and molecular tagging. By means of specific amplified products, sequence characterized amplified primers may be developed to distinguish genetically the different bentgrass species in the future. AFLP analysis may be used in identifying bentgrass genotypes and clusters, constructing core collections and screening for duplicate, or misclassified accessions in germplasm collections.


   ACKNOWLEDGMENTS
 
Funding for this study was provided by the Michigan Experiment Station. We wish to thank M. McGrath and staff for the use of their equipment, and also thank J. Kelly and D. Sleper for helpful comments on the manuscript.

Received for publication August 31, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
A. B. Hollman, J. C. Stier, M. D. Casler, G. Jung, and L. A. Brilman
Identification of Putative Velvet Bentgrass Clones Using RAPD Markers
Crop Sci., March 28, 2005; 45(3): 923 - 930.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
G. V. Vergara and S. S. Bughrara
Genetic Differentiation of Tetraploid Creeping Bentgrass and Hexaploid Redtop Bentgrass Genotypes by AFLP and their Use in Turfgrass Breeding
Crop Sci., May 1, 2004; 44(3): 884 - 890.
[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 (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Agricola
Right arrow Articles by Vergara, G. V.
Right arrow Articles by Bughrara, S. S.
Related Collections
Right arrow Turfgrass Management
Right arrow Turfgrass

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