High SSR diversity but little differentiation between accessions of Nordic timothy (Phleum pratense L.)

High SSR diversity but little differentiation between accessions of Nordic timothy (Phleum pratense L.) P. TanhuanPää and O. Manninen Biotechnology and Food Research, MTT Agrifood Research Finland, Jokioinen, Finland Tanhuanpää, P. and Manninen, O. 2012. high SSR diversity but little differentiation between accessions of nordic timothy (Phleum pratense L.). – Hereditas 149: 114–127. Lund, Sweden. eiSSn 1601-5223. Received 20 September 2011. accepted 29 May 2012. a large collection of genebank accessions of the hexaploid outcrossing forage grass species timothy (Phleum pratense L.) was for the first time analysed for SSR diversity on individual, population and regional level. Timothy is the most important forage grass species in the nordic countries. eighty-eight timothy accessions from nordic countries and eight accessions around europe were analysed with recently developed simple sequence repeat (SSR) markers. Timothy proved to be very polymorphic: the 13 selected SSRs amplified a total of 499 polymorphic alleles, the number of alleles per SSR locus varying from 15 to 74. Taking all SSR alleles together, the observed number in each accession ranged from 95 to 203. Levels of diversity were found to be significantly different between countries, vegetation zones and different cultivar types. however, the differentiation between accessions was low: most of the variation (94%) in the studied timothy material was due to variation within accessions and only 5% was between accessions and 1% between countries. Lack of geographical differentiation may reflect the outcrossing and hexaploid nature of timothy. Our results showed that neutral SSR markers are suitable for demonstrating levels of diversity but not alone adequate to resolve population structure in timothy. nordic timothy material seems to be diverse enough for breeding purposes and no decline in the level of diversity was observed in varieties compared to wild timothy populations. Challenges in analysing SSR marker data in a hexaploid outcrosser were discussed. Pirjo Tanhuanpää, Plant Genomics, Biotechnology and Food Research, MTT Agrifood Research Finland, FI-31600 Jokioinen, Finland. e-mail: [email protected] hereditas 149: 114–127 (2012) © 2012 The authors. This is an Open access article. DOi: 10.1111/j.1601-5223.2012.02244.x Timothy (Phleum pratense L.) is a cool-season perennial grass species distributed naturally throughout europe and parts of north africa and asia. Wild populations of P. pratense represent a polyploid series from diploids to octoploids. The cultivated form of timothy is hexaploid. The uniformity of the molecular profile in agricultural P. pratense suggests that the formation of this hexaploid is probably post-glacial (St e w a r t et al. 2011). The genomic composition of hexaploid timothy has not been fully resolved yet, but there is some evidence that the genome contains four doses of bertolonii genome and two doses of rhaeticum. Both of these genomes derive from the same progenitor and are not very differentiated which explains that both hexasomic and tetradisomic inheritance has been reported in timothy (St e w a r t et al. 2011). Timothy is cultivated for hay, silage and pasture across the northern hemisphere. in nordic countries timothy is the most important forage grass species due to adaptation to the cool and relatively humid northern climate. The main goal in timothy breeding for this region is to combine high yield, good winter survival, and high feeding quality. Timothy breeding relies on broad genetic variation and utilisation of heterosis, which can be achieved by combining genetically distant individuals with good combining ability in a synthetic variety. Therefore, plant breeders have to be sure that they have sufficient genetic variation available for their breeding programmes. Genetic diversity within a plant species reflects both the life history traits and distribution of the species. Perennial, outcrossing species are known to have higher genetic diversity and less differentiation among populations than annual self pollinators (Ha m r i c k and Go d t 1996). Timothy is a perennial wind pollinating species where hexaploidy is expected to further rise the level of diversity. The abundant centre model (Br o w n 1984) presumes reduced neutral genetic diversity within peripheral compared to more central populations (ec k e r t et al. 2008). although the distribution of hexaploid timothy covers most of europe (co n e r t 1998), at the northern margin, namely northern boreal and alpine vegetation zones, harsh winter conditions may limit survival of timothy. This may be reflected in the levels of diversity. Previous studies have also shown that genetic diversity of plant populations may either increase or decrease with increasing altitude (Ya n et al. 2009). nordGen, the nordic Genetic Resource Center, has a collection of 716 accessions of timothy, originating mostly from nordic countries. Sixty-four of these are cultivars, others represent natural populations or old landraces. nordic countries represent a wide geographical region hereditas 149 (2012) SSR diversity in Nordic timothy 115 with varying growth conditions from southern nemoral zone in Denmark to northern alpine vegetation zone in northern norway. in addition to latitude, growth conditions are also affected by longitude since conditions in western norway are maritime and in eastern Finland more continental. Three hundred and seventy-three timothy accessions in the nordGen collection have been previously characterised for morphological and agronomic traits in Finland, norway, iceland and Sweden during 1995–1996 (<www.nordgen.org/index.php/skand/ content/view/full/344>). Characterisations were mostly made on coarse, relative scale and variation within each accession was not taken into account. This data gives an overall picture of the phenotypic variation present in the collection. however, it doesn’t fully describe the levels of genetic diversity, the genetic structure of variation between and among populations nor the genetic distances between individuals or populations. There are many molecular marker systems available for diversity analyses, from which we chose simple sequence repeats (SSRs) for studying diversity in timothy. Primers for 355 SSR loci in timothy have been developed (ca i et al. 2003), and some of the loci have been located on a diploid timothy map (ca i et al. 2009). SSRs are mostly codominantly inherited, very polymorphic, and with the use of different fluorescent labels, can be multiplexed in PCR. The information content per locus is bigger in SSRs compared to dominant markers because homoand heterozygotes are detected. however, in polyploid species interpretation of exact marker genotypes is not straightforward and SSR alleles are usually analysed as presence/absence markers. This study is part of a wider nordic collaborative research project, where the variation of nordGen timothy collection was evaluated both on phenotypic and genotypic level. here we report the results of the assessment of genetic diversity using SSR markers. Our aim was to study whether geographical location (vegetation zone, latitude, longitude, altitude) affects the level of genetic diversity. in addition, we studied if genetic markers could find a population structure in the nordic timothy material and thus help finding heterotic groups among the collection of timothy to be used in variety breeding. MaTeRiaL anD MeThODS Plant material eighty-eight timothy accessions from nordic countries (Table 1, Fig. 1) were selected from nordGen collection based on geographical distribution and previous phenotyping data to represent as wide geographical and trait variation as possible, and 15–20 randomly selected individuals per accession were analysed. Most of the accessions, namely 59, were classified as wild accessions, 17 as landraces and 11 as varieties or breeders material. accessions were divided to six groups according to country of origin: norway (26), Finland (25), Sweden (25), Denmark (10), iceland (2), and exotic (8) including all origins outside nordic countries. exotic accessions were obtained from different genebank collections. accessions with known geographical coordinates were divided to six vegetation zones (mo e n 1999): 1  nemoral (11), 2  boreonemoral (15), 3  southern boreal (14), 4  middle boreal (22), 5  northern boreal (11), and 6  alpine (2). Dnas were extracted using the method of ti n k e r et al. (1993) with the following modifications: lyophilised leaves were crushed with a FastPrep FP120 Cell Disrupter (BiO 101, Thermo Savant, Waltham, Ma, uSa), in 1 ml CTaB (hexadecyltrimethyl-ammonium bromide) buffer supplied with 70 u of ribonuclease a (Omega Bio-tek, norcross, Ga, uSa) and 0.05 mg of proteinase K (Finnzymes, espoo, Finland). extractions were first done with phenol/chloroform/isoamyl alcohol (25:24:1) and then with chloroform. Dna concentrations were measured using the GeneQuant ii Rna/Dna Calculator (Pharmacia Biotech Ltd., Cambridge, uK). SSR analyses SSRs developed for timothy (ca i et al. 2003) were used for assessing diversity in the selected accessions. at the beginning of the study, 35 timothy SSRs were selected using the following criteria: strong amplification (ca i et al. 2003), preferably SSRs with trinucleotide repeats (SSRs containing trinucleotide or higher order repeats have less stuttering: Ho l t o n 2001), and some SSRs which have been localised on one position on the diploid timothy map (ca i et al. 2009). The SSRs were optimised and tested for their polymorphism, multiplexing possibilities, and easiness of interpretation. One primer of each primer pair was labelled with a fluorescent dye, FaM (5-carboxyfluorescein), heX (hexachloro6-carboxyfluorescein) or TeT (6-carboxytetrachlorofluorescein) to enable separation and visualisation of amplification products with a MegaBaCe 500 Sequencer (Ge healthcare, Buckinghamshire, uK) using MegaBaCe eT400-R Size Standard. Thirteen best SSRs (Table 2) were selected for final analyses and were amplified using two different PCR programs in a PTC-220 Dna engine Dyad Peltier Thermal Cycler (MJ Research, Waltham, Ma, uSa) or a Bio-Rad Dna engine Tetrad 2 Thermal Cycler (Bio-Rad, hercules, Ca, uSa). The first five SSRs in Table 2 were amplified with five cycles of 15 s at 94°C, 15 s at 65°C, and 30 s at 72°C, followed by 30 similar cycles except that the annealing temperature was 60°C. The program started with an initial denaturation step of 5 min at 94°C and was 116 P. Tanhuanpää & O. Manninen hereditas 149 (2012) Table 1. Ninety-six accessions of Phleum pratense ssp. pratense analysed in the study with 499 SSR markers (each SSR allele treated as a separate marker). number code accession no. Genebank name Country Cultivar type1Latitude Longitude altitude2 Veg. zone3 no. of ind. Observed no. of markers no. of private markers aa4ai5PWD6 1 nGB10828 nordgen Va88108 Denmark W 1 19 120 0 113.1 28.6 31.0 2 nGB10829 nordgen Va88112 Denmark W 1 19 141 2 131.9 32.9 38.7 3 nGB10830 nordgen Va88119 Denmark W 1 19 124 0 116.7 28.8 30.7 4 nGB10831 nordgen hF88266 Denmark W 1 19 148 2 137.9 28.4 32.5 5 nGB15461 nordgen Vildbjerg aC0103 Denmark W 56°11′ n 8°49′58″ e 40 1 19 188 0 171.5 33.3 44.0 6 nGB16650 nordgen ejsing Denmark W 56°31′21″ n 8°47′11″ e 1 19 142 0 132.9 31.2 36.2 7 nGB1672 nordgen BiLBO Denmark CV 19 162 1 149.7 31.7 40.1 8 nGB1675 nordgen POTa Denmark CV 20 119 0 110.1 30.0 32.5 9 nGB4053 nordgen SR SaLTuM Mh0202 Denmark W 57°14′ n 9°46′ e 6 1 19 145 0 135.2 30.6 36.2 10 nGB4548 nordgen nR FaRuP Mh0202 Denmark W 55°21′ n 8°41′ e 2 1 20 154 0 139.4 30.3 35.1 11 nGB132 nordgen LiPinLahTi Me0901 SeP a Finland L 63°28′ n 29°18′ e 100 4 19 186 0 167.5 31.1 39.1 12 nGB9285 nordgen OTTO Finland CV 18 188 1 175.9 35.9 46.8 13 nGB14394 nordgen KäRKÖLä hM0102 Finland W 60°55′16″ n 25°17′27″ e 3 19 175 0 162.2 32.7 42.6 14 nGB14399 nordgen MahLaMäKi Mh0104 Finland W 61°22′00″ n 22°56′19″ e 70 3 20 179 0 162.5 32.9 41.7 15 nGB14403 nordgen näReKuMPu Mh0103 Finland L 61°57′14″ n 28°26′05″ e 3 20 203 2 180.8 35.0 46.3 16 nGB14404 nordgen PaaTTinen Mh0201 Finland L 60°35′11″ n 22°22′16″ e 2 20 176 0 158.4 33.6 40.5 17 nGB14415 nordgen huOLiLa Mh0204 Finland L 60°41′00″ n 21°45′06″ e 2 20 141 0 128.0 30.6 35.0 18 nGB14417 nordgen MeDVaSTÖ Mh0101 Finland W 60°06′02″ n 24°37′36″ e 25 2 18 121 0 115.2 28.4 32.5 19 nGB14419 nordgen KiiKaOJa Mh0201 Finland W 61°30′33″ n 22°32′10″ e 3 20 194 1 174.4 32.3 42.3 20 nGB747 nordgen nuVVuS aK0401 Finland W 69°50′ n 26°19′ e 160 6 20 144 1 131.4 30.3 34.7 21 nGB748 nordgen uTSJOKi aK0602 Finland W 69°55′ n 27°03′ e 70 6 20 169 1 153.6 33.7 39.8 22 nGB754 nordgen haLOSenRanTa eh0101 Finland W 66°41′ n 27°30′ e 145 4 19 183 0 168.4 34.7 43.9 23 nGB757 nordgen PeKKaLa eh0703 Finland W 66°21′ n 26°52′ e 145 4 19 189 0 173.5 36.2 44.7 24 nGB1095 nordgen LaiTaSaaRi Me0201 Finland L 64°51′ n 25°56′ e 4 20 192 1 170.2 34.2 43.1 25 nGB1096 nordgen TuOMiOJa Me0201 Finland L 64°36′ n 25°02′ e 4 18 191 1 177.5 33.6 44.0 26 nGB1107 nordgen JYRinKi Me0101 Finland L 63°55′ n 24°26′ e 4 19 188 0 171.5 33.6 43.5 27 nGB1111 nordgen MäLäSKä Me0101 Finland L 64°24′ n 26°19′ e 4 20 194 1 172.7 33.6 44.3 28 nGB151 nordgen VaRiSLahTi Me0102 Finland L 62°42′ n 28°42′ e 115 3 18 190 0 176.5 33.9 41.1 29 nGB1115 nordgen KiLPau Me0101 Finland L 64°20′ n 25°07′e4 19 194 0 178.8 35.1 44.6 30 nGB1119 nordgen KaTeRMa Me0401 Finland L 64°03′n 29°09′ e 4 19 189 0 170.9 31.9 40.5 (Continued) hereditas 149 (2012) SSR diversity in Nordic timothy 117 Table 1. (Continued). number code accession no. Genebank name Country Cultivar type1Latitude Longitude altitude2 Veg. zone3 no. of ind. Observed no. of markers no. of private markers aa4ai5PWD6 31 nGB1122 nordgen nääDänMaa Me0202 Finland L 62°30′ n 28°14′ e 3 19 188 0 174.7 35.1 45.5 32 nGB2791 nordgen nORRGÅRD aP0101 Finland L 63°32′ n 22°31′ e 3 19 186 2 169.8 33.3 42.5 33 nGB2798 nordgen LÅnGÅMinne aP0201 Finland L 62°54′ n 21°43′ e 3 20 186 1 164.9 33.6 41.2 34 nGB2836 nordgen LanKaMaa aP0202 Finland L 62°23′ n 26°14′ e 3 19 184 0 166.8 32.1 41.5 35 nGB4066 nordgen TaMMiSTO Finland CV 19 180 1 164.5 32.7 38.6 36 nGB4140 nordgen KORPa iceland L 19 193 1 176.2 33.3 43.9 37 nGB4141 nordgen aDDa iceland CV 19 117 0 111.2 31.3 34.6 38 nGB7557 nordgen KLeVeLanD 01-5-43-2 norway W 63°13′ n 11°03′ e 300 4 20 178 0 160.9 33.7 41.7 39 nGB7559 nordgen SVenDGÅRD 01-5-43-4 norway W 63°13′ n 11°03′ e 100 4 15 152 0 152.0 33.9 42.9 40 nGB7573 nordgen ØVRe heRSTaD 01-5-44-3 norway W 63°52′ n 11°16′ e 200 3 20 164 1 146.6 32.2 39.2 41 nGB7577 nordgen VOLDen 01-5-44-8 norway W 63°22′ n 9°56′ e 20 3 17 179 2 170.4 33.8 42.8 42 nGB7592 nordgen SKJØLSViK 01-5-46-5 norway W 62°57′ n 7°48′ e 20 3 19 182 0 168.4 35.3 44.9 43 nGB7597 nordgen MÅna 01-6-48-13 norway W 62°06′ n 10°37′ e 600 5 20 155 0 141.4 32.8 40.3 44 nGB7709 nordgen nORDSKOT 01-2-13-6 norway W 67°50′ n 14°50′ e 15 4 19 186 1 169.9 32.6 42.7 45 nGB10785 nordgen SanDBu 01-6-49-4 norway W 61°52′ n 9°07′ e 420 5 19 95 0 86.8 32.6 28.9 46 nGB13647 nordgen LeVeLD norway P 18 106 1 102.6 32.6 34.1 47 nGB17194 nordgen ifjord 1-1-2-2 norway W 70°27′42″ n 27°06′30″ e 5 20 166 0 149.5 31.4 37.2 48 nGB17198 nordgen Karasjok 1-1-3-2 norway W 69°28′31″ n 25°30′23″ e 5 18 179 1 166.4 32.4 42.5 49 nGB2169 nordgen BODin norway CV 19 167 0 153.4 30.3 38.8 50 nGB2170 nordgen VÅTi7702 norway B 17 119 0 115.0 34.0 35.2 51 nGB2180 nordgen GRinDSTaD norway CV 18 185 1 169.2 31.2 39.5 52 nGB2917 nordgen KLOMSeT 01-6-54-6 norway W 59°28′ n 8°36′ e 130 3 19 164 0 152.3 35.1 42.5 53 nGB2918 nordgen huSeTeR 01-9-70-1 norway W 59°40′ n 11°23′ e 135 2 20 164 0 150.2 32.1 39.0 54 nGB2922 nordgen SØRhuS 01-6-48-2 norway W 62°06′ n 10°40′ e 520 5 18 178 0 166.5 34.4 43.3 55 nGB2927 nordgen ØSTeRØYa 01-9-71-1 norway W 59°05′ n 10°12′ e 10 2 20 182 1 164.0 34.2 42.7 56 nGB2930 nordgen GRauTeKnaPP 01-6-54-7 norway W 59°10′ n 8°49′ e 95 2 20 194 0 174.6 34.4 44.7 57 nGB4226 nordgen haTLeSTaD 01-7-56-3 norway W 61°21′ n 6°06′ e 5 17 158 0 151.6 34.9 42.2 58 nGB4227 nordgen hÅRKLau 01-7-56-4 norway W 61°25′ n 6°15′ e 5 19 155 1 146.8 35.6 43.6 59 nGB4231 nordgen GJeRDÅKeR 01-7-58-1 norway W 60°40′ n 6°30′ e 100 5 19 156 1 143.6 31.3 38.4 118 P. Tanhuanpää & O. Manninen hereditas 149 (2012) 60 nGB4508 nordgen eneBO 01-6-48-5 norway W 61°15′ n 12°19′ e 550 4 20 161 0 149.2 33.2 38.9 61 nGB4523 nordgen FOSS 01-9-71-3 norway W 59°25′ n 11°21′ e 100 2 19 188 1 170.8 35.1 45.3 62 nGB7548 nordgen naMSVaTn 01-5-40-1 norway W 64°58′ n 13°34′ e 500 5 18 150 0 140.8 29.8 35.4 63 nGB7551 nordgen SOLeM 01-5-42-1 norway W 63°45′ n 9°45′ e 20 3 17 168 0 161.2 34.5 44.2 64 nGB722 nordgen KuOSSenJaRKa JP0404 Sweden W 66°41′ n 19°45′ e 260 5 19 115 0 108.8 29.5 32.7 65 nGB728 nordgen PJeSKeR Ph0405 Sweden W 65°32′ n 19°42′ e 350 4 18 182 0 171.2 33.9 43.4 66 nGB11428 nordgen JOnaThan Sweden CV 16 139 0 138.0 34.2 39.8 67 nGB11430 nordgen aRGuS Sweden CV 19 154 1 141.0 31.7 37.0 68 nGB13226 nordgen RaGnaR Sweden CV 17 111 0 107.9 32.8 34.3 69 nGB14224 nordgen SÖnDRaRP iB0101 Sweden W 57°36′02″ n 14°26′39″ e 268 2 20 175 1 158.4 33.6 42.2 70 nGB14236 nordgen LÖVhuLT iB0103 Sweden W 57°39′11″ n 14°45′23″ e 269 2 19 170 1 156.6 31.2 40.0 71 nGB731 nordgen RÖRMYRBeRG JP0204 Sweden W 64°40′ n 19°09′ e 350 4 20 182 2 163.2 31.9 41.1 72 nGB16958 nordgen LYa LJunGheD FO0201 Sweden W 56°24′13″ n 12°53′42″ e 175 1 18 152 0 142.3 30.7 37.7 73 nGB16975 nordgen nORRa KYLSäTeR FO0103 Sweden W 58°36′04″ n 11°59′51″ e 117 2 19 176 1 161.8 33.5 42.2 74 nGB733 nordgen SÖDRa SunDeRBYn Me0101 Sweden L 65°40′ n 21°52′ e 20 4 15 185 1 185.0 36.2 46.6 75 nGB16977 nordgen RYR, STORa BeRGeT FO0101 Sweden W 58°48′16″ n 12°29’25″ e 47 2 19 179 1 165.0 33.5 43.1 76 nGB16981 nordgen BRäCKeTORP FO0501 Sweden W 59°02′48″ n 12°29′26″e166 2 16 153 0 149.8 33.8 41.7 77 nGB17061 nordgen STORa ROThuLT haJ0201 Sweden W 57°55′53″ n 15°40′33″ e 180 2 20 187 0 168.5 33.4 43.3 78 nGB1306 nordgen BRaTTÅKeR GB0101 Sweden W 64°18′ n 19°33′ e 344 4 18 184 2 173.6 32.9 41.4 79 nGB1310 nordgen STORhäGGSJÖ GB0104 Sweden W 63°59′ n 20°00′ e 100 4 19 138 0 130.9 34.6 41.9 80 nGB1320 nordgen SKaRPMYRBeRG PR0601 Sweden W 64°37′ n 16°18′ e 400 5 20 194 0 173.4 33.5 42.5 81 nGB1327 nordgen haMMaRn PR0401 Sweden W 63°48′n 20°29′ e 10 4 19 162 0 151.4 33.6 41.2 82 nGB1330 nordgen äLGSJÖ Sh0302 Sweden W 64°13′ n 17°29′ e 390 4 20 185 0 166.0 34.7 43.3 83 nGB1331 nordgen VäSTanSJÖ Sh0102 Sweden W 63°45′ n 18°59′ e 170 4 19 170 0 158.4 32.8 42.1 84 nGB1332 nordgen KLuBBSJÖ Sh0301 Sweden W 63°51′ n 19°07′ e 250 4 18 185 1 172.4 34.0 43.1 85 nGB1537 nordgen eSKeLheM TL0104 Sweden W 57°29′ n 18°10′ e 2 18 139 1 131.7 30.7 36.1 86 nGB2530 nordgen RäMne GJ0301 Sweden W 59°00′ n 12°04′ e 110 2 19 165 0 152.8 31.9 39.6 87 nGB4349 nordgen BeneSTaD JK1506 Sweden W 55°31′ n 13°54′ e 40 1 20 171 1 153.4 32.4 40.6 88 nGB4350 nordgen BOaRP SB2106 Sweden W 55°56′ n 13°47′ e 180 1 18 175 1 163.9 34.3 42.6 89 Pi381926 GRin France P 19 131 0 122.3 31.1 34.6 90 Pi406317 GRin Russia P 19 165 1 151.7 31.8 38.5 (Continued) hereditas 149 (2012) SSR diversity in Nordic timothy 119 Table 1. (Continued). number code accession no. Genebank name Country Cultivar type1Latitude Longitude altitude2 Veg. zone3 no. of ind. Observed no. of markers no. of private markers aa4ai5PWD6 91 ihaR151908 ihaR Germany P 19 150 0 137.4 31.4 34.8 92 Pi210426 GRin Greece P 18 146 2 138.9 31.5 38.3 93 Pi325461 GRin Russia P 19 170 8 157.8 31.7 39.8 94 Pi204480 GRin Turkey P 19 158 3 144.3 31.6 37.6 95 14G2400116 RiCP Czech Republic P 19 186 2 170.5 34.2 44.0 96 RCaT040682 RCaT hungary W 20 157 6 143.5 30.5 39.1 1CV  advanced cultivar, L  traditional cultivar, landrace, B  breeding, research material, genetic stock, P  pending, unknown cultivar type, W  wild population, weedy. 2meters above sea level. 3vegetation zones, according to Moen 1999. 4corrected number of all markers in each accession. 5mean number of all alleles observed in each individual. 6mean number of pairwise differences (PWD) (euclidean distances) between individuals in each accession. Fig. 1. Geographic location of 71 timothy accessions. number codes are presented in Table 1. followed by a final extension step of 7 min at 65°C. The following eight SSRs in Table 2 were amplified with the PCR program described in ca i et al. (2003). The PCR amplification reactions in 10 ml contained 0.25 u of FiRePol Dna polymerase i (Solis BioDyne Ou, Tartu, estonia), the buffer B with 2.5 mM MgCl2 supplied by the enzyme manufacturer, 100 mM each dnTP, 10 ng of Dna, and 125–500 nM each primer. Suitable SSR combinations were found with the FastPCR software (ka l e n d a r et al. 2009), and the 13 SSRs were multiplexed in five PCR (those amplified together are grouped in Table 2). Data analyses allele phenotypes of the plants were visually scored using a binary code (1/0) for the presence or absence of allele peaks without knowing the doses of the alleles. When calculating genetic distances each SSR allele was thus treated as a separate marker locus. however, for the POPDiST program (to m i u k et al. 2009), the allele phenotype was recorded locuswise i.e. the allelic content of an individual at each of the 13 SSR loci was described. Genetic diversity of an accession was described in three different ways: 1) corrected number of all alleles ( markers) in each accession (aa), where the observed number of alleles was corrected to a sample size n  15 120 P. Tanhuanpää & O. Manninen hereditas 149 (2012) Table 2. SSRs used in the diversity analysis of timothy accessions. SSR Repeat motif Repeat class Fluorescent label exp. size (bp)1 allele size range (bp) no. of alleles no. of alleles/ accession Mean no. of alleles/ individual Most common allele2 no. of private alleles3 Miss. inf. (%) Linkage group4 a03a07 (TG)33 perfect heX 140 94–207 50 4–17 2.10 0.64 5 2.5 unknown C02C08 (aaG)13 perfect FaM 241 212–278 24 4–19 2.69 0.61 0 2.7 LG6 C01B11 (TTC)16 perfect TeT 194 153–249 31 9–21 3.91 0.76 5 0.8 unknown C02h01 (TTC)17 perfect FaM 146 95–201 53 8–26 2.93 0.44 8 2.4 unknown D01e04 (Caa)8(Taa)10 compound TeT 158 95–288 56 8–23 2.82 0.67 11 1.9 LG1 B03F07 (TC)14 perfect heX 130 112–172 32 2–18 1.44 0.29 2 15.1 unknown C01e11 (TTC)11 perfect FaM 123 98–142 15 2–10 1.12 0.53 0 18.8 LG5 a09h08 (TG)16 perfect TeT 255 230–267 15 4–9 3.55 0.86 5 1.1 unknown D01G10 (TGa)7(CGa)4舰 … (TGa)16 compound and interrupted FaM 231 208–336 74 7–24 2.51 0.44 9 2.4 LG2 B03a09 (Ga)19 perfect TeT 226 193–247 40 8–22 2.62 0.28 3 0.3 LG4 a03e06 (TTG)28 perfect heX 238 166–290 49 3–22 2.46 0.91 10 0.5 unknown D01h08 (aaT)13 perfect FaM 147 116–170 20 4–15 1.88 0.28 1 7.1 LG6 a10a10 (Ca)31 perfect TeT 232 161–243 40 6–22 2.68 0.68 7 3.2 LG3 1according to Cai et al. 2003. 2Occurence of the most common allele. 3alleles present in only one accession. 4Refers to the diploid timothy map (Cai et al. 2009). hereditas 149 (2012) SSR diversity in Nordic timothy 121 different distance indices: euclidean distance from arlequin, nei’s distance (ne i 1972) from Genalex, and Tomiuk and Loeschke distance from Popdist) with geographic distance (km) was tested using a Mantel-test (ma n t e l 1967) in the software Genalex. Mantel test was also used to compare different genetic distance indices. ReSuLTS Diversity in SSR loci Thirteen SSRs (Table 2) were selected to assess genetic diversity in timothy accessions. Of these, five included a dinucleotide motif and eight a trinucleotide motif, of which two were compound ones. in some cases the allelic series (allele sizes fit the assumption of increments of two or three nucleotides) was perfect (C02C08, B03F07, C01e11, D01h08) but usually some alleles were missing. Generally, the allele sizes followed neatly the increment of two or three bases. however, a few extra alleles that did not fit the allele series existed in all SSRs except C01e11. in some SSRs (C02h01, D01G10, B03a09, a03e06), there seemed to be another allele series differing from the common one with one base pair (Fig. 2). The existence of this other series most probably was the outcome of an indel mutation in the SSR amplicon. as a consequence, the size of all the alleles arisen thereafter had shifted with one base pair. The 13 selected SSRs amplified a total of 499 polymorphic alleles, the number of alleles per SSR locus varying from 15 (C01e11 and a09h08) to 74 (D01G10) (Table 2). The average repeat length of alleles in SSR loci correlated positively (r  0.69) with the total number of alleles in the loci. Most of the alleles were quite rare, ca. 40% occurred in not more than 1% of all the individuals, or with thousand times of resampling without replacement, 2) mean number of all alleles observed in each individual (ai) and 3) mean number of pairwise differences (PWD) (euclidean distances) between individuals in each accession, which was counted with the program aRLeQuin ver. 2.000 (Sc H n e i d e r et al. 2000). Differences in the level of diversity between different groups like countries, vegetation zones (mo e n 1999), or cultivar type were analysed by anOVa Proc GLM (SaS enterprise Guide 4.3). Correlations were counted between diversity and latitude, longitude and altitude (Proc CORR, SaS enterprise Guide 4.3) for those accessions where information of collection site map coordinates or elevation was available. Genetic divergence between accessions or groups was analysed by analysis of molecular variance (aMOVa) (ex c o f f i e r et al. 1992) using the program Genalex 6.4 (Pe a k a l l and Sm o u S e 2006). Significance of the results was tested by permuting the Dna marker data 999 times. a neighbor-joining (nJ, Sa i t o u and ne i 1987) dendrogram was constructed using the program MeGa ver. 4 (ta m u r a et al. 2007). The genetic distances between accessions for the dendrogram were calculated with the program POPDiST (to m i u k et al. 2009), where the estimation of genetic distances is based on grouping of allele phenotypes (distance measure of to m i u k et al. 1998), in which case the degree of ploidy is of no importance. as far as we know, POPDiST is the only program for diversity studies that can handle codominant markers in polyploids. Principal coordinates analysis (PCa) based on nei’s genetic distances between accessions was performed using the software Genalex 6.4 (Pe a k a l l and Sm o u S e 2006). Map coordinates were availabe for 71 accessions. Correlation between genetic distance (described with Allele size (bp) 192.8 201.2 203.1 204.1 205.0 206.0 206.9 207.9 209.0 209.7 210.7 211.7 212.7 213.7 214.5 215.6 216.4 217.5 218.4 219.4 220.2 221.3 222.2 223.2 224.1 225.1 226.3 227.1 228.1 229.0 230.9 232.9 233.7 234.8 236.8 238.8 240.6 242.5 244.2 246.5 No. of individuals 0 100 200 300 400 500 series 2 series 1 217.5 series 1 series 2 Fig. 2. an example of the two allelic series in SSR locus B03a09. 122 P. Tanhuanpää & O. Manninen hereditas 149 (2012) were more diverse than Danish accessions when aa was compared (Tukey’s test, p  0.05). For PWD, Denmark showed less diversity than Finland, norway or Sweden (Tukey’s test, p  0.05). Danish accessions were also less diverse than Finnish, Swedish, norwegian or exotic based on ai. accessions originating from southern boreal or middle boreal vegetation zone were more diverse than those from nemoral or alpine vegetation zone when aa or PWD were compared (Tukey’s test, p  0.05). Vegetation zones explained 33% of the variation in diversity levels between timothy accessions (Table 3). Significant differences between vegetation zones were also observed for ai but they were very small and explained only a minor fraction of variation between individuals (Table 3). accessions with the cultivar type L, meaning landrace or traditional, locally cultivated accession, had a higher aa when compared to cultivars or wild accessions. Landraces also were more diverse than cultivars based on PWD or ai (Tukey’s test, p  0.05). no correlation was observed between latitude or altitude and the diversity indices. however, a weak correlation was observed between longitude and aa (r  0.29, p  0.013). Genetic divergence between accessions and groups aMOVa was performed in order to divide the total genetic variation into three components: variation within accessions, among accessions and among groups. according to aMOVa analysis, most of the variation (94%) in the studied timothy material was due to variation within accessions and only 5% was between accessions and 1% between countries (Table 5). no genetic divergence was observed between vegetation zones or cultivar types (aMOVa, p  0.05). no clear clustering of accessions based on countries or any other grouping was seen in either PCa (Fig. 3) or nJ dendrogram (Supplementary material appendix a1 Fig. a1). in PCa, the first two axes explained 42.4% of the variation among the 96 accessions. Most of the accessions clustered together apart from a couple of exceptions. Genetic distance matrices counted with different ways 10% of the accessions. This usually caused a high occurence of the most common allele in a SSR locus (Table 2). The mean number of alleles in an individual varied from 1.12 to 3.91 depending on the SSR locus. Very few individuals (1.2%) carried six alleles in any SSR locus. The most heterozygous SSRs were C01B11 and a09h08: 66% and 54% of the individuals in the study, respectively, contained four alleles or more in these loci. On the other hand, the alleles in the SSR loci B03F07 and C01e11 often occurred alone (in 48% and 67% of the individuals, respectively). This is perhaps spurious, due to the existence of null alleles, which is also reflected by the large number of apparent missing information in these loci (15.1% and 18.8%, respectively, Table 2). The five SSRs with a dinucleotide repeat amplified an average of 35.4 alleles compared to 40.3 alleles amplified by the eight SSRs with a trinucleotide repeat (Table 2). The average observed number of alleles/accession was 11.7 and 13.4, and of private alleles 3.8 and 4.9 in SSRs with diand trinucleotide repeats, respectively (results not shown). however, none of these differences were statistically significant (t-test, p  0.05). Genetic diversity within accessions Taking all the 499 SSR alleles ( individual markers) together, the observed number in each accession ranged from 95 (nGB10785) to 203 (nGB14403) (Table 1). Most of the markers were polymorphic i.e. very few existed in all individuals of an accession. The number of private alleles i.e. alleles that did not exist in any other accession was generally low but accessions Pi325461 (from Russia) and RCaT040682 (from hungary) included eight and six private alleles, respectively. ai ranged from 28.4 (nGB14417) to 36.2 (nGB733 and nGB757) (Table 1). Genetic diversity within accessions measured as PWD varied from 28.9 (nGB10785) to 46.8 (nGB9285) (Table 1). Levels of diversity were found to be significantly different between countries, vegetation zones and different cultivar types (anOVa, Table 3, 4). Finnish accessions Table 3. ANOVA table showing F-values, significance levels P and R2 for comparisons of different groups for their levels of SSR diversity. Here AI represents the number of alleles on individual level. Total number of alleles (aa) number of pairwise differences (PWD) number of alleles per individual (ai) Diversity index df F p R2F p R2F p R2 Grouping accession 95 3.52 < 0.001 0.16 Country of origin 5 4.15 0.002 0.19 4.40 0.001 0.20 14.15 < 0.001 0.04 Vegetation zone 5 6.86 < 0.001 0.33 6.78 < 0.001 0.33 13.29 < 0.001 0.04 Cultivar type 2 8.58 < 0.001 0.17 4.46 0.014 0.10 6.05 0.002 0.01