Hereditas 138: 101 – 113 (2003) Assessment of genetic variation in timothy (Phleum pratense L.) using RAPD and UP-PCR YANG-DONG GUO, TAPANI YLI-MATTILA and SEPPO PULLI Laboratory of Plant Physiology and Molecular Biology, Department of Biology, Uni6ersity of Turku, Turku, Finland Guo, Y.-D., Yli-Mattila, T. and Pulli, S. 2003. Assessment of genetic variation in timothy (Phleum pratense L.) using RAPD and UP-PCR. — Hereditas 138 : 101– 113. Lund, Sweden. ISSN 0018-0661. Received September 3, 2002. Accepted April 9, 2003 DNA-based fingerprinting technologies including random amplified polymorphic DNA (RAPD) and universally primed PCR (UP-PCR), a novel method for studying genetic variation, were employed as genetic markers for assessing genetic diversity and relationships in timothy (Phleum pratense L.). This study sought to identify the genetic background of the genotypes used in timothy breeding. Thirty eight genotypes from fifteen countries were used as test materials. RAPD and UP-PCR dendrograms based on 132 (from 3 primers) and 44 highly reproducible bands, respectively, were analyzed. The electrophoretic gels showed that the PCR products were informative and polymorphic. Different geographic genotype groups were distinguished according to the combined RADP and UP-PCR results. The results demonstrate that methods based on molecular fingerprinting can be used for timothy identification. Yang-Dong Guo, Laboratory of Grass Breeding, National Agricultural Research Center for Hokkaido Region, National Agricultural Research Organization, Hitsujigaoka, JP-062 -8555 Sapporo, Japan. E-mail: yaguo@affrc.go.jp Timothy (Phleum pratense L.) is one of the major forage crops in global food production, and the most important grass in the northern latitudes. Timothy is valued for its winter hardiness, good palatability, and moderate nutritional feed value. It is an allohexaploid (2n=6x =42) (CAI and BULLEN 1994). The highly outcrossing nature of timothy, however, makes breeding of new varieties with improvements in quantitative traits like yield, protein content, digestibility slow. This is mainly because of the low degree of genetic homozygosity commonly obtainable in this species which necessitates selection for desirable traits to be carried out with very heterozygous material. Genetic variation among individuals within a species and their interrelationships have conventionally been assessed using morphological and agronomic traits, or by a biochemical test such as isozyme analysis (FORD and BALL 1991; LU et al. 2002). However, these approaches are subject to environmental influences and their overall effectiveness in estimating genetic relatedness has been questioned. Consequently, it is now widely accepted that information generated from DNA-based polymorphisms provides the best estimate of genetic diversity (WAUGH 1997). Several molecular marker systems have been used for genetic mapping and biodiversity studies. These systems include random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP) and simple sequence repeat (SSR) (RIDOUT and DONINI 1999). Detection of nucleotide sequence variability and its exploitation as genetic marker have revolutionized many aspects of plant genetics. Initially, variation at the DNA level was detected as restriction fragment length polymorphism (RFLP). However, over the last decade, PCR technology has significantly influenced almost all areas of molecular biology, and modifications of the basic procedure have promoted the development of numerous assays for detecting variation at the nucleotide level (WAUGH 1997). Nevertheless, most of these assays require prior DNA sequence information to enable the design and synthesis of oligonucleotide primers which flank the target sequence and allow selective DNA amplification by PCR. RAPD analysis (WELSH and MCCLELLAND 1990; WILLIAMS et al. 1990) is quick and well suited for the efficient non-radioactive DNA fingerprinting of genotypes. It relies on the use of single, short (usually 10 bp) arbitrary oligomers as primers for PCR-mediated amplification of genomic DNA between closely spaced inverted sequences. RAPD technique has been commonly used to study variation in cereals and grasses such as wheat (FAHIMA et al. 1999; CAO et al. 1999), rice (QIAN et al. 2001), maize (PEJIC et al. 1998), barley (DE BUSTOS et al. 1998), rye (MATOS et al. 2001; PERSSON et al. 2001), Texas bluegrass (RENGANAYAKI et al. 2001) and bromegrass (FERDINANDEZ et al. 2001). The geographical distribution of accessions and genetic diversity were important evidences for the origin of the crops. The RAPD tech- 102 Y.-D. Guo et al. nique was used in genetic variation and geographic diversity studies in the weeds Silene latifolia (VELLEKOOP et al. 1996), Finnish tansy (KESKITALO et al. 1998) and in red algae (VALATKA et al. 2000; GABRIELSEN et al. 2002). It was also used for the detection of somaclonal variation in rice (YANG et al. 1999). However, RAPD has some drawbacks such as lack of reproducibility of amplification. Some of these problems can be resolved by using universally primed PCR (UP-PCR), (YLI-MATTILA et al. 1997a; BULAT et al. 1998; LUBECK et al. 1999). It is a modification of RAPD and was originally developed in the same year as RAPD (BULAT and MIRONENKO 1990). UP-PCR is a PCR fingerprinting method similar to the well-known RAPD technique in that it is possible to amplify DNA from any organism without previous knowledge of DNA sequences and to generate multibanding profiles following gel electrophoresis (BULAT et al. 1998). UP-PCR markers have been used in studies on the genetic variation within fungi (YLIMATTILA et al. 1997a,b, 2002; BULAT et al. 1998; LUBECK et al. 1999) and plants (BULAT et al. 1994). The main differences between RAPD and UP-PCR are that in the latter, longer (16 bp) semi-random primers of unique design and relatively high annealing temperatures (about 55°C) are used. This results in the amplification of numerous DNA fragments (up to 100) randomly distributed throughout the fungi genome (BULAT et al. 1998). The UP-primers consist of the variable part (3% end, 8–10 nt) and ‘‘natural’’ minisatellite-like sequences (5% end, 6– 10 nt), which can be found in any genome. The 5% end is capable of generating numerous PCR products with any kind of template DNA and it stabilizes the primer hybridization at high annealing temperatures (BULAT et al. 1998). The random 3% end of the primer is the most important part which has to match the target sequence with 100 % reliability. The 3% end is designed to avoid the amplification of phylogenetically conserved regions of the genome (BULAT et al. 1998; BULAT and MIRONENKO 1990; PAAVANEN-HUHTALA 2000). The main advantages of UP-PCR over RAPD are the reproducibility and the banding profiles which consist of higher numbers of bands than most RAPDs, thus facilitating identification of specific markers, and at the same time showing conserved bands (YLI-MATTILA et al. 1997a; BULAT et al. 1998; LUBECK et al. 1999). Knowledge of germplasm diversity and interrelationships among breeding materials and varieties is important for the genetic improvement of plants. It is therefore useful for timothy breeders to know the genetic background of the breeding materials and varieties. The dual aims of this study were to deter- Hereditas 138 (2003) mine genetic diversity, using RAPD and UP-PCR markers, and to investigate the relationships among timothy genotypes. This is the first attempt in the application molecular techniques for studying genetic variation and interrelationships in timothy. MATERIAL AND METHODS Plant material Thirty eight genotypes of timothy from 15 countries were used in this study (Table 1). About 50 seeds of each were sown into plastic pots filled with 1:1 peat/ sand mixture. Plants were watered weekly and fertilized with NPK (9 %–16 %–22 %, w/w). When the plants reached 3– 5 cm in height, they were used to extract DNA. Plant growth conditions were: 24°C/ 18°C, day/night temperature and a 16-h photoperiod Table 1. Genotypes and country of origin of timothy used for this study. Genotype Source Våti 7702 Våti 7701 Engmo Bodin Grinstad Kämpe II Saga SvA, 0918 SvA, 0896 Erecta Rvp Alma Iki Tuukka Tammisto Heidemij Ligntasso Lirocco Liphlea Liglory Comtal Adda Zolis Topas Foka Drummond Farol Glenmore Winmore Vetrovsky Sobol Barmidi Phlewiola Mariposa Kiritappu Kumpu Hokushu Nosappu Akkeshi Norway Norway Norway Norway Norway Sweden Sweden Sweden Sweden Belgium Finland Finland Finland Finland Germany Germany Germany Germany Germany France Iceland Lithuania Denmark Poland Canada Canada Canada Canada Czech Republic Czech Republic The Netherlands Netherlands USA Japan Japan Japan Japan Japan Hereditas 138 (2003) at 80 mmol m − 2 s − 1 supplemented by fluorescent lamps. DNA extraction RAPD and UP-PCR analysis was performed using DNA from 25 plantlets per genotype. Total DNA was extracted from 0.20 g of fresh leaf tissue of 10– 12 days old young plantlets. The leaves were cut into small pieces and put into an eppendorf tube. Plant tissues were ground in liquid nitrogen in advance and either kept frozen or lyophilized. DNA extraction protocol was based on the CTAB method (WEISING et al. 1995) without RNAse treatment. Plant DNA was stored in TE buffer (1 mM Tris/HCl pH 7.8; 0.1 mM EDTA) at 4°C (for instant use) or at −20°C (for longer storage). DNA concentration was measured by running agarose gels with Boehringer/ Mannheim marker VI. The isolated DNA may still contain a considerable amount of RNA, polysaccharides, proteins which are tightly bound to DNA, and other contaminants (WEISING et al. 1995). Contamination of DNA by these compounds inhibits PCR reaction. In this study, good quality DNA was white and no good DNA was dark brown/brown. Good quality DNA was obtained by using very young plantlets (aged 10– 12 days) for DNA extraction. RAPD-PCR amplification and electrophoresis Amplification was performed by a PTC-200 Peltier Thermal Cycler. The PCR reaction mixture (25 ml) contained 0.4 mM dNTPs, 580 nM primer, 2.5 ml 10× Dynazyme reaction buffer, 0.5 unit Dynazyme polymerase (Finnzymes, Espoo, Finland), 1 ml DNA and deionized sterile water. The amplification conditions were: one cycle of 94°C for 90 s; 39 cycles of 93°C for 35 s, 35°C for 25 s, 72°C for 95 s. After the final cycle, one cycle of extension at 72°C for 180 s was added. Bands were detected visually from the photographs of the 1 % agarose gel (LE, analytical grade, Promega) in 0.5× TBE buffer solution with ethidium bromide. Amplification was repeated 2– 3 times for each sample and only reproducible results were accepted into the final data matrix. In RAPD, primers Operon B-1 to B-20, 91373, 91299, X, Y, 91300, Operon A-1 and Operon A-3 were tested (Table 2). UP-PCR amplification and electrophoresis UP-PCR was performed in 25 ml volume containing 0.4 mM dNTPs, 580 nM primers, 2.5 ml 10 × Dynazyme reaction buffer and 0.5 unit Dynazyme polymerase. The program for UP-PCR consisted of one cycle of 3 min at 94°C followed by 30 cycles of 50 s Assessment of genetic 6ariation in timothy 103 at 92°C, 70 s at 55°C and 60 s at 72°C. After the final cycle, one cycle of 72°C for 180 s was added. Annealing conditions were optimized by performing a series of trial PCR amplifications for different primers. Amplification products of UP-PCR were electrophoresed in 1.8 % Metaphor agarose gel (FMC BioProducts, Rockland, ME, USA) in 0.5× TBE buffer solution containing ethidium bromide. In UPPCR, primers AA2M2, HE 45, AS 19 and As 15 were tested (Table 2). Data scoring and statistical analysis Amplified fragments, named by the primer used and the molecular weight in base pairs (bp), were scored for the presence (1) or absence (0) of bands of the same size and a matrix of the different marker phenotypes was assembled. The gels were scored for the presence or absence of only those major bands which showed a reproducible pattern among genotypes. The program GEL (PATZEKIN V and KLOPOV N, Petersburg Nuclear Physics Institute, Russia) was used for processing 1 D gel images. Data produced by the RAPD and UP-PCR analyses were analyzed by PHYLIP 3.5 (FELSENSTEIN 1993) using the unweighted pair group method with arithmetic mean (UPGMA) and neighbor-joining (NJ) method. Qualitative similarity coefficients (DICE) were calculated between all pairs of electrophoretic phenotypes according to the equation: Sxy =2nxy/(nx +ny), where nxy is the number of shared bands and nx and ny are the number of bands in the electrophoretic phenotypes x and y respectively (NEI and LI 1979). Distance (1-DICE) matrices in RAPD and UP-PCR patterns obtained by the Distnew program (KLOPOV N, Petersburg Nuclear Physics Institute, Russia) were used to calculate the pair-wise genetic distances between genotypes. These distance matrices were then employed to construct dendrograms using the UPGMA and NJ methods. UPGMA trees can be created by reducing all information on genetic characters to pairwise estimates of similarity or dissimilarity. RESULTS RAPD fingerprints of strains Primers which had informative and polymorphic products resolvable by electrophoresis were selected. Out of 27 primers, 25 produced 354 scoreable bands (Table 2), and the primer OPB-2 generated highly reproducible and consistently well-amplified bands, ranking in size from 2410 bp to 359 bp (Fig. 1). Usually, products below 330 bp or above 2500 bp gave faint and non-reproducible bands which were 104 Y.-D. Guo et al. Hereditas 138 (2003) Table 2. Sequences of oligonucleotide primers used for RAPD and UP-PCR amplification of timothy total DNA and the polymorphism obtained. Primer Sequence (5% to 3%) Total bands Polymorphic bands Percentage polymorphis Band Mol. weight (bp) RAPD-PCR primers 91373 91299 X Y 91300 Operon A-1 Operon A-3 Operon B-1 Operon B-2 Operon B-3 Operon B-4 Operon B-5 Operon B-6 Operon B-7 Operon B-8 Operon B-9 Operon B-10 Operon B-11 Operon B-12 Operon B-13 Operon B-14 Operon B-15 Operon B-16 Operon B-17 Operon B-18 Operon B-19 Operon B-20 CGTAGTGGTG CGATTCGGCG GATAACGCAC CGAGACACAC CGAGGTTCGC CAGGCCCTTC AGTCAGCCAC GTTTCGCTCC TGATCCCTGG CATCCCCCTG GGACTGGAGT TGCGCCCTTC TGCTCTGCCC GGTGACGCAG GTCCACACGG TGGGGGACTC CTGCTGGGAC GTAGACCCGT CCTTGACGCA TTCCCCCGCT TCCGCTCTGG GGAGGGTGTT TTTGCCCGGA AGGGAACGAG CCACAGCAGT ACCCCCGAAG GGACCCTTAC 7 6 5 12 37 9 6 14 37 28 9 11 8 11 0 14 14 58 0 9 9 7 5 9 9 9 11 7 2 5 5 21 7 3 11 29 22 5 6 4 7 0 11 13 11 0 7 6 4 3 4 5 6 10 100.0 33.3 100.0 41.7 56.9 77.8 50.0 78.6 78.4 78.6 55.6 54.6 50.0 63.6 – 78.6 92.9 18.9 – 77.8 66.7 57.1 60.0 44.4 55.6 66.7 90.9 1217-750 1465-523 1568-582 1761-497 1548-327 2135-572 1790-389 2898-415 2410-359 2679-569 2720-438 2180-547 2136-590 2045-463 – 2368-501 1648-679 2485-383 – 1936-664 1879-467 1753-430 2254-723 1549-447 2538-636 1946-468 1845-554 UP-PCR primers AA2M2 HE 45 AS 19 AS 15 GAGCGACCCAGAGCGG GTAAAACGAGGCCAGT GAGGGTGGCGGCTAG GGCTAAGCGGTCGTTAC 44 7 11 9 19 6 4 5 43.2 85.7 36.4 55.6 2341-337 1749-304 1136-362 1598-637 not used for data analysis. In RAPD study, the primers OPB-2, OPB-11 and 91300 were finally selected for dendrogram building, because they gave highly polymorphic and reproducible banding patterns. Genetic relationships among populations by RAPD RAPD analysis revealed extensive polymorphism among the different genotypes of timothy. In the pairwise distance matrix (Table 3) ca 50 % of the 1-DICE coefficient values were lower than 50 %. An UPGMA dendrogram based on the 132 RAPD bands (using primers OPB-2, OPB-11 and 91300) clustered the 38 genotypes within distinct groups according to their geographical origin with the exception of the American variety Mariposa (Fig. 3). Two main groups were identified at the similarity level of 72 %: In the first group, except for Mariposa, all other genotypes originated from central and northern Eu- rope. It could be divided into two subgroups, of which one included all Norwegian genotypes and another included all Finnish and Swedish genotypes, except one Swedish genotype (Kämpe II), which was closer to Norwegian genotypes. The second group included timothy genotypes from Canada, Germany, Denmark, The Netherlands and Japan. It could be divided into two subgroups, of which one included all Japanese and Dutch genotypes and another included all Canadian and German genotypes. The Polish genotype Foka was separated from the two main groups. The similar clusters were also obtained by the NJ method (dendrogram not shown). UP-PCR fingerprints of strains and genetic relationships among populations Of the four primers tested (Table 2), AA2M2 gave the highest polymorphism. Altogether, 44 clear repeatable UP-PCR bands were obtained (Fig. 2). In Hereditas 138 (2003) Assessment of genetic 6ariation in timothy 105 Fig. 1. RAPD patterns of timothy genotypes using the primer OPB-2. The MW lanes are mol. wt.markers (Boehringer/Mannheim, marker VI) and NC is negative control. the pairwise distance matrix (Table 4), the low 1DICE coefficient values was smaller than that in RAPD analysis (Table 3), but the distribution of low distance values was similar to that in RAPD analysis. The UP-PCR UPGMA dendrogram was made by using primer AA2M2. These 38 genotypes were clustered into four main groups at the similarity level of 83 % in the UPGMA dendrogram (Fig. 4). The first main group contained all Japanese and Norwegian genotypes. All of the German genotypes, the Danish and Polish genotype and most of the Canadian genotypes (three out of four) were clustered in the second main group. The genotypes from northern Europe were dominant in groups 3 and 4. Three of the Finnish genotypes were clustered in group 4. In the NJ dendrogram, the subgroups were generally the same as in the UPGMA tree (dendrogram not shown). Genetic relationships among populations by combination of RAPD and UP-PCR An UPGMA tree was finally made by combining the binary data resulting from RAPD and UP-PCR (Fig. 5). This tree largely confirmed the RAPD results and showed the clustering of timothy genotypes into two main groups. In the first group, genotypes from northern Europe were dominant, with Norwegian and Finnish genotypes clustered to different subgroups at the similarity level of 82 % and 84 %, respectively. The second main group consisted of the genotypes originating from Canada, Germany, The Netherlands and Japan. It could be divided into two subgroups, of which one included all Japanese and Dutch genotypes and the other one all Canadian and German genotypes. In combination NJ dendrogram of RAPD and UP-PCR, both classifications of the RAPD tree and the UP-PCR tree were visible (dendrogram not shown). DISCUSSION Timothy is thought to have been introduced to North America from northern Europe in the early 1700s as the name of Herd’s grass, where it is considered indigenous. Cultivated strains were later reintroduced into Britain from North America. However, at that time it was being cultivated in Sweden under the name A8 ngkampe (MARTIN et al. 1976; BARNES et al. 1995; BERG et al. 1996). Timothy cultivars are heterogenous populations, most timothy cultivars are synthetics, i.e. the advanced generation of two to many parental clones that were interpollinated. Parental clones are selected on the basis of phenotype, the performance of the individual plants, or combining ability, which is determined through progeny testing. In timothy breeding, mass and recurrent mass selection have been the most frequently used methods (BARNES et al. 1995; BERG et al. 1996). Some cultivars have been selected from land races or ecotypes. In other cases, elite germplasm has been recombined and mass selection applied to the progenies. Some Canadian cultivars have been developed by recurrent mass selection from European populations (BERG et al. 1996). Thus, it is easy to understand that the Canadian genotypes are always close to some European (e.g. German) genotypes (Fig. 3 –5). 106 Y.-D. Guo et al. Table 3. Data matrix of pairwise distances constructed from RAPD banding. Distances smaller than 0.4 underlined. Hereditas 138 (2003) Hereditas 138 (2003) Assessment of genetic 6ariation in timothy 107 Fig. 2. UP-PCR patterns of timothy genotypes using the primer AA2M2. The MW lanes are mol. wt.markers (Boehringer/Mannheim, marker VI) and NC is negative control. Northern Europe is a center of timothy diversity and has abundant genetic resources, including local varieties and wild relatives. Northern European genotypes were classified in all of the groups in the dendrograms (Fig. 3– 5). Some Finnish varieties such as Iki, Alma and Tuukka were developed from ecotypes and landraces indigenous to Finland. The relatively high genetic similarity of Finnish genotypes is demonstrated in Fig. 4 and 5. This is in agreement with a genotype evaluation study of timothy in which ten phenologic, agronomic and wintering characters were selected for cluster analysis using Ward minimum variance method. Finnish varieties ‘‘Alma’’ and ‘‘Tuukka’’ grouped close together (LARSEN and HONNE 2001). According to Dr. Jönsson, of Svalöv Weibull AB in Sweden (pers. comm.), some Swedish genotypes, including Kämpe II, Saga, SvA, 0896 and SvA, 0918, have tight genetic linkages with a Finnish background, due to the geographic proximity of Sweden and Finland. Kämpe II is a selection out of Kämpe, which was a selection out of Finnish wild timothy from the Kuopio area. ‘‘Saga’’, ‘‘SvA, 0896’’ and ‘‘SvA, 0918’’ are crosses between A, 0850 and Finnish wild materials from Halola. The above relationship in timothy genetic background was further confirmed by our results (Fig. 3–5). Genetic similarities among Norwegian timothy genotypes are relatively high: in combination of RAPD and UP-PCR binary matrices analysis of UPGMA dendrogram, all five Norwegian genotypes are located in one subgroup (Fig. 5). According to Dr. Larsen of the Norwegian Crop Research Institute, Norway (pers. comm.), ‘‘Engmo’’ and ‘‘Bodin’’ are both local strains from central Norway, and ‘‘Våti 7701’’ is a synthetic population of clones from crosses between ‘‘Bodin’’ and ‘‘Engmo’’. ‘‘Grinstad’’ is a local strain from southeastern Norway. ‘‘Våti 7702’’ is a true synthetic population of 14 selected clones from local populations and some foreign varieties. The UP-PCR dendrogram showed the Japanese genotypes ‘‘Kumpu’’ and ‘‘Hokushu’’ to be identical (Fig. 4). In RAPD and UP-PCR combined dendrogram, all Japanese genotypes were grouped together (Fig. 5), showing their high similarity. Timothy was introduced in Japan from North America in 1870s (SHIMOKOHJI 1998). According to Dr. Yamada of the National Agricultural Research Center for Hokkaido Region, Japan (pers. comm.), most Japanese varieties have been crossed between Japanese strains (Hokkaido ecotypes) and foreign varieties. In RAPD, UP-PCR and combination dendrograms, the northern Europe geographical races (i.e. Norwegian and Finnish), the central Europe geographical races (i.e. German and Dutch), Canadian geographical races and Japanese geographical races were distinguished. The distribution of these races seemed no correlation with climatic or soil data. Perhaps the present-day distribution of the population of timothy is not the result of a process of selection for different gene combinations in different geographical areas, but as the result of a past migration. This hypothesis of timothy is similar with that of the weed Silene latifolia which spread over Europe (VELLEKOOP et al. 1996) and Finnish tansy (KESKITALO et al. 1998). The distributions of the RAPD and UP-PCR characters provide support for the hypothesis that north- 108 Y.-D. Guo et al. Table 4. Data matrix of pairwise distances constructed from UP-PCR banding. Distances smaller than 0.4 underlined. Hereditas 138 (2003) Hereditas 138 (2003) Assessment of genetic 6ariation in timothy 109 Fig. 3. UPGMA dendrogram showing genetic relationships among the 38 timothy populations based on similarity (DICE) coefficient from RAPD data. ern Europe is the native region of timothy, it spread to central Europe and northern America, then it was introduced to Japan from northern America. In the RAPD and UP-PCR combination dendrogram, the genotypes from northern Europe were dominant in the first group and genotypes from northern America, central Europe and Japan were located in the second group (Fig. 5). That timothy spread from northern Europe to northern America then to Japan is clear (MARTIN et al. 1976; BARNES et al. 1995; BERG et al. 1996; SHIMOKOHJI 1998) and support the present results. The results from this study showed that RAPD and UP-PCR techniques measured sufficient polymorphism for DNA typing, and may be powerful tools for the genetic dissection of timothy. The polymorphism between timothy genotypes was smaller than in the red alga Furcellaria lumbricalis (VALATKA 110 Y.-D. Guo et al. Hereditas 138 (2003) Fig. 4. UPGMA dendrogram showing genetic relationships among the 38 timothy populations based on similarity (DICE) coefficient from UP-PCR data. et al. 2000), which may be due to different primers used and the sexual reproduction in timothy. Asexually reproducing species tend to have lower levels of within-populational genetic diversity and higher levels of population differentiation compared to sexually reproducing ones (SOSA and LINDSTRÖM 1999). It seems that the resolution of UP-PCR is somewhat better than that in RAPD in resolving differences and similarity groups between the timothy genotypes. There was little contradiction between UP-PCR and RAPD results, which is in agreement with previous results obtained in Fusarium (YLI-MATTILA et al. 1996; 1997a), Trichoderma, Gliocladium (BULAT et al. 1998) and Verticillium fungi (MITINA and YLIMATTILA 2002). As a general conclusion, we indicated that timothy genetic variation is largely dependent on the geographic diversity. Most timothy cultivars are synthetics. Using different geographic background populations as parental clones will promote heterosis in timothy breeding. The present study demonstrates the indigenous center and biodiversity of this species. In addition, future analysis of genetic variation by combination of new UP-PCR and RAPD primers and other molecular markers like AFLP and SSR and morphological and physiological characteristics Hereditas 138 (2003) Assessment of genetic 6ariation in timothy 111 Fig. 5. UPGMA dendrogram showing genetic relationships among the 38 timothy populations based on similarity (DICE) coefficient from combined RAPD and UPPCR data. will allow us to understand better of timothy genetic variation. ACKNOWLEDGEMENTS The authors thank Dr. Toshihiko Yamada (National Agricultural Research Center for Hokkaido Region, Japan), Dr. Hans-Arne Jönsson (Svalöv Weibull AB, Sweden) and Dr. Arild Larsen (The Norwegian Crop Research Institute, Norway) for providing timothy genetic background information. We thank Mr. Pertti Parssinen (Boreal Plant Breeding Finland) for the provision of timothy seed and for the technical discussion. This work was supported by the Finnish Ministry of Agriculture and Forestry. REFERENCES Barnes, R. F., Miller, D. A. and Nelson, C. J. 1995. 112 Y.-D. Guo et al. Forages. – In: An introduction to grassland agriculture. Vol. I. Iowa State Univ. Press, pp. 305 – 311. Berg, C. C., Mcelroy, A. R. and Kunelius, H. T. 1996. Timothy. – In: Moser, Buxton and Casler (eds), Coolseason forage grasses. Madison Wisconsin Publishers, Madison, Wisconsin, USA, pp. 643– 665. Bulat, S. A. and Mironenko, N. 1990. Species identity of the phytopathogenic fungi Pyrenophora teres Dreschler and P. graminea Ito & Kuribayashi. – Mikologiya i Fitopatologia 24: 435 –441. Bulat, S. A., Mironenko, N. V., Lapteva, M. N. and Strelchenko, P. P. 1994. Polymerase chain reaction with universal primers (UP PCR) and its application to plant genome analysis. – In: R. P. Adams, J. S. Miller, E. M. Golenberg and J. E. Adams (eds), Conservation of plant genes II: utilization of ancient and modern DNA. Monographs in systematic botany from the Missouri botanical garden, Vol. 48. Missouri Botanical Garden, p. 113 –129. Bulat, S. A., Lubeck, M., Mironenko, N. et al. 1998. UP-PCR analysis and ITS1 ribotyping of strains of Trichoderma and Gliocladium. – Mycol. Res. 102: 933 – 943. Cai, Q. and Bullen, M. R. 1994. Analysis of genome-specific sequences in Phleum species: identification and use for study of genomic relationships. – Theor. Appl. Genet. 88: 831 – 837. Cao, W., Scoles, G., Hucl, P. and Chibbar, R. N. 1999. The use of RAPD analysis to classify Triticum accessions. – Theor. Appl. Genet. 98: 602 – 607. de Bustos, A., Casanova, C., Soler, C. and Jouve, N. 1998. RAPD variation in wild populations of four species of the genus Hordeum (Poaceae). – Theor. Appl. Genet. 96: 101– 111. Fahima, T., Sun, G. L., Beharav, A. et al. 1999. RAPD polymorphism of wild emmer wheat populations, Triticum dicoccoides, in Israel. – Theor. Appl. Genet. 98: 434– 447. Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package) version 3.5. Univ. of Washington, Seattle, USA. Ferdinandez, Y. S. N., Somers, D. J. and Coulman, B. E. 2001. Estimating the genetic relationship of hybrid bromegrass to smooth bromegrass and meadow bromegrass using RAPD markers. – Plant Breeding 120: 149– 153. Gabrielsen, T. M., Brochmann, C. and Rueness, J. 2002. The Baltic Sea as a model system for studying postglacial colonization and ecological differentiation, exemplified by the red alga Ceramium tenuicorne. – Mol. Ecol. 11: 2083– 2095. Ford, B. A. and Ball, P. W. 1991. Allozyme diversity and genetic relationships among North American members of the short-beaked taxa of Carex sect. Vesicariae (Cyperaceae). – Syst. Bot. 16: 116 – 131. Keskitalo, M., Linden, A. and Valkonen, J. P. T. 1998. Genetic and morphological diversity if Finnish tansy (Tanacetum 6ulgare L., Asteraceae). – Theor. Appl. Genet. 96: 1141 –1150. Larsen, A. and Honne, B. I. 2001. Evaluation of local populations for timothy breeding. – In: P. Monjardino, A. da Camara and V. Carnide (eds), Breeding for stress tolerance in fodder crops and amenity grasses. Proc. 23rd meeting of the fodder crops and amenity grasses section of EUCARPIA. Azores, Portugal, 2000, p. 46 – 50. Lu, H., Li, J. S., Liu, J. L. and Bernardo, R. 2002. Allozyme polymorphisms of maize populations from Hereditas 138 (2003) southwestern China. – Theor. Appl. Genet. 104: 119 – 126. Lubeck, M., Alekhina, I. A., Lubeck, P. S. et al. 1999. Delineation of Trichoderma harzianum into two different genotypic group by a highly robust fingerprinting method, UP-PCR, and UP-PCR product cross-hybridization. – Mycol. Res. 103: 289 – 298. Martin, J. H., Leonard, W. H. and Stamp, D. L. 1976. Principles of field crop production, 3rd edn. – Macmillan Publishing Co, pp. 581 – 587. Matos, M., Pinto-Carnide, O. and Benito, C. 2001. Phylogenetic relationships among portuguese rye based on isozyme, RAPD and ISSR markers. – Hereditas 134: 229 –236. Mitina, G. and Yli-Mattila, T. 2002. RAPD-PCR, UPPCR and rDNA sequence analyses of entomopathogenic Verticillium lecanii and Beau6eria bassiana strains. – J. Russian Phytopathol. Soc. (in press). Nei, M. and Li, W. H. 1979. Mathematical models for studying genetic variation in terms of restriction endonucleases. – Proc. Natl. Acad. Sci. USA 76: 5269 – 5273. Paavanen-Huhtala, S. 2000. Molecular based assays for the determination of diversity and identification of Fusarium and Gliocladium fungi. PhD thesis, Univ. of Turku, Finland. Pejic, I., Ajmone-Marsan, P., Morgante, M. et al. 1998. Comparative analysis of genetic similarity among maize inbred lines detected by RFLPs, RAPDs, SSRs, and AFLPs. – Theor. Appl. Genet. 97: 1248 – 1253. Persson, K., Diaz, O. and Bothmer, R. von 2001. Extent and patterns of RAPD variation in landraces and cultivars of rye (Secale cereale L.) from northern Europe. – Hereditas 134: 237 – 243. Qian, W., Ge, S. and Hong, D. Y. 2001. Genetic variation within and among populations of a wild rice Oryza granulata from China detected by RAPD and ISSR markers. – Theor. Appl. Genet. 102: 440 – 449. Renganayaki, K., Read, J. C. and Fritz, A. K. 2001. Genetic diversity among Texas bluegrass genotypes (Poa arachnifera Torr.) revealed by AFLP and RAPD markers. – Theor. Appl. Genet. 102: 1037 – 1045. Ridout, C. J. and Donini, P. 1999. Use of AFLP in cereals research. – Trends Plant Sci. 4: 76 – 79. Shimokohji, H. 1998. Timothy. – In: T. Sanbuichi, T. Tochiya and H. Sasaki (eds), Crop breeding in Hokkaido. Hokkaido Kyodo Kumiai Tsushinsha, p. 245 –263 (in Japanese). Sosa, P. A. and Lindström, S. C. 1999. Isozymes in macroalgae (seaweeds): genetic differentiation, genetic variability and applications in systematics. – Eur. J. Phycol. 34: 427 – 442. Valatka, S., Mäkinen, A. and Yli-Mattila, T. 2000. Analysis of genetic diversity of Furcellaria lumbricalis (Gigartinales, Rhodophyta) in the Baltic Sea by RAPR-PCR technique. – Phycology 39: 109 – 117. Vellekoop, P., Buntjer, J. B., Maas, J. W. and van Brederode, J. 1996. Can the spread of agriculture in Europe be followed by tracing the spread of the weed Silene latifolia. A RAPD study. – Theor. Appl. Genet. 92: 1085 –1090. Waugh, R. 1997. RAPD analysis: use for genome characterization, tagging traits and mapping. – In: Clark, MS (ed.), Plant molecular biology, a laboratory manual. Springer, pp. 305 –333. Weising, K., Nybom, H., Wolff, K. and Meyer, W. 1995. DNA fingerprinting in plants and fungi. – CRC Press, pp. 55– 56. Hereditas 138 (2003) Welsh, J. and McClelland, M. 1990. Fingerprinting genomes using PCR with arbitrary Primers. – Nucl. Acid Res. 18: 7213 – 7219. Williams, J. G. K., Bubelik, A. R., Livak, K. et al. 1990. DNA polymorphisms amplification by arbitrary primers are useful as genetic markers. – Nucl. Acid Res. 18: 6531 – 6535. Yang, H., Tabei, Y., Kamada, H. et al. 1999. Detection of somaclonal variation in cultured rice cells using digoxigenin-based random amplified polymorphic DNA. – Plant Cell Rep. 18: 520 – 526. Yli-Mattila, T., Paavanen, S., Hannukkala, A. et al. 1996. Isozyme and RAPD-PCR analyses of Fusarium a6enaceum strains in Finland. – Plant Pathol. 45: 126 – 134. Assessment of genetic 6ariation in timothy 113 Yli-Mattila, T., Mironenko, N. V., Alekhina, I. A. et al. 1997a. Universally primed polymerase chain reaction analysis of Fusarium a6enaceum isolated from wheat and barley in Finland. – Agric. Food Sci. Finland 6: 25 – 36. Yli-Mattila, T., Paavanen-Huhtala, S., Hannukkala, A. et al. 1997b. RAPD-PCR, UP-PCR, rDNA RFLP, rDNA sequence and isozyme analyses of Fusarium isolates in Finland. – Cereal. Res. Comm. 25: 629 –630. Yli-Mattila, T., Paavanen-Huhtala, S., Bulat, S. A. et al. 2002. Molecular, morphological and phylogenetic analysis of Fusarium a6enaceum/F. arthrosporioides/F. tricinctum species complex-a polyphasic approach. – Mycol. Res. 106: 655 – 669.