Flora 206 (2011) 151–157

Contents lists available at ScienceDirect Flora

journal homepage: www.elsevier.de/flora

Genetic diversity of a relict species, sibirica (L.) Cass. ()

Anna Smídovᡠa,b,∗, Zuzana Münzbergová a,c, Ivana Plackovᡠa a Institute of Botany, Academy of Sciences of the Czech Republic, Zámek 1, 252 43 Pruhonice,˚ Czech Republic b Department of Botany, Faculty of Science, Palacky University, Slechtitelˇ u˚ 11, 783 71 Olomouc, Czech Republic c Department of Botany, Faculty of Science, Charles University, Benátská 2, 128 01 Prague 2, Czech Republic article info abstract

Article history: Rare plant species can be divided into naturally, ‘old rare’ species and anthropogenically, ‘new rare’ Received 12 November 2009 species. Many recent studies explored genetic diversity of ‘new rare’ species. Less is, however, known Accepted 12 March 2010 about genetic diversity of ‘old rare’ species. We examined isozyme genetic variability of 20 populations of an ‘old rare’ plant species, (Asteraceae) in the Czech and Slovak Republic. It is a long- Keywords: lived perennial herb with mixed-mating breeding system, widely distributed from East Asia to European Allozyme Russia, with few isolated relict populations in the remaining part of Europe. Genetic diversity The results showed high genetic diversity within populations (80.8%) and a low level of genetic dif- Ligularia sibirica Old rare species ferentiation (FST = 0.179). Genetic distance between populations correlated significantly with geographic Population size distance. There was also a significant positive correlation between genetic diversity and population size. Postglacial migration This is probably caused by destruction of habitats in last centuries and subsequent decrease of population size. Patterns of genetic diversity suggest that the recent distribution is a result of stepwise postglacial migration of the species and subsequent natural fragmentation. We conclude that L. sibirica populations preserve high levels of genetic diversity and are not yet threatened by genetic factors. However, this may change if changes in habitat conditions continue. © 2010 Elsevier GmbH. All rights reserved.

Introduction such a ‘new rare’ species may still reflect genetic diversity in the past (Booy et al., 2000). Genetic consequences of habitat fragmentation are a topic of In contrast to recently declining ‘new rare’ species, genetic many recent studies (reviewed by Honnay and Jacquemyn, 2007; diversity of so-called ‘old rare’ species is expected to show long Leimu et al., 2006; Young et al., 1996). It has been shown that term consequences of habitat fragmentation (Reisch, 2001; Van small plant populations often face increased inbreeding depres- Rossum and Prentice, 2004). ‘Old rare’ species are species that are sion and reduced genetic variation due to genetic drift (Frankham naturally rare in a specific area, occurring in small, isolated popu- et al., 2006; Young et al., 1996). This often leads to decreased fit- lations (sensu Huenneke, 1991). To correctly understand the status ness of individuals as well as decreased ability to adapt to changes of such a species, we need to investigate the history of their popu- in environment (Booy et al., 2000; Leimu et al., 2006). lations using evidence from population genetics as well as from Most of the studies on effects of habitat fragmentation are deal- biogeography, ecology and vegetation science (Pott, 1995). ‘Old ing with processes that occurred in the last decades or centuries due rare’ species include narrow endemics (Pegtel, 1998) and relicts, to human activities (Gonzales and Hamrick, 2005) such as changes such as glacial and postglacial relicts (Hofhanzlová and Fér, 2009; in land use (e.g. Jacquemyn et al., 2003; Robinson and Sutherland, Pegtel, 1998; Rasmussen and Kollmann, 2004). 2002; Stoate et al., 2001). These studies are mainly focusing on Studies of genetic diversity of endemic plant species have shown anthropogenically rare (‘new rare’) species, i.e. species that were less than half of the genetic diversity of widespread species, low formerly much more common in a particular area, and popula- percentage of polymorphic loci and mean number of alleles per tions became smaller, less abundant, and more isolated because polymorphic locus (Hamrick and Godt, 1989). In contrast, studies of human influence (sensu Huenneke, 1991). Genetic diversity of on European glacial relict species with arctic–alpine distribution range such as Biscutella laevigata (Dannemann, 2000), Saxifraga aizoides (Lutz et al., 2000) and Saxifraga paniculata (Reisch, 2001) have shown much higher genetic diversity, percentage of poly- morphic loci and number of alleles than endemics or species with ∗ Corresponding author at: Institute of Botany, Academy of Sciences of the Czech narrow distribution range (Hamrick and Godt, 1989). Republic, Zámek 1, 252 43 Pruhonice,˚ Czech Republic. E-mail addresses: [email protected] (A. Smídová),ˇ The Central European relict plant species thus clearly repre- [email protected] (Z. Münzbergová), [email protected] (I. Placková).ˇ sent a specific group of species with unique patterns of genetic

0367-2530/$ – see front matter © 2010 Elsevier GmbH. All rights reserved. doi:10.1016/j.flora.2010.03.003 152 A. Smídovᡠet al. / Flora 206 (2011) 151–157 diversity. The glacial relicts with arctic–alpine distribution range in mosaics of these communities (Procházka and Pivnicková,ˇ have received considerable attention in the past (e.g. Bauert et 1999). al., 1998; Dannemann, 2000; Lutz et al., 2000; Reisch, 2001). The species has a wide Eurosiberian distribution range. The main Another group of relict species, i.e. species with a large con- continuous distribution range is from East Asia to Southern Siberia tinuous distribution area from East Asia to the western part of and to the European part of Russia. In Europe, there are few dis- Russia and with the western edge of distribution range in Europe, junct populations in Estonia, Latvia, Poland, Hungary, Romania, are still largely ignored. In Europe such species occur in few Croatia, Bulgaria, the Slovak Republic, the Czech Republic, Austria, isolated localities (Meusel and Jäger, 1992). To our knowledge, and France (Mattauch, 1936; Meusel and Jäger, 1992). Localities in genetic variability of relict west range-peripheral populations was these countries are rather distant from the continuous distribution studied only in Angelica palustris (Dittbrenner et al., 2005), Iris range of the species, originated most probably in the early post- aphylla (Wróblewska, 2008; Wróblewska et al., 2003), Iris sibirica glacial period and thus represent rare remnants of former more (Kostrakiewicz and Wróblewska, 2008) and Stipa capillata (Hensen extensive distribution. Therefore, L. sibirica is considered to be a et al., 2009; Krzakowa and Michalak, 2007). These studies have postglacial relict (Hendrych, 2003). The species has currently 10 shown variable results. While Dittbrenner et al. (2005) found a sim- localities with more than 50 individuals in the Czech Republic and ilar level of genetic variability as revealed for most glacial relict about 15 localities in the Slovak Republic. species, Kostrakiewicz and Wróblewska (2008) found very low L. sibirica is classified as a ‘critically endangered’ species in the level of genetic variability, similar to endemic species. The results Czech Republic (Procházka, 2001) and as a ‘vulnerable’ species in of the latter study may be related to the extensive clonal growth in the Slovak Republic (Feráková et al., 2001). This species is also pro- Iris sibirica. tected by EU Habitat Directive, Annex II of the Council of European In the present study, we use isozyme electrophoresis to inves- Communities (1992). For manipulation of this species and enter- tigate genetic variability and genetic structure of the relict plant ing its localities, we obtained permission from the Ministries of species, Ligularia sibirica in the Czech and Slovak Republic. L. sibir- Environment of the Czech and Slovak Republic. ica has a wide continuous Eurosiberian distribution from East Asia to the European part of Russia. In central and western Europe this species has only few isolated localities (Meusel and Jäger, 1992) Study populations and is listed among the species of European importance (Council of European Communities, 1992). Due to recent habitat change, For our study, we sampled all 10 populations in the Czech the species began to decline at some localities (Hendrych, 2003). Republic and 10 selected populations in the Slovak Republic cov- We can therefore expect that genetic diversity in these populations ering most of the populations containing more than 50 adult might be reduced as in the new rare species. individuals (Fig. 1). Populations were considered as separate, when We asked the following questions: (1) What is the present-day located at least 100 m apart from each other. isoenzyme variability of L. sibirica and its distribution within and The populations in the Czech Republic are located in three among populations, and among geographical regions? (2) What regions. The first one (containing 5 populations) is in north- ◦   ◦   is the population genetic differentiation? (3) Is there a correla- ern Bohemia near the village Jestrebíˇ (50 36 8 –50 36 22 N, ◦   ◦   tion between population size and genetic variability? We aimed 14 36 51 –14 37 21 E). The second one (containing 4 pop- to interpret the results in relation to biology of L. sibirica and dis- ulations) is in central Bohemia near the village Reckovˇ ◦   ◦   ◦   ◦   cuss how the present pattern of genetic variability is affected by (50 29 38 –50 30 3 N, 14 51 56 –14 54 40 E). The last pop- the possible history of the populations. ulation is in southern Bohemia in Posumavíˇ near a fishpond ‘Olsina’ˇ (48◦4734N, 14◦67E). Eight of the sampled populations in the Slovak Republic are in a large area in NP Slovensky´ raj ◦   ◦   ◦   ◦   Materials and methods (48 52 32 –48 55 40 N, 20 13 50 –20 21 44 E). One isolated population is in NP Nízke Tatry near the village Liptovská Teplickaˇ ◦   ◦   Study species (48 58 2 N, 20 5 23 E) and one rather distant is in Salvatorské lúky approximately 25 km west of Presovˇ (49◦245N, 20◦5630E). Ligularia sibirica (L.) Cass. (Asteraceae) is a perennial hemicryp- The distance between populations within regions varied from tophyte 50–160 (210) cm high. It has ground rosettes of leaves and a 100 m to approximately 10 km. The minimum distance between short rhizome. During flowering from mid July to the end of August, the Czech and Slovak populations is 410 km. L. sibirica creates 1–8 flowering stems with 20 (up to 55) flower Habitat conditions in most of the study localities seem suitable heads per inflorescence on average. The flower heads are yellow- for L. sibirica with exception of northern Bohemia region in the ish and bisexual. The species is entomogamous, pollinated mainly Czech Republic (populations CZ N/3, CZ N/4 and CZ N/5, Table 1). by bees (Slavík, 2004). The oval achenes (further referred to as Here the fens had been fragmented and partly meliorated in the seeds) are on average 2.16 mg heavy (Smídovᡠunpubl.). Their pap- past (Hendrych, 2003). pus forms a tuft of hairs longer than the achenes (Hegi, 1929; Slavík, Historical data about distribution of L. sibirica in the Czech and 2004). Seeds are either dispersed by wind over a short distance or Slovak Republic are available since the early 19th century. They are by gravity. Terminal velocity of L. sibirica seeds varies between 0.86 not very detailed but they cover all main regions from which the and 1.18 m s−1 (Smídovᡠunpubl.). The species is diploid (2n = 60) species is currently known. L. sibirica was first mentioned from the (Liu, 2004). Both sexual and vegetative reproduction occurs. It has a Czech Republic in 1814 from northern Bohemia (Hendrych, 2003). ‘phalanx’ type of clonal plant growth form which represents a tight- The localities in central Bohemia are known since 1843. The locality packed advancing front of ramets with slow radial spread (sensu in southern Bohemia was found in 1984. The nativity of this locality Doust, 1981). was questioned, but no clear conclusion on this issue was drawn L. sibirica is a plant species which prefers full sunlight, but (Hendrych, 2003; Slavík, 2004). Populations in the Slovak Republic it also grows under open tree canopy where it diminishes seed in the NP Slovensky´ raj are known since 1866. The isolated popu- production (Kukk, 2003). It occurs in a variety of wetlands, such lation in the Salvatorské lúky was first mentioned in 1853 (Domin, as humid grasslands, alkaline mires and alluvial woodlands with 1940). The locality in the NP Nízke Tatry was recorded by botanists the groundwater level at surface. The optimum of this species only in year 2000 (Turis, 2000), but the species was present at the is in the alliances Caricion davallianae, Magnocaricion elatae and locality probably for at least 50 years (Turis unpubl). A. Smídovᡠet al. / Flora 206 (2011) 151–157 153

Fig. 1. Geographic map of sampling localities of L. sibirica populations (codes in Table 1) in the Czech and Slovak Republic.

Sampling and genetic analyses a higher genetic variation than adults. Mixing seedlings and adults in different proportions would lead to biased estimates of genetic Allozyme variation was analyzed for a total of 400 L. sibirica diversity of the populations (Mandák et al., 2006; Dostálek et al., individuals sampled from 20 populations in the Czech and Slovak 2010). Leaf tissue was kept on ice during transport to laboratory Republic (Table 1). In each population 20 randomly selected indi- and stored in a refrigerator until extraction. viduals separated by at least 5 m from each other were sampled. Approximately 70 mg of leaf tissue was mechanically ground We decided not to sample seedlings, because seedlings often have with Dowex-Cl (1-X8) and quartz sand and homogenized on

Table 1 Estimates of population size and genetic variability for the studied populations of Ligularia sibirica. The populations are divided according to states (CZ and SK) and regions (N,

C, S, SR, NT and P). A = mean number of alleles per locus, AP = mean number of alleles per polymorphic locus, P = percentage of polymorphic loci, TA = total number of alleles, U = number of unique alleles, HO = observed heterozygosity, HE = expected heterozygosity, FIS = fixation index. Standard deviations are given in parentheses.

Population Abbrev. Estimated AAP PTAUHE HO FIS pop. size (n individuals)

Czech Republic (CZ) 2 Northern Bohemia (N) Vojenské louky CZ N/1 126 2.182 2.857 63.64 25 0 0.314 0.291 0.073 Baronsky´ rybník CZ N/2 243 2.272 3.143 63.64 25 0 0.238 0.264 −0.107 Louky u Robeceˇ CZ N/3 193 2.000 2.429 63.64 22 0 0.292 0.364 −0.245 Rákosina u Robeceˇ CZ N/4 323 2.182 2.857 63.64 24 0 0.294 0.318 −0.084 Slunecníˇ dvur˚ CZ N/5 123 2.091 2.857 63.64 21 0 0.238 0.272 −0.140 Central Bohemia (C) Reckovˇ CZ C/1 ∼10,000 2.545 3.125 72.73 26 0 0.373 0.370 0.008 Louky u NV CZ C/2 ∼10,000 2.454 3.125 72.73 27 0 0.320 0.370 −0.156 Klokockaˇ CZ C/3 ∼7000 2.182 3.167 54.55 24 0 0.292 0.277 0.052 Valcha CZ C/4 85 2.182 3.143 63.64 24 0 0.283 0.273 0.033 Southern Bohemia (S) Olsinaˇ CZ S/1 224 2.182 3.000 63.64 23 0 0.256 0.250 0.022 Slovak Republic (SK) 3 NP Slovensky´ raj (SR) 2 Pusté Pole SK SR/1 ∼28,000 2.182 2.625 72.73 24 0 0.331 0.386 −0.166 Mokrá SK SR/2 ∼300 2.272 2.750 72.73 25 0 0.329 0.418 −0.272 Vysoky´ SK SR/3 ∼5000 2.364 2.875 72.73 26 0 0.291 0.341 −0.171 DobsinᡠSK SR/4 ∼20,000 2.545 3.125 72.73 28 0 0.359 0.404 −0.128 Stratená dolina SK SR/5 ∼1000 2.364 2.875 72.73 26 0 0.329 0.386 −0.175 Malé Zajfy SK SR/6 ∼150 2.091 2.714 63.64 23 0 0.270 0.359 −0.332 Vernár SK SR/7 ∼18,000 2.273 3.000 63.64 25 0 0.295 0.382 −0.295 Biela Voda SK SR/8 ∼200 2.182 2.857 63.64 24 0 0.279 0.286 −0.027 NP Nízke Tatry (NT) Liptovská Teplickaˇ SK NT/1 ∼1500 2.273 3.000 63.64 25 0 0.240 0.282 −0.176 Presovˇ (P) Salvatorské lúky SK P/1 ∼2000 2.182 3.167 54.55 24 0 0.296 0.314 −0.059 Means 2.250 (0.144) 2.934 (0.203) 65.91 (5.806) 24.55 (1.637) 0.360 (0.269) 0.331 (0.318) −0.117 (0.119) 154 A. Smídovᡠet al. / Flora 206 (2011) 151–157 ice in extraction buffer (0.1 M Tris–HCl, pH 8.0, 70 mM of 2- Table 2 mercaptoethanol, 26 mM sodium metabisulfite, 11 mM ascorbic Results of the analyses of molecular variance (AMOVA) for 20 populations of Ligularia sibirica. acid, 4% (w/v) polyvinylpyrrolidone). The extraction and elec- trophoresis followed Kaplan et al. (2002) except for diaphorase Locus Among regions (%) Among populations Within (DIA, E.C. 1.6.99.3). The staining solution of diaphorase consisted within regions (%) populations (%) of 100 ml Tris–HCl buffer, pH 8, 26 mg NADH, 10 mg MTT and 4 mg Df 1 18 778 of 2,6-dichlorphenol-indophenol. All gels were incubated in the LAP 0.258 NS 18.989*** 80.752*** AAT-1 17.043** 9.748*** 73.208*** dark at 35 ◦C until bands appeared. Afterwards, all gels were thor- AAT-2 12.899 NS 15.573*** 71.528*** oughly rinsed in distilled water, dried between two cellophanes 6PGDH-1 12.318 NS 18.036*** 69.646*** and stored. Staining recipes are available on request. The follow- 6PGDH-2 16.177*** 11.039*** 72.784*** ing 11 enzymes were analysed: ADH, AAT, DIA, EST, LAP, PGI, PGM, SOD-1 4.211 NS 8.596*** 87.193*** − PRX, 6-PGDH, SHDH and SOD. From those we chose five enzyme SOD-2 0.208 NS 1.544** 101.336** DIA-1 8.048*** 2.739*** 89.213*** systems, which reveal a clear pattern for genetic analyses. These DIA-2 −1.114 NS 19.055*** 82.057*** five enzyme systems are: aspartate aminotransferase (AAT, E.C. 2.6.1.1), diaphorase (DIA, E.C. 1.6.99.3), leucyl aminopeptidase (LAP, Overall 6.54*** 12.65*** 80.81*** E.C. 3.4.11.1), superoxide dismutase (SOD, E.C. 1.15.1.1) and 6- Note: Overall values are results from overall AMOVA analyses, i.e. not averaged over phosphogluconate dehydrogenase (6-PGDH, E.C. 1.1.1.44). A total single loci. The P values are derived using 1000 permutations of the data. Significance of 11 loci were assayed for this species–AAT (2), DIA (3), LAP (1), levels *P < 0.01, **P < 0.005, ***P < 0.001. NS = not significant. SOD (3) and 6-PGDH (2).

occurred only in the Slovak Republic. The mean number of alleles Data analyses per locus (A) was 2.250 (range: 2.000–2.545), alleles per polymor- phic locus (AP) was 2.934 (range: 2.429–3.167). The percentage of Standard measures of genetic diversity were calculated with the polymorphic loci (P) was 65.91% (range: 54.55–72.73%). Total num- program Arlequin (Arlequin 3.11; Excoffier and Schneider, 2005) ber of alleles (TA) was 24.55 and ranged from 21 to 28. The mean for each population. Calculated measures of genetic diversity are expected population heterozygosity (HE) was 0.360 and ranged mean number of alleles per locus (A) and per polymorphic locus from 0.238 to 0.373. The mean observed heterozygosity (HO) was (AP), percentage of polymorphic loci (P), total number of alleles (TA) 0.331 and ranged from 0.264 to 0.418 (Table 1). and number of unique alleles (U). We calculated mean observed Analysis of molecular variance (AMOVA) showed that 80.8% of (HO) and mean expected (HE) heterozygosity for all loci (Nei, 1973) molecular variance can be found within populations, 12.7% among using Popgene (version 1.32; Yeh and Boyle, 1997). populations within Czech and Slovak Republic and only 6.5% among Variance components and their significance level for variation the two republics (Table 2). The results of AMOVA for the study among groups, among populations within groups and within popu- regions showed 82% variation within populations, 10.6% among lations were conducted using AMOVA (Arlequin 3.11; Excoffier and populations within regions and 7.4% among regions. All propor- Schneider, 2005). Variance components were calculated separately tions of variance were significantly different from zero (AMOVA, for regions and the two republics (Czech and Slovak Republic). P < 0.001). In Popgene (version 1.32; Yeh and Boyle, 1997) we calcu- The overall fixation index, FIS, of individuals relative to their lated summary F-statistics (FIS, FIT and FST). FST was calculated as population was −0.116 and the mean overall inbreeding coefficient, genetic differentiation between each population compared with FIT, was 0.084. Genetic differentiation between populations over all all other 19 pooled populations. We also calculated Wright’s fix- loci and populations, FST, was 0.179 (Table 3). ation index (FIS) to determine deviations from Hardy–Weinberg expectations for each locus and population using the method of Wright (1978). Using 2 test, we tested the significance of devia- Table 3 tions from Hardy–Weinberg equilibrium at each polymorphic locus F statistics and average gene flow for Ligularia sibirica per polymorphic locus. (A) and population. Over all loci (B) for each study region and two states (Czech and Slovak Republic). FIS = fixation index, FIT = inbreeding coefficient, FST = genetic differentiation between We did not estimate level of gene flow (Nm) because studied each population compared with all 19 other pooled populations. populations do not fit the assumptions of equal number of individ- uals per population and no spatial structure. Also the newly reduced Locus FIS FIT FST populations have clearly not yet reached equilibrium between (A) migration and genetic drift (Whitlock and McCauley, 1999). LAP 0.209 0.371 0.205 We tested associations between log-transformed geographic AAT-1 0.088 0.287 0.218 AAT-2 0.094 0.318 0.247 distance and corresponding genetic distance between pairs of pop- 6PGDH-1 0.017 0.281 0.268 ulations (FST) with a Mantel test (Mantel, 1967) for the whole data 6PGDH-2 −0.083 0.159 0.224 set and then also for the two countries and three regions separately SOD-1 −0.063 0.071 0.126 − − using Arlequin (Arlequin 3.11; Excoffier and Schneider, 2005). SOD-2 0.965 0.959 0.003 DIA-1 −0.800 −0.635 0.092 We tested the relationship between population size and differ- DIA-2 0.517 0.610 0.194 ent measures of genetic diversity using linear regression (Statistica Overall −0.116 0.084 0.179 version 7.0; StatSoft, Inc., 2004). (B) Regions Northern Bohemia −0.096 0.010 0.097 Results Central Bohemia −0.017 0.075 0.091 Southern Bohemia 0.022 0.022 NA NP Slovensky´ raj −0.194 −0.034 0.134 Nine of the 11 loci (81.81%) were polymorphic at least in one NP Nízke Tatry −0.176 −0.176 NA population. Only SOD-3 and DIA-3 were monomorphic across all Presovˇ −0.059 −0.059 NA populations. Overall 32 different alleles were detected in the poly- States morphic loci. Two alleles, LAP-a and 6PGDH-2-b, occurred only Czech Republic −0.051 0.083 0.128 − in the Czech Republic and three alleles LAP-e, LAP-f and SOD-1-b, Slovak Republic 0.179 0.007 0.158 A. Smídovᡠet al. / Flora 206 (2011) 151–157 155

0.45 Table 4 The relationship between different measures of genetic variation of L. sibirica. 0.40 A = mean number of alleles per locus, AP =mean number of alleles per polymorphic 0.35 locus, P = percentage of polymorphic loci, TA = total number of alleles, HO = observed heterozygosity, HE = expected heterozygosity, FIS = fixation index. Significance val- 0.30 ues in bold, P < 0.05. 0.25 AAP PTAHO HE FIS ST 0.20 F A

0.15 AP 0.54 P 0.64 −0.21 0.10 TA 0.91 0.47 0.57

0.05 HO 0.43 −0.32 0.67 0.47 H 0.62 0.02 0.58 0.61 0.75 0.00 E FIS 0.07 0.51 −0.32 0.02 −0.63 0.04 -0.05 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 log distance (km) ity (HO; r = 0.517; P = 0.02). The different response variables were Fig. 2. The relationship between pairwise population FST values and logarithm of partly correlated (Table 4). geographical distance (km). The relationship is significant (r = 0.395, P < 0.001, Man- tel test). Dots around 2.6 log distances represent comparisons of Czech vs. Slovak populations. Discussion

Out of the total 145 tests for deviance of 9 loci/population Main species traits that affect the amount of genetic diversity of from Hardy–Weinberg equilibrium, 66 (45.5%) were significant plant species according to Hamrick and Godt (1989) are: life form, (P < 0.01). Of these, each 50% had positive FIS values, indicating geographical range and breeding system. Long-lived, perennial, a heterozygote deficit, and 50% had negative FIS values, indicat- widespread, outcrossing species have generally more genetic diver- ing a homozygote deficit. These deviations were scattered across sity than short lived, selfing or mixed-mating, narrow endemic populations and loci with no obvious trends. Overall, eight out species (Hamrick and Godt, 1989, 1996). L. sibirica is a long-lived, of nine polymorphic loci deviated significantly (P < 0.01) from perennial herb with mixed-mating breeding system, in Europe clas- Hardy–Weinberg equilibrium. Of these, 75% had positive FIS values sified as a relict. With this combination of species traits, genetic and remaining 25% had negative FIS values. diversity of L. sibirica identified using allozymes should be some- Most of the pairwise genetic distances (FST) between popula- where in between low and high level, when compared with mean tions (189/190) were significant (P < 0.05). The FST values ranged genetic diversity values based on allozymes given in Hamrick and from 0.420 to −0.015. Pairwise genetic distances were signifi- Godt (1989). Observed values are on one hand close to endemic cantly correlated with geographic distances (Mantel test; r = 0.395, species (mean number of alleles per locus (A) = 3.00; L. sibirica: P < 0.001) (Fig. 2). This correlation was also significant within 2.250, SD ± 0.144), on the other hand roughly followed expectation the Czech Republic (Mantel test; r = 0.445, P = 0.002) and within for widespread species (percentage of polymorphic loci (P) = 59%; the Slovak Republic (Mantel test; r = 0.485, P < 0.001) when cal- L. sibirica: 65.91%, SD ± 5.806 and mean expected heterozygosity culated separately. However, when testing regions, a positive (HE) = 0.202; L. sibirica: 0.360, SD ± 0.269). A similar pattern was significant correlation was found only among populations in region found also in studies of some other rare species based on allozymes, NP Slovensky´ raj in the Slovak Republic (Mantel test; r = 0.715, suggesting that the variation in genetic diversity between different P < 0.001) but not among regions northern Bohemia and central rare species can be large (e.g. Dannemann, 2000; Lutz et al., 2000). Bohemia in the Czech Republic (Mantel test; r = 0.392, P = 0.263 and The analysis of molecular variance revealed most of the genetic r = 0.649, P = 0.163, respectively). variability (80.8%) within populations of L. sibirica, 12.7% variance There are significantly positive correlations of logarithm of esti- among populations within regions and only 6.5% among regions. In mated population size and measures of genetic diversity. Mean case of other disjunctly distributed taxa the analysis showed sim- expected heterozygosity (HE; r = 0.561; P = 0.010) increased with ilar high levels of intra-population diversity (Dannemann, 2000; population size (Fig. 3). Similar results were found for mean num- Dittbrenner et al., 2005; Lutz et al., 2000; Wróblewska et al., 2003). ber of alleles per locus (A; r = 0.631, P = 0.003), total number of Such a higher genetic variability within, rather than among popu- alleles (TA; r = 0.600, P = 0.005) and mean observed heterozygos- lations is typical for outcrossing, perennial plant species (Hamrick and Godt, 1989, 1996). Thus, our findings may result from particu- 0.38 lar life history traits of L. sibirica, such as both sexual and vegetative reproduction, pollination by bees and long life span that tend to pre- 0.36 serve genetic variability within populations (Loveless and Hamrick, 0.34 1984). The level of genetic differentiation between populations in L. 0.32 sibirica (FST = 0.179) corresponds to values reported for outcrossing

E 0.30 species analysed using allozymes (e.g. Hamrick and Godt, 1996; H Loveless and Hamrick, 1984). L. sibirica is a predominantly out- 0.28 crossing species, but it is self-compatible, and selfing is thus also a ˇ 0.26 possible way of reproduction (Smídová unpubl.). Genetic distance significantly correlated with the logarithmic 0.24 geographic distance in the whole data set as well as in each country

0.22 separately. This finding is in agreement with isolation by distance 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 model (Ellstrand and Elam, 1993) and corresponds with many other Log population size studies testing correlation by distance in natural plant popula- Fig. 3. The relationship between estimated population size and expected heterozy- tions (e.g. Hensen et al., 2005; Reisch, 2001; Wróblewska, 2008). gosity of L. sibirica (r = 0.561; P = 0.01). This trend, however, was not observed for northern and central 156 A. Smídovᡠet al. / Flora 206 (2011) 151–157

Bohemia when testing the regions separately. Observed isolation All the above suggests that L. sibirica is an ‘old rare’ species with by distance may be generally caused by limited seed or pollen the ability to maintain a good level of genetic variability for a longer dispersal (Gaudeul et al., 2000). In our study, significant genetic period of time. In this study we also found a significant effect of differentiation was found even for populations that are 100 m population size on genetic diversity measured as mean number of apart. This finding is in agreement with Ehrlich and Raven (1969), alleles per locus, total number of alleles, observed and expected het- who observed that already distances of 15 m to a few kilometres erozygosity, i.e. a pattern commonly reported to ‘new rare’ species can effectively isolate two populations of insect pollinated plant (e.g. Fischer and Matthies, 1998; Leimu et al., 2006, for reviews; species. Long-distance dispersal of seeds is also quite unlikely. Non- but see Bachmann and Hensen, 2007; Oostermeijer et al., 1994) theless, L. sibirica seems to have a good wind-dispersal potential of and as well in ‘old rare species’ (Dittbrenner et al., 2005; Reisch seeds (i.e. diaspores are achenes with pappus) and the terminal et al., 2003). Thus our finding is in agreement with the hypothesis velocity is quite high (0.86–1.18 m s−1, Smídovᡠunpubl.). Within that small populations cannot maintain genetic diversity as high the seed dispersal values presented by Soons (2006) for wind dis- as those found in larger populations (Frankham et al., 2006). Such persed wetland , L. sibirica belongs to species with maximal a feature is usually explained by genetic drift, by either random dispersal distances of several tens of meters. Additionally it has to loss of alleles, founder effects during colonization and/or inbreed- be kept in mind that generally most seeds are dispersed close to ing events (Soulé, 1986). Genetic drift, concretely random loss of the parental plant (Nathan and Muller-Landau, 2000; Tackenberg alleles, may be true for some small sized populations of L. sibir- et al., 2003). ica in northern Bohemia which have experienced relatively recent High within population genetic variability together with low reduction of population size due to negative changes of habitat con- genetic differentiation may suggest high gene flow. Alternatively, ditions in last centuries, such as melioration and fragmentation such pattern may be a result of past events (Van Rossum and of fens (Hendrych, 2003). Founder effect may be the explanation Prentice, 2004). One hypothesis explaining this pattern in L. sibirica for one population (CZ S/1) which has been suggested that it may could be that distribution of genetic diversity within and between be of secondary origin by intentional planting (Hendrych, 2003; populations still reflects gene flow in the past before the formerly Slavík, 2004). We assume that if this was the case, the expected larger populations were fragmented (Hensen et al., 2005; Lowe genetic diversity would be even lower than it would correspond to et al., 2004; Peterson et al., 2008). The second hypothesis to be its size. This was, however, not the case and we thus do not have considered is that current gene flow connects populations that are any support for this speculation. geographically closer to each other more efficiently than popula- tions divided by greater distances (Hensen et al., 2005; Peterson Conclusion et al., 2008). This option is likely for northern and central Bohemia regions. We assume that both above proposed hypotheses might be The results of our study on genetic diversity of an ‘old rare’ true for L. sibirica. Similar results were found for other glacial relict species, L. sibirica showed high genetic diversity within populations species (e.g. Dannemann, 2000; Lutz et al., 2000; Reisch et al., 2003). and low level of genetic differentiation between populations which Populations of the glacial relicts were probably more widespread in roughly follows values for outcrossing species. Genetic distance Pleistocene (Reisch, 2001) and have maintained their high genetic between populations correlated significantly with geographic dis- diversity in spite of approximately 12,000 years of isolation. tance suggesting the isolation by distance model for our study Similarly to glacial relicts L. sibirica was probably more species. This model is in agreement with our hypothesis that Ligu- widespread in the past and its populations are presumably laria sibirica migrated to Europe via slow unidirectional stepwise naturally fragmented for approximately 10,000–7000 years. How- migration. There was also a significant positive correlation between ever, in literature we can find several theories on the origin genetic diversity and population size. This is probably caused by of the species in Central and Western Europe. One, but today destruction of habitats of the species in fens in northern Bohemia, overcome idea is import of seeds with hay to feed the Cos- the Czech Republic, in last centuries and subsequent decrease of sack horses during the Napoleonic Wars from Russia (Kneblová, population size. Nevertheless, L. sibirica populations still preserve 1950; Procházka and Pivnicková,ˇ 1999). Another hypothesis is a high levels of genetic diversity and are not yet threatened by genetic rare long-distance dispersal by migrating or overwintering birds factors. However, this may change if changes in habitat conditions (Marsáková-Nˇ emejˇ cková,ˇ 1973; Walter and Straka, 1970). Such continue. rare events would have to be repeated several times and several founder effects would lead to loss of alleles and to homozygos- ity (Hewitt, 1999). Our results thus contradict this hypothesis. The Acknowledgments commonly accepted explanation is migration of L. sibirica sometime during the glacial period (Pleistocene, Mattauch, 1936) or in post- We would like to thank workers of protected areas of the Czech glacial period (early Holocene, Hendrych, 2003; Meusel and Jäger, and Slovak Republic in which L. sibirica occurs, for providing us with 1992). Most probably L. sibirica came to Europe from north-east information about localities, J. Smídaˇ for preparing the map and T. in early Holocene, when the climatic conditions and vegetation in Heinken for help in field and useful comments on the manuscript. Europe were similar to conditions in the current centre of its distri- This study was supported by grant MSMTˇ 2B06178 and partly also bution in southern central Siberia (Hendrych, 2003) (high moisture by MSMTˇ 0021620828 and AV0Z60050516. and temperature in summer, low temperature in winter, Meusel and Jäger, 1992). High level of genetic variability and low genetic References differentiation observed in L. sibirica is characteristic for slower uni- directional stepwise migration (isolation by distance, Gaudeul et al., Bachmann, U., Hensen, I., 2007. Is declining Campanula glomerata threatened by 2000; Hewitt, 1999). We have no clear idea how widespread L. sibir- genetic factors? Plant Species Biol. 22, 1–10. ica was in that time in Europe, but certainly, it was present in a much Bauert, M.R., Kälin, M., Baltisberger, M., Edwards, P.J., 1998. No genetic variation detected within isolated relict populations of Saxifraga cernua in the Alps using broader area than today. Consequently, most of the recent popu- RAPD markers. Mol. Ecol. 7, 1519–1527. lations are remnants of former larger populations that persisted in Booy, G., Hendriks, R.J.J., Smulders, M.J.M., Van Groenendael, J.M., Vosman, B., 2000. climatically suitable locations with appropriate hydrological con- Genetic diversity and the survival of populations. Plant Biol. 2, 379–395. Council of European Communities, 1992. Council Directive 92/43/EEC of 21 May ditions (Hendrych, 2003). This hypothesis is further supported by on the conservation of natural habitats and of wild fauna and flora. Off. J. Eur. the positive relationship between genetic and geographic distance. Commun. 35, 7–50. A. Smídovᡠet al. / Flora 206 (2011) 151–157 157

Dannemann, A., 2000. Der Einfluss von Fragmentierung und Populationsgröße auf Loveless, M.D., Hamrick, J.L., 1984. Ecological determinants of genetic structure in die genetische Variation und Fitness von seltenen Pflanzenarten am Beispiel plant populations. Annu. Rev. Ecol. Evol. Syst. 15, 65–95. von Biscutella laevigata (Brassicaceae). In: Diss. Bot., vol. 330. J. Cramer, Berlin, Lowe, A., Harris, S., Ashton, P., 2004. Ecological Genetics: Design, Analysis, and Appli- Stuttgart. cation. Blackwell, Oxford. Dittbrenner, A., Hensen, I., Wesche, K., 2005. Genetic structure and random amplified Lutz, E., Schneller, J.J., Holderegger, R., 2000. Understanding population history for polymorphic DNA diversity of the rapidly declining Angelica palustris (Apiaceae) conservation purposes: Population genetic of Saxifraga aizoides () in Eastern Germany in relation to population size and seed production. Plant in the lowlands and lower mountains north of the Alps. Am. J. Bot. 87, 583–590. Species Biol. 20, 191–200. Mandák, B., Bímová, K., Mahelka, V., Placková,ˇ I., 2006. How much genetic variation Domin, K., 1940. Vegetacníˇ obrazy ze Slovenska, vol. 2. Carpatica, Praha Ser. B, pp. is stored in the seed bank? A study of Atriplex tatarica (Chenopodiaceae). Mol. 51–72. Ecol. 15, 2653–2663. Dostálek, T., Münzbergová, Z., Placková,ˇ I., 2010. Genetic diversity and its effect on Mantel, N., 1967. The detection of disease clustering and a generalized regression fitness in an endangered plant species, Dracocephalum austriacum L. Conserv. approach. Cancer Res. 27, 209–212. Genet., doi:10.1007/s10592-009-9879-z. Marsáková-Nˇ emejˇ cková,ˇ M., 1973. Popelivka sibirská,ˇ vol. 28. Ochr. Prírody,ˇ Praha, Doust, L.L., 1981. Population dynamics and local specialization in a clonal perennial pp. 15–16. (Ranunculus repens). I. The dynamics of ramets in contrasting habitats. J. Ecol. Mattauch, F., 1936. Über die Ligularia sibirica (L.) Cass., vol. 7. Natur Heimat, Aussig, 69, 743–755. pp. 57–62. Ehrlich, P.R., Raven, P.H., 1969. Differentiation of populations. Science 165, Meusel, H., Jäger, J., 1992. Vergleichende Chorologie der zentraleuropäischen Flora, 1228–1232. vol. 3. Karten. G. Fischer, Jena. Ellstrand, N.C., Elam, D.R., 1993. Population genetic consequences of small popu- Nathan, R., Muller-Landau, H.C., 2000. Spatial patterns of seed dispersal, their deter- lation size: implications for plant conservation. Annu. Rev. Ecol. Evol. Syst. 24, minants and consequences for recruitment. Trends Ecol. Evol. 15, 278–285. 217–242. Nei, M., 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Excoffier, L.G.L., Schneider, S., 2005. Arlequin ver. 3.0. An integrated software pack- Sci. U.S.A. 70, 3321–3323. age for population genetics data analysis. Evol. Bioinform. Online 1, 47–50. Oostermeijer, J.G.B., Eijck, M.W., Nijs, J.C.M., 1994. Offspring fitness in relation to Feráková, V., Maglocky,´ S.,ˇ Marhold, K., 2001. Cervenˇ y´ zoznam paprad’orastov a population size and genetic variation in the rare perennial plant species Gentiana semennych´ rastlín Slovenska (December 2001). In: Baláz,ˇ D., Marhold, K., Urban, pneumonanthe (Gentianaceae). Oecologia 97, 289–296. P. (Eds.), Cervenˇ y´ zoznam rastlín a zivoˇ cíchovˇ Slovenska, vol. 20. Ochrana Pegtel, M.D., 1998. Rare vascular plant species at risk: recovery by seeding? Appl. Prírody, Banská Bystrica, pp. 48–81. Veg. Sci. 1, 67–74. Fischer, M., Matthies, D., 1998. RAPD variation in relation to population size and Peterson, A., Bartish, I.V., Peterson, J., 2008. Effects of population size on genetic plant fitness in the rare Gentianella germanica (Gentianaceae). Am. J. Bot. 85, diversity, fitness and pollinator community composition in fragmented popula- 811–819. tions of Anthericum liliago L. Plant Ecol. 198, 101–110. Frankham, R., Ballou, J.D., Briscoe, D.A., 2006. Introduction to Conservation Genetics. Pott, R., 1995. The origin of grassland plant species and grassland communities in Cambridge University Press, Cambridge. Central Europe. Fitosociologia (Padua) 29, 7–32. Gaudeul, M., Taberlet, P., Till-Bottraud, I., 2000. Genetic diversity in an endangered Procházka, F., Pivnicková,ˇ M., 1999. Ligularia sibirica (L.) Cass. In: Ceˇ rovskˇ y,´ J., Fer- alpine plant, Eryngium alpinum L. (Apiaceae), inferred from amplified fragment áková, V., Holub, J., Maglocky,´ S.,ˇ Procházka, F. (Eds.), Cervenᡠkniha ohrozenych´ length polymorphism markers. Mol. Ecol. 9, 1625–1637. a vzácnych druhov rastlín a zivoˇ cíchovˇ SR a CR,ˇ vol. 5. Vysˇsieˇ rostliny. Príroda Gonzales, E., Hamrick, J.L., 2005. Distribution of genetic diversity among disjunct a.s., Bratislava, p. 219. populations of the rare forest understory herb, Trillium reliquum. Heredity 95, Procházka, F. (Ed.), 2001. Cernˇ ya´ cervenˇ y´ seznam cévnatych´ rostlin Ceskéˇ republiky 306–314. (stav v roce 2000), vol. 18. Príroda,ˇ Praha, pp. 1–166. Hamrick, J.L., Godt, M.J.W., 1989. Allozyme diversity in plant species. In: Brown, Rasmussen, K.K., Kollmann, J., 2004. Poor sexual reproduction on the distribution A.H.D., Clegg, M.T., Kahler, A.L., Weir, B.S. (Eds.), Plant Population Genetics, limit of the rare tree Sorbus torminalis. Acta Oecol. 25, 211–218. Breeding and Genetic Resources. Sinauer, Sunderland, MA, pp. 43–63. Reisch, Ch., 2001. Climatic oscillations and the fragmentation of plant populations – Hamrick, J.L., Godt, M.J.W., 1996. Effects of life history traits on genetic diversity in genetic diversity within and among populations of the glacial relict plants Sax- plant species. Philos. Trans. Roy. Soc. Lond. B 351, 1291–1298. ifraga paniculata (Saxifragaceae) and Sesleria albicans (Poaceae). Dissertation, Hegi, G., 1929. Illustrierte Flora von Mitteleuropa. 1st ed., vol. 6/2. A.Pichler’s Witwe University Regensburg. und Sohn, Wien, pp. 796–799. Reisch, Ch., Poschlod, P., Wingender, R., 2003. Genetic variation of Saxifraga pan- Hendrych, R., 2003. Poznatky o druhu Ligularia sibirica v Cechách.ˇ Preslia 75, 39–69. iculata Mill. (Saxifragaceae): molecular evidence for glacial relict endemism in Hensen, I., Oberprieler, C., Wesche, K., 2005. Genetic structure, population size, and central Europe. Biol. J. Linn. Soc. 80, 11–21. seed production of Pulsatilla vulgaris Mill. (Ranunculaceae) in central Germany. Robinson, A.R., Sutherland, W.J., 2002. Post-war changes in arable farming and bio- Flora 200, 3–14. diversity in Great Britain. J. Appl. Ecol. 39, 157–176. Hensen, I., Kilian, C., Wagner, V., Durka, W., Pusch, J., Wesche, K., 2009. Low genetic Slavík, B., 2004. Ligularia Cass. In: Slavík, B., Stˇ epánková,ˇ J. (Eds.), Kvetenaˇ Ceskéˇ variability and strong differentiation among isolated populations of the rare Republiky, vol. 7. Academia, Praha, pp. 306–309. steppe grass Stipa capillata L. in Central Europe. Plant Biol., doi:10.1111/j.1438- Soons, M.B., 2006. Wind dispersal in freshwater wetlands: knowledge for conserva- 8677.2009.00227.x. tion and restoration. Appl. Veg. Sci. 9, 271–278. Hewitt, G.M., 1999. Post-glacial re-colonization of European biota. Bot. J. Linn. Soc. Soulé, M.E., 1986. Viable Populations for Conservation. Cambridge University Press, 68, 87–112. Cambridge. Hofhanzlová, E., Fér, T., 2009. Genetic variation and reproduction strategy of Gen- StatSoft, Inc., 2004. STATISTICA (data analysis software system), version 7. tiana pannonica in different habitats. Flora 204, 99–110. www.statsoft.com. Huenneke, L.F., 1991. Ecological implications of genetic variation in plant popula- Stoate, C, Boatman, N.D., Borralho, R.J., Rio Carvalho, C., de Snoo, G.R., Eden, P., 2001. tions. In: Falk, D.A., Holsinger, K.E. (Eds.), Genetics and Conservation of Rare Ecological impacts of arable intensification in Europe. J. Environ. Manage. 63, Plants. Oxford University Press, Oxford, pp. 31–44. 337–365. Honnay, O., Jacquemyn, H., 2007. Susceptibility of common and rare plant species to Tackenberg, O., Poschlod, P., Bonn, S., 2003. Assessment of wind dispersal potential the genetic consequences of habitat fragmentation. Conserv. Biol. 21, 823–831. in plant species. Ecol. Monogr. 73, 191–205. Jacquemyn, H., Van Rossum, F., Brys, R., Endels, P., Hermy, M., Triest, L., De Blust, G., Turis, P., 2000 Charakteristika raselinískˇ Národného parku Nízke Tatry. In: 2003. Effects of agricultural land use and fragmentation on genetics, demogra- Stanová, V., (Ed.), RaseliniskᡠSlovenska. DAPHNE, Institútˇ aplikovanej ekológie, phy and population persistence of the rare Primula vulgaris and implications for Bratislava, pp. 73–74. conservation. Belg. J. Bot. 136, 5–22. Van Rossum, F., Prentice, H.C., 2004. Structure of allozyme variation in Nordic Silene Kaplan, Z., Placková,ˇ I., Stˇ epánek,ˇ J., 2002. Potamogeton x fluitans (P. natans × P. lucens) nutans (Caryophyllaceae): population size, geographical position and immigra- in the Czech Republic II. Isozyme analysis. Preslia 74, 187–195. tion history. Biol. J. Linn. Soc. 81, 357–371. Kneblová, V., 1950. Reservace u Reckovaˇ a Ligularia sibirica (L.) Cass., vol. 3. Ces.ˇ Bot. Walter, H., Straka, H., 1970. Arealkunde – Floristisch-historische Geobotanik. Ein- Listy, Praha, pp. 17–21. führung in die Phytologie III/2. Ulmer, Stuttgart. Kostrakiewicz, K., Wróblewska, A., 2008. Low genetic variation in subpopulations Whitlock, M.C., McCauley, D.E., 1999. Indirect measures of gene flow and migration: on an endangered clonal plant Iris sibirica in southern Poland. Ann. Bot. Fenn. FST =/ 1/(4Nm+1). Heredity 82, 117–125. 45, 186–194. Wright, S., 1978. Evolution and the Genetics of Populations, vol. IV. Variability Within Krzakowa, M., Michalak, M., 2007. Genetic variability of selected marginal popula- and Among Natural Populations. University of Chicago Press, Chicago. tions of Stipa capillata L. Biol. Lett. 44, 127–135. Wróblewska, A., 2008. From the center to the margins of geographical range: molec- Kukk, Ü., 2003. The distribution of Ligularia sibirica (L.) Cass. in Estonia and changes ular history of steppe plant Iris aphylla L. in Europe. Plant Syst. Evol. 272, 49–65. in its population. Biul. Ogrodów Bot. 12, 11–22. Wróblewska, A., Brzosko, E., Czarnecka, B., Nowosielski, J., 2003. High levels of Leimu, R., Mutikainen, P., Koricheva, J., Fischer, M., 2006. How general are positive genetic diversity in populations of Iris aphylla L., (Iridaceae), an endangered relationships between plant population size, fitness and genetic variation? J. species in Poland. Bot. J. Linn. Soc. 142, 65–72. Ecol. 94, 942–952. Yeh, F.C., Boyle, T.J.B., 1997. Population genetic analysis of co-dominant and domi- Liu, J.-Q., 2004. Uniformity of karyotypes in Ligularia (Asteraceae: ), a nant markers and quantitative traits. Belg. J. Bot. 129, 157. highly diversified of the eastern Qinghai–Tibet Plateau highlands and Young, A., Boyle, T., Brown, T., 1996. The population genetic consequences of habitat adjacent areas. Bot. J. Linn. Soc. 144, 329–342. fragmentation for plants. Trends Ecol. Evol. 11, 413–418.