The Arabidopsis arenosa complex is closely related to the model plant Arabidopsis thaliana. Species and subspecies in the complex are mainly biennial, predominantly outcrossing, herbaceous, and with a distribution range covering most parts of latitudes and the eastern reaches of Europe. In this study we present the first comprehensive evolutionary history of the A. arenosa species complex, covering its natural range, by using chromosome counts, nuclear AFLP data, and a maternally inherited marker from the chloroplast genome [trnL intron (trnL) and trnL/F intergenic spacer (trnL/F-IGS) of tRNALeu and tRNAPhe, respectively]. We unravel the broad-scale cytogeographic and phylogeographic patterns of diploids and tetraploids. Diploid cytotypes were exclusively found on the Balkan Peninsula and in the Carpathians while tetraploid cytotypes were found throughout the remaining distribution range of the A. arenosa complex. Three centers of genetic diversity were identified: the Balkan Peninsula, the Carpathians, and the unglaciated Eastern and Southeastern Alps. All three could have served as long-term refugia during Pleistocene climate oscillations. We hypothesize that the Western Carpathians were and still are the cradle of speciation within the A. arenosa complex due to the high species number and genetic diversity and the concurrence of both cytotypes there.
Citation: Schmickl R, Paule J, Klein J, Marhold K, Koch MA (2012) The Evolutionary History of the Arabidopsis arenosa Complex: Diverse Tetraploids Mask the Western Carpathian Center of Species and Genetic Diversity. PLoS ONE7(8): e42691. https://doi.org/10.1371/journal.pone.0042691
Editor: Simon Joly, Montreal Botanical Garden, Canada
Received: February 27, 2012; Accepted: July 11, 2012; Published: August 3, 2012
Copyright: © Schmickl et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was supported by DFG grant KO 2302/5-2 (to MAK). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Pleistocene climatic fluctuations strongly shape the evolutionary history of new species (e.g., grasshoppers ) and the distribution of genetic variants at the population level within species (e.g., Arabidopsis  and hominins ). In numerous studies of the European flora and fauna, these aspects of Pleistocene climatic oscillations have been investigated, but mainly confined to a single species. In this study we consider a species complex of wild relatives of the model plant Arabidopsis thaliana (L.) Heynh. with various taxa and cytotypes to investigate the hypothesis that past and ongoing gene flow between taxa and ploidal levels contributes to the ability of populations to adapt and survive in rapidly changing environments, particularly during the Pleistocene , .
The Arabidopsis arenosa species complex is one of three major species complexes within the genus Arabidopsis –, formerly treated as Cardaminopsis , . Arabidopsis arenosa and its segregates might not only represent the most ancestral species complex compared to the other two major species lineages, namely Arabidopsis lyrata (L.) O'Kane & Al-Shehbaz and A. halleri (L.) O'Kane & Al-Shehbaz , but the various taxa also harbor greater genetic diversity than any other Arabidopsis species and show a remarkable broad spectrum of ecological adaptations from high alpine regions in the High Tatras to sand dune vegetation in Scandinavia , . A taxonomic overview of the various taxa of the A. arenosa species complex, including ploidal level and geographic distribution, is provided in Table 1. Arabidopsis arenosa is a colline, montane, and subalpine species complex with a mainly Central European distribution range including parts of the Alps and Carpathians. Only a few studies have been attempted to unravel the evolutionary history of the A. arenosa complex , . Several studies focused on the natural hybrid A. suecica (Fr.) Norrl., which is of allopolyploid origin with the maternal parent A. thaliana and a paternal parent from the A. arenosa species complex , , confirmed also by artificial crosses . Polyploidisation, mainly tetraploidisation, is frequent in several taxa of the A. arenosa complex , indicating repeated independent polyploidisation events. Introgression, the stable integration of genetic material from one species into another through repeated backcrossing, was observed between members of the A. arenosa and A. lyrata complexes , .
According to different authors, the Arabidopsis arenosa complex comprises several taxa at various taxonomic levels. The complex has been treated as one species [A. arenosa (L.) Lawalrée] with two subspecies of partly overlapping distribution ranges in Central Europe : the mainly tetraploid subsp. arenosa (2n = 16/32), also occurring in northern Europe, growing mainly on siliceous bedrock and sandy soil, and the tetraploid subsp. borbasii (Zapał.) O'Kane & Al-Shehbaz (2n = 32), growing predominantly on calcareous bedrock and additionally found in the Carpathians. Diploid A. neglecta (Schult.) O'Kane & Al-Shehbaz (2n = 16) was described mainly from the Carpathians and rarely from the Alps, but its occurrence in the Alps is doubtful, since in the Alps this taxon has been introduced as Cardaminopsis arenosa var. intermedia (Kovats) Hayek . Based on morphological and karyological data, several additional, mainly diploid Carpathian taxa at the species and subspecies level have been proposed, which were at that time attributed to the genus Cardaminopsis , . These names were, however, never validly published and kept as nomina provisoria (nom. prov.) , pending ongoing studies aimed at clarifying their exact taxonomic status: Arabidopsis carpatica, nom. prov. (2n = 16), A. nitida, nom. prov. (2n = 16), A. petrogena (A. Kern) V.I. Dorof. subsp. petrogena (2n = 16), and A. petrogena subsp. exoleta, nom. prov. (2n = 32). In general, taxonomic concepts in the A. arenosa species complex are strongly debated .
There is an increasing interest in A. arenosa as a model system for adaptation to calcareous versus siliceous bedrocks (Koch and Widmer, ongoing studies; Bomblies et al., ongoing studies), or character trait research such as shade-tolerance (Bomblies et al., ongoing studies). Additionally, A. arenosa is interesting in terms of hybrid speciation, as it is the paternal parent of the natural allopolyploid A. suecica. Studies on the genomic consequences of hybridization are underway, and a first assembly of the A. arenosa genome is available (wiki.bioinformatics.ucdavis.edu/index.php/Arabidopsis_arenosa_whole_genome_assembly). Working with A. arenosa as a model system needs careful consideration of the evolutionary history of the taxa one is investigating, particularly the distribution of natural genetic variation within and among taxa.
The following two aspects are the focus of our research: Unravelling the broad-scale cytogeographic and phylogeographic patterns of diploids and tetraploids: We are particularly interested in contact zones of populations with different or mixed ploidal levels, as they can indicate ongoing species differentiation. The second is detecting centers of genetic diversity: In the northern hemisphere late Quaternary climate oscillations, especially the last glacial maximum (LGM), about 26,500 to 19,000–20,000 years ago, had the most severe influence on present-day distribution and diversity of plant taxa. Arabidopsis arenosa is distributed both in regions that remained largely unglaciated during Pleistocene climate oscillations and in areas formerly covered by glaciers, making it well-suited for comparative studies of evolution in changing environments.
Materials and Methods
The accession list is provided in Table S1. Geographic distribution of single accessions is shown in Fig. 1A. Ploidy information was obtained from 214 accessions (126 populations, one to eleven individuals with five flowers each), AFLP data from 356 accessions (275 populations, one to seven individuals each), and plastid trnL/F sequence data from 365 accessions (260 populations, one to eleven individuals each) (Table S1).
Maximal glaciation and mountain glaciers of the LGM are drawn according to Ehlers and Gibbard . The borders of the seven geographic regions are indicated (see also Table S1, where the affiliation of each accession to one of these regions is listed). A: Visualization according to taxonomy. Seven entities are distinguished: A. arenosa subsp. arenosa, A. arenosa subsp. borbasii, A. carpatica, A. neglecta, A. nitida, and A. petrogena, following Měsíček ,  and Kolník , and Arabidopsis arenosa var. intermedia from the Alps. B: Visualization according to ploidal level (diploids and tetraploids). Ploidal level estimates were only available for a subset of accessions. C: Visualization according to chloroplast DNA suprahaplotypes.
Mitotic chromosome preparations
Ploidy was determined both in this and a previous publication  and additionally from herbarium vouchers from the Herbarium of the Natural History Museum Vienna on which Polatschek had indicated chromosome numbers. The respective source of ploidy determination is recorded in Table S1. Cytological methods and light microscopy were applied according to Schmickl et al. .
DNA isolation and amplified fragment length polymorphisms (AFLPs)
AFLP analysis was performed using a standard protocol  with the following modifications: Approximately 200–500 ng DNA was digested and ligated in a 15 µl reaction mix containing T4 ligase buffer and ATP solution (Bioline, USA), 50 mM NaCl, 0.75 µg BSA, 1.5 U T4 ligase (Bioline, USA), 1 U MseI and 5 U EcoRI (New England Biolabs, USA), 0.37 µM EcoRI adapter and 3.67 µM MseI adapter. The reaction mix was incubated for 3 h at 37°C, followed by an inactivation step for 10 minutes at 65°C. The restriction-ligation product was subsequently diluted tenfold. In the pre-selective PCR 2.5 µl of the diluted restriction-ligation product was used in a total reaction volume of 12.5 µl containing PCR buffer II [Applied Biosystems (ABI), USA], 2 mM MgCl2, 0.8 mM dNTP mix, 0.2 µM EcoRI-A primer (5′-GAC TGC GTA CCA ATT CA-A-3′), 0.2 µM MseI-C primer (5′-GAT GAG TCC TGA GTA AC-C-3′), and 0.25 U AmpliTaq polymerase (ABI). The reactions were held at 72°C for 2 min followed by 20 cycles of 94°C for 20 s, 56°C for 30 s, and 72°C for 2 min with a final 30 s extension at 60°C. The pre-selective PCR product was visualized on a 1.5% agarose gel and diluted tenfold. For selective PCR we used 2.5 µl of the diluted pre-selective PCR product as template in a total reaction volume of 12.5 µl. The PCR mix contained 1× GoldTaq buffer (ABI), 2.5 mM MgCl2, 0.8 mM dNTP mix, 0.08 µM EcoRI fluorescence labelled primer, 0.2 µM Mse primer [EcoRI-AGG(TET)/MseI-CTC, EcoRI-AAC(6-FAM)/MseI-CTG, EcoRI-AAG(HEX)/MseI-CAC], and 0.5 U AmpliTaq Gold (ABI). The reactions were held at 95°C for 5 min followed by 13 cycles of 94°C for 30 s, 65°C→56°C (−0.7°C per cycle) for 1 min, and 72°C for 1 min, followed by 23 cycles of 94°C for 30 s, 56°C for 1 min, and 72°C for 1.5 min with a final 8 min extension at 72°C.
Three differentially fluorescence labelled PCR products of the same sample were multiplexed and diluted, and the fragments were electrophoretically separated on a MegaBase 500 sequencer together with an ET-ROX 550 size standard (Amersham Biosciences, USA). For each run a total of 48 samples were analyzed, including one standard sample, one negative control, one repeat within the runs, and several other replicates (altogether 6.5%), as recommended by Bonin et al. . Raw data were visualized and the fragments in the range of 60–513 bp manually scored using GeneMarker version 1.9 (SoftGenetics, USA). Processed data were exported as a presence/absence matrix.
TrnL/F amplification and sequencing
For the cpDNA markers trnL intron and trnL/F intergenic spacer (trnL/F-IGS), primers, PCR cycling scheme, purification of the amplified fragment, cycle sequencing, and sequencing on a MegaBace 500 sequencer followed the protocol of Schmickl et al. . Amplified sequences of trnL/F-IGS included the complete trnL/F-IGS and the first 18 bases of the trnF gene.
AFLP genetic diversity statistics and Principal Component Analysis (PCA)
Several statistical parameters were computed using the R script AFLPdat , R 2.9.2 environment  for geographic and taxonomic groups: proportion of variable markers (FP) and Nei's gene diversity (HE ). The following seven geographic regions were considered: (1) Balkan Peninsula (Balk), (2) Carpathians (Carp), (3) unglaciated Eastern and Southeastern Alps (UnglaESEAlps), (4) glaciated Eastern Alps (GlaEAlps), (5) glaciated Western Alps (GlaWAlps), (6) unglaciated Central Europe (UnglaCentrEur), and (7) glaciated northern Europe (GlaNEur). These regions are illustrated in Figure 1. Regarding taxonomy, six entities were distinguished following Měsíček ,  and Kolník : A. arenosa subsp. arenosa, A. arenosa subsp. borbasii, A. carpatica, A. neglecta, A nitida, and A. petrogena. In order to analyze and display the similarity among the AFLP genotypes, a Principal Component Analysis (PCA) was performed using MVSP version 3.1 (Kovach Computing Services, UK). Pairwise Euclidean distance was applied as distance measure and, alternatively, both Jaccard and simple match coefficient .
Plastid trnL/F sequence definition, network analysis, genetic diversity statistics, and map reconstruction
Plastid trnL/F sequences were defined as i) haplotypes and ii) suprahaplotypes following our previous studies (e.g., ). Haplotypes (i) are characterized by varying (in sequence and structure) trnF pseudogenes in the 3′-region of the trnL/F-IGS close to the functional trnF gene: Haplotypes belonging to one suprahaplotype (ii) share the same base order throughout the whole sequence except for the pseudogene-rich region, where they vary in both length and base content. Mutation rate within the pseudogene-rich region is about 10 to 20 times higher than within the non-coding spacer and intron regions . Therefore, our cpDNA dataset is based on trnL/F suprahaplotypes only. Suprahaplotypes differ from each other by single point mutations and/or indels. Newly defined trnL/F haplotypes were assigned to GenBank numbers [FJ477684–FJ477690, FJ477705–FJ477716] (Table S1). The network was constructed using TCS version 1.21  using the statistical parsimony algorithm . Single gaps (except polyT stretches) were coded as single additional binary characters. Genetic diversity statistics were performed with Arlequin version 3.11 : Genetic diversity was estimated as effective genetic diversity according to Gregorius (Va) , nucleotide diversity π and Nei's unbiased gene diversity HE.
In order to visualize the geographical data, ArcView version 8.2 (ESRI, USA) was used. The maximum extent of the ice sheets during the LGM was taken from Ehlers and Gibbard .
Chromosome counts identify diploids exclusive to the Balkan Peninsula and the Carpathian Mountains
Two ploidal levels, diploid and tetraploid, were observed within our sampling (Fig. 1B). Diploids were exclusively found in southeastern and eastern Europe on the Balkan Peninsula, in northern Hungary, and in the Carpathians. In contrast, tetraploids have a large distribution range and occur from the Julian Alps (Slovenia) in the south, the Western Carpathians (Slovakia) in the east, France and Belgium in the west, and Scandinavia in the north. Several regions were reported as areas of recent, mainly anthropogenically influenced colonisation after 1890  (Belgium, Finland, France, Great Britain, Greenland), frequently along railway tracks. The Balkan Peninsula and the Western Carpathians were the only regions where both diploid and tetraploid populations were found. Populations of mixed ploidal levels were not observed, but can not be completely ruled out, as only a limited number of populations from the Eastern Alps (n = 28) and the Western Carpathians (n = 9) were analyzed with more than one individual per population.
AFLP data indicate similar values for gene diversity throughout Europe and demonstrate high gene diversity of a widespread tetraploid
Diversity statistics, based on AFLP data, showed similar values for gene diversity throughout the whole distribution range of the A. arenosa species complex (Table 2), ranging from HE = 0.133 (GlaNEur) to HE = 0.159 (UnglaESEAlps). The proportion of variable markers differed more strongly between geographic regions, ranging from FP = 0.411 (Balk) to FP = 0.906 (UnglaESEAlps). However, the proportion of variable markers is biased with respect to sample size (Balk: n = 9, UnglaESEAlps: n = 114) and, therefore, not a valid measurement of genetic diversity. Genetic diversity patterns in the A. arenosa complex need additional consideration in terms of taxonomy, as for certain regions, e.g., Carp, numerous taxa are described, and in other regions, e.g., GlaNEur, only one taxon is found. Gene diversity of the different taxa ranged from HE = 0.157 (A. arenosa subsp. arenosa) and HE = 0.155 (A. arenosa subsp. borbasii) to HE = 0.138 (A. carpatica and A. petrogena) and HE = 0.125 (A. neglecta). The proportion of variable markers is, again, highly correlated with sample size (A. neglecta: n = 6, FP = 0.296; A. arenosa subsp. arenosa: n = 277, FP = 0.985) and so should be treated with caution.
PCA according to regions (Fig. 2A) resulted in overlapping groups of AFLP genotypes from nearly all regions, except for accessions from parts of Carp and UnglaESEAlps. Groups of AFLP genotypes according to taxonomy also largely overlapped (Fig. 2B). Widespread A. arenosa subsp. arenosa but also subsp. borbasii formed large groups of AFLP genotypes in comparison to A. carpatica, A. neglecta, and A. petrogena (A. nitida was omitted, as it was represented by one accession only). This finding underlines the high genetic plasticity of the tetraploids (A. arenosa subsp. arenosa and subsp. borbasii) in contrast to the mainly diploids [A. carpatica (exclusively diploid), A. neglecta (predominantly diploid), A. petrogena (partially diploid)]. PCA according to ploidal levels (Fig. 2C) revealed two partly overlapping clusters of diploids and tetraploids, but due to many accessions without ploidal level estimates the two clusters could actually be more strongly intermingled.
Each symbol represents an individual. A: Visualization according to geographic regions. The following seven geographic regions were considered: (1) Balkan Peninsula (Balk), (2) Carpathians (Carp), (3) unglaciated Eastern and Southeastern Alps (UnglaESEAlps), (4) glaciated Eastern Alps (GlaEAlps), (5) glaciated Western Alps (GlaWAlps), (6) unglaciated Central Europe (UnglaCentrEur), and (7) glaciated northern Europe (GlaNEur). These regions are illustrated in Figure 1. B: Visualization according to taxonomy. Arabidopsis arenosa var. intermedia is marked with an asterisk. C: Visualization according to ploidal level. Data lacking ploidal level estimates are marked in grey.
The Balkan Peninsula, the Carpathians and the unglaciated Eastern and Southeastern Alps are the three centers of chloroplast sequence diversity of the A. arenosa species complex
Based on trnL/F sequence data, we detected three centers of genetic diversity of the A. arenosa complex (Table 2). The unglaciated Eastern and Southeastern Alps displayed highest effective diversity according to Gregorius (Va = 4.14), highest nucleotide diversity (π = 0.483%), and highest Nei's gene diversity (HE = 0.764) within the whole dataset. The Balkan Peninsula was detected as the second center of genetic diversity (Va = 2.94, π = 0.395%, HE = 0.733). The third center of genetic diversity, the Carpathians, was characterized by gene diversity values similar to those of the Balkan Peninsula (Va = 3.49, π = 0.352%, HE = 0.719). In contrast to these three genetically highly diverse regions, which remained largely unglaciated during Pleistocene climate oscillations, formerly glaciated regions showed reduced values of effective genetic diversity according to Gregorius, nucleotide and Nei's gene diversity: The part of the Eastern Alps formerly covered by glaciers was characterized by Va = 2.23, π = 0.310%, and HE = 0.561. The formerly glaciated Western Alps showed strongly reduced values (Va = 1.24, π = 0.109%, HE = 0.205). Glaciated northern Europe was characterized by Va = 1.81, π = 0.315%, and HE = 0.491. Although Central Europe remained largely unglaciated during Pleistocene climate oscillations, genetic diversity was also reduced (Va = 1.64, π = 0.241%, HE = 0.403). In contrast to AFLP data, effective genetic diversity according to Gregorius, nucleotide and Nei's gene diversity of diploid A. carpatica (Va = 3.71, π = 0.514%, HE = 0.751) exceeded that of tetraploid A. arenosa subsp. arenosa (Va = 3.48, π = 0.410%, HE = 0.714).
In all geographic regions, except GlaWAlps, unique suprahaplotypes were found, which occurred in one region only (Figs. 1C, 3A). These unique types are nearly all derived from the “core” suprahaplotypes A, B, and E: AY (Balk); AE, AZ, BA, BE, P, Y (Carp); F, O, Z (UnglaESEAlps); AW, AX, I, M (GlaEAlps); BB, BC (UnglaCentrEur); W (GlaNEur). In contrast to the suprahaplotypes, which were shared between regions (Fig. 3A) and also between taxa (Fig. 3B; A, B, C, E, L, Q, U), these regionally unique suprahaplotypes were found in exclusively one taxon, except P and Y (Fig. 3B). Regarding ploidal levels (Fig. 3C), diploids and tetraploids shared numerous suprahaplotypes (A, AV, B, C, E, L, U) but also had unique ones (diploids: AY, BA, BE, P, Y; tetraploids: AU, AW, AX, AZ, BB, BC, D, Q).
The sizes of the circles indicate the relative frequency of a suprahaplotype. Geographic regions, taxonomic entities, and cytotypes are indicated with the same colours as in Figure 2. A: Visualization according to geographic regions. B: Visualization according to taxonomy. Arabidopsis arenosa var. intermedia is marked with an asterisk. C: Visualization according to ploidal level.
The Western Carpathian Mountains are the cradle of speciation within the A. arenosa complex
Speciation is often accompanied by polyploidisation . Within the A. arenosa complex two ploidal levels, diploid and tetraploid, were observed. Contact zones of these two cytotypes were localized in the northwestern part of the Balkan Peninsula and in the Western Carpathians, and, consequently, at least one independent polyploidisation event can be assumed for each of these regions. Regarding the Western Carpathians, we observed a mosaic pattern of diploids and tetraploids, which is in congruence with Měsíček  (see also ). In the contact zones of diploids and tetraploids in the Western Carpathians we found no populations with mixed ploidal levels. This finding can be discussed with respect to the origin of the tetraploids: Populations of mixed ploidal levels could either be the result of a relatively recent autopolyploidisation event or the result of secondary contact of formerly allopatric populations with different ploidal levels. The genus Melampodium L., in particular the white-rayed species complex with mainly diploids and tetraploids, is an example for postglacial formation of polyploids via autopolyploidisation. Intrapopulational cytotype mixture was reported, but emphasized as rare . In contrast, secondary contact of formerly allopatric populations with different ploidy is the explanation for populations of mixed ploidal levels (diploids and tetraploids or tetraploids and hexaploids) in the Knautia arvensis agg. . Therefore, we conclude, that the lack of mixed ploidy populations and the lack of uneven ploidy and aneuploidy in the A. arenosa complex excludes recent polyploidisation events as well as secondary contact zones. We assume ancient polyploidisation, probably dated several glacial cycles ago. Slightly different ecological adaptations of diploids and tetraploids might have favored such a distinct pattern of diploid and tetraploid populations. Diploid A. neglecta subsp. neglecta, for example, is found on siliceous substrates, mostly in high alpine habitats above the tree line, where it grows along mountain streams. The tetraploid subspecies robusta (previously recognized as Cardaminopsis neglecta subsp. robusta , but never validated by publication), corresponds to a taxon, which, although it also occurs on siliceous substrates, is found in different mountain ranges and at lower altitudes compared to typical subspecies (mostly around or below the tree line [Kolník, unpubl. data]). This is comparable to Senecio carniolicus Willd. with an altitudinal, ecological gradient composed of mainly diploid and hexaploid populations in the Eastern Alps .
Additionally, the high genetic diversity, based on plastid trnL/F, in the Western Carpathians is an indicator for past and also ongoing speciation within the A. arenosa complex. Not only the high genetic diversity but also the high number of species and subspecies underline the Western Carpathians as a cradle of A. arenosa speciation. Numerous proposed taxa are unique for the Carpathians: A. neglecta is considered to comprise two subspecies, one diploid (subsp. neglecta) and one tetraploid (subsp. robusta) [18; Měsíček, unpubl. data]. Arabidopsis petrogena is exclusively described from the Carpathians as diploid (subsp. petrogena) and as tetraploid (subsp. exoleta [18; Měsíček, unpubl. data]). Arabidopsis nitida is the third unique taxon of the Carpathians. And within broadly defined A. borbasii a Carpathian subspecies (subsp. carpatica, referred to as A. carpatica, nom. prov. , and as C. borbasii subsp. carpatica ) is discussed. However, taxonomy of this highly diverse species complex needs to be revised in the near future.
Long-term evolution in two glacial refugia: the Carpathians and the unglaciated Eastern/Southeastern Alps
In numerous studies of both plant and animal species three classical LGM refugia were reported, based on the fossil record  and species and genetic diversity : the Balkan Peninsula, the Iberian Peninsula, and the Appenin. Out of these three Pleistocene refugia, the Balkan Peninsula was emphasized as the most important refugium, especially for tree species , but also for upper and lower montane taxa of especially eastern European distribution , . The Balkan Peninsula probably served as one of the major refugial areas for the A. arenosa complex, which will be discussed in more detail in the following section. Although the Carpathians remained largely unglaciated during Pleistocene climate oscillations, except in the south, they have been reported as a refugium for only a few plant species so far, including temperate trees ,  and herbs, e.g., Campanula alpina Jacq. . For the A. arenosa complex, the Western Carpathians were already described as a cradle of speciation, and they probably served as a second major refugial area. In contrast, the unglaciated eastern Austrian Alps have long been assumed as a glacial refugium. Along the eastern border of the Austrian Alps a cryptic refugium for tree species was already suggested , according to palynological data and supported by phylogeographic analyses . Additionally, the northeastern Limestone Alps have already been described as rich in subalpine endemics , and this view was further supported by Tribsch and Schönswetter : Based on species diversity and the number of endemics, they suggested refugia for numerous calcicolous and silicicolous plants. As A. arenosa is both a colline-montane and subalpine species complex, only parts of the subalpine refugia of the northeastern Limestone Alps overlapped with the A. arenosa refuge area of the unglaciated Eastern and Southeastern Alps. This is in congruence with the view that the eastern edge of the northeastern Limestone Alps is an area of periglacial survival for montane plant species , .
Parallel evolution in the Eastern Alps and the Western Carpathians
According to trnL/F sequence data, the unglaciated Eastern/Southeastern Alps and the Carpathians formed two distinct genetic groups: the Alps characterized by suprahaplotype B and the Carpathians by suprahaplotype L. Although other suprahaplotypes, e.g., A and E, were shared between these two mountain ranges, we assume strong barriers to gene flow between the Alps and Carpathians, due to the Pannonian Basin, which constituted a lowland barrier for montane to subalpine taxa since the Holocene warming. Long-term genetic isolation between the Alps and Carpathians was also proposed for several other plant species, such as Campanula alpina . Additional studies from Ranunculus glacialis L.  and Rosa pendulina L.  support this view.
However, an alternative explanation for genetic differentiation between A. arenosa populations from the Alps and Carpathians has to be considered. Based on our karyologic data, independent colonisation of both mountain ranges from the Balkan Peninsula can be hypothesized: Compared to all other regions, exclusively diploid cytotypes (except for one accession in the northern part) were found on the Balkan Peninsula. Diploid A. arenosa could have migrated northwards and experienced multiple polyploidisation events, especially in the Carpathians, where taxa of both diploid and tetraploid cytotype are described. The Alps were probably colonised by (a) tetraploid cytotype(s) of A. arenosa, which could have originated in the Julian Alps, where only tetraploids were found. The Balkan Peninsula as the ancient refuge area for the A. arenosa species complex is additionally supported by the occurrence of an Arabidopsis species endemic to the Balkan Peninsula, Arabidopsis croatica. This endemic is closely related to the A. arenosa complex, based on ITS sequence data (internal transcribed spacer region of nuclear encoded ribosomal DNA) , .
Taxa in refuge areas like the Eastern Alps and the Western Carpathians probably underwent long-term adaptational processes, which could result in adaptation to the same ecological niche in the two mountain ranges in parallel. Indeed, we found one example for parallel evolution in the Eastern Alps and the Western Carpathians: Diploid A. neglecta grows on siliceous bedrock along mountain streams in alpine habitats of the High Tatras (Slovakia), a similar ecological niche in the Eastern Alps (Wölzer Tauern, Styria) is occupied by a tetraploid taxon corresponding to Cardaminopsis arenosa var. intermedia (Kovats) Hayek . Besides sharing similar ecological demands, the two taxa are discussed to be morphologically more similar to each other than to any other member of the A. arenosa complex. However, AFLP data did not confirm that the vicariant populations from the Eastern Alps and the High Tatras represent a single species (Fig. 2B). This needs further experimental confirmation based on a broader sampling of both taxa. Interestingly, two other species from the Brassicaceae co-occur in the same or neighbouring alpine regions on siliceous bedrock along mountain streams: hexaploid Cochlearia tatrae Borbás in the High Tatras (Slovakia) and diploid Cochlearia excelsa Zahlbr. ex Fritsch in the Eastern Alps (Seckauer Tauern, Styria). In parallel to A. neglecta and C. arenosa var. intermedia, these two Cochlearia species evolved independently and most likely within the last approximately 100,000 years , . Parallel evolution of species pairs implies effective reproductive isolation between the two mountain ranges and limited multiple immigration.
Tetraploid A. arenosa subsp. arenosa is a highly genetically diverse taxon
Based on both AFLP and chloroplast sequence data, genetic diversity was not strongly reduced in regions formerly covered by glaciers (GlaEAlps, GlaWAlps, GlaNEur) in comparison to formerly unglaciated regions (UnglaCentrEur). In all these four regions, exclusively tetraploid A. arenosa subsp. arenosa and subsp. borbasii were found. Local, periglacial survival could serve as one explanation for the high genetic diversity of especially A. arenosa subsp. arenosa. In Central Europe, numerous populations are restricted to relict habitats on exposed rocks in low mountain ranges, e.g., the Black Forest, the Eifel, the Elbe Sandstone Mountains, the Harz Mountains, and the Swabian Mountains, where they often co-occur with Pleistocene relict species such as Dianthus gratianopolitanus Vill. [Koch, unpubl. data]. A second explanation could be the lack of genetic bottlenecks and the maintenance of large effective population sizes during postglacial migration into formerly glaciated regions, probably enhanced by the plant's biennial life cycle. In a third alternative, gene flow between different taxa and/or ploidal levels could account for the high genetic diversity in especially tetraploid A. arenosa subsp. arenosa, probably before this taxon migrated from the Eastern Alps and Western Carpathians, its putative refuge areas, to Central Europe. Gene flow between different taxa of the A. arenosa species complex is documented by numerous hybrids, including triploid ones, reported from the Western Carpathians . Gene flow between different ploidal levels was recently described for A. lyrata : Based on an isolation with migration model analysis , , gene flow from diploids to tetraploids and vice versa was hypothesized. We assume that all three factors could have contributed to the high genetic diversity of tetraploid A. arenosa, which could have resulted in its ability to be a successful coloniser on various different substrate types (e.g., limestone, sandstone, granite, basalt) in various different habitat types (“natural” sites on rocks, gravel, and sand; anthropogenically influenced sites such as railway tracks and rural areas).
Information about accession details and experimental results: taxonomic unit, taxon name on herbarium voucher, herbarium/herbarium voucher no., accession no., geographic region, latitude/longitude, locality, collector/date of collection, ploidal level/publication or other source, AFLPs, trnL intron type, trnL intron GenBank no., trnL/F-IGS type, trnL/F-IGS GenBank no., trnL intron+trnL/F-IGS type, trnL intron+trnL/F-IGS suprahaplotype. Accessions, for which AFLP data were obtained, are marked (see AFLPs).
We thank the curator of the Herbarium of the Natural History Museum Vienna, Ernst Vitek, for providing plant material, Adolf Polatschek for providing various chromosome counts and sharing his knowledge during the course of the project, Susanne Ball and Michaela Wernisch for laboratory assistance, Martin Kolník for locality data, Jürgen Ehlers and Phil Gibbard for shapefiles of the LGM, and Graham Muir for critically reading the manuscript.
Conceived and designed the experiments: MAK RS. Performed the experiments: RS JP JK. Analyzed the data: RS JP JK MAK. Wrote the paper: RS MAK JP JK KM.
- 1. Carstens BC, Knowles LL (2007) Shifting distributions and speciation: species divergence during rapid climate change. Mol Ecol 16: 619–627.
- 2. Beck JB, Schmuths H, Schaal BA (2008) Native range genetic variation in Arabidopsis thaliana is strongly geographically structured and reflects Pleistocene glacial dynamics. Mol Ecol 17: 902–915.
- 3. Dennell RW, Martinón-Torres M, Bermúdez de Castro JM (2011) Hominin variability, climatic instability and population demography in Middle Pleistocene Europe. Quaternary Sci Rev 30: 1511–1524.
- 4. Schmickl R, Jørgensen MH, Brysting AK, Koch MA (2010) The evolutionary history of the Arabidopsis lyrata complex: a hybrid in the amphi-Beringian area closes a large distribution gap and builds up a genetic barrier. BMC Evol Biol 10: 98.
- 5. Schmickl R, Koch MA (2011) Arabidopsis hybrid speciation processes. Proc Natl Acad Sci U S A 108: 14192–14197.
- 6. O'Kane SL Jr, Al-Shehbaz IA (1997) A synopsis of Arabidopsis (Brassicaceae). Novon 7: 323–327.
- 7. Al-Shehbaz IA, O'Kane SL Jr, Price RA (1999) Generic placement of species excluded from Arabidopsis (Brassicaceae). Novon 9: 296–307.
- 8. Clauss MJ, Koch MA (2006) Poorly known relatives of Arabidopsis thaliana. Trends Plant Sci 11: 449–459.
- 9. Koch MA, Matschinger M (2007) Evolution and genetic differentiation among relatives of Arabidopsis thaliana. Proc Natl Acad Sci U S A 104: 6272–6277.
- 10. Koch MA, Wernisch M, Schmickl R (2008) Arabidopsis thaliana's wild relatives: an updated overview on systematics, taxonomy and evolution. Taxon 57: 933–943.
- 11. Säll T, Jakobsson M, Lind-Halldén C, Halldén C (2003) Chloroplast DNA indicates a single origin of the allotetraploid Arabidopsis suecica. J Evol Biol 16: 1019–1029.
- 12. Jakobsson M, Hagenblad J, Tavaré S, Säll T, Halldén C, et al. (2006) A unique recent origin of the allotetraploid species Arabidopsis suecica: Evidence from nuclear DNA markers. Mol Biol Evol 23: 1217–1231.
- 13. Comai L, Tyagi AP, Winter K, Holmes-Davis R, Reynolds SH, et al. (2000) Phenotypic instability and rapid gene silencing in newly formed Arabidopsis allotetraploids. Plant Cell 12: 1551–1568.
- 14. Měsíček J (1970) Chromosome counts in Cardaminopsis arenosa agg. (Cruciferae). Preslia 42: 225–248.
- 15. Jørgensen MH, Ehrich D, Schmickl R, Koch MA, Brysting A (2011) Interspecific and interploidal gene flow in Central European Arabidopsis (Brassicaceae). BMC Evol Biol 11: 346.
- 16. Al-Shehbaz IA, O'Kane SL Jr (2002) Taxonomy and phylogeny of Arabidopsis (Brassicaceae). The Arabidopsis Book doi: 10.1199/tab.0001.
- 17. Hayek Av (1908–1914) Flora von Steiermark 1–2(1). Berlin: Verlag von Gebrüder Bornträger.
- 18. Měsíček J (1998) Cardaminopsis. In: Marhold K, Hindák F, editors. Zoznam nižších a vyšších rastlín Slovenska – Checklist of non-vascular and vascular plants of Slovakia. Bratislava: VEDA. 395–396.
- 19. Kolník M (2007) Arabidopsis. In: Marhold K, Mártonfi P, Mereďa P Jr, Mráz P, editors. Chromosome number survey of the ferns and flowering plants of Slovakia. Bratislava: VEDA. 94–102 Available: http://www.chromosomes.sav.sk. Accessed 2012 Jul 15.
- 20. Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15.
- 21. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, et al. (1995) AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res 23: 4407–4414.
- 22. Bonin A, Bellemain E, Eidesen PB, Pompanon F, Brochmann C, et al. (2004) How to track and assess genotyping errors in population genetics studies. Mol Ecol 13: 3261–3273.
- 23. Ehrich D (2006) AFLPdat: a collection of R functions for convenient handling of AFLP data. Mol Ecol Notes 6: 603–604.
- 24. R Development Core Team (2009) A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available: http://www.R-project.org. Accessed 2012 Jul 15.
- 25. Nei M (1987) Molecular Evolutionary Genetics. New York: Columbia University Press.
- 26. Kosman E, Leonard KJ (2005) Similarity coefficients for molecular markers in studies of genetic relationships between individuals for haploid, diploid, and polyploid species. Mol Ecol 14: 415–424.
- 27. Koch MA, Dobeš C, Matschinger M, Bleeker W, Vogel J, et al. (2005) Evolution of the trnF(GAA) gene in Arabidopsis relatives and the Brassicaceae family: monophyletic origin and subsequent diversification of a plastidic pseudogene. Mol Biol Evol 22: 1032–1043.
- 28. Clement M, Posada D, Crandall KA (2000) TCS: a computer program to estimate gene genealogies. Mol Ecol 9: 1657–1659.
- 29. Templeton AR, Crandall KA, Sing CF (1992) A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 132: 619–633.
- 30. Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evol Bioinform Online 1: 47–50.
- 31. Gregorius H-R (1978) The concept of genetic diversity and its formal relationship to heterozygosity and genetic distance. Math Biosci 41: 253–271.
- 32. Ehlers J, Gibbard PL (2004) Quaternary glaciations – extent and chronology. Amsterdam: Elsevier.
- 33. Meusel H, Jäger E, Weinert E (1965) Vergleichende Chorologie der zentraleuropäischen Flora. Jena: Gustav Fischer Verlag.
- 34. Soltis DE, Buggs RJA, Doyle JJ, Soltis PS (2010) What we still don't know about polyploidy. Taxon 59: 1387–1403.
- 35. Stuessy TF, Weiss-Schneeweiss H, Keil DJ (2004) Diploid and polyploid cytotype distribution in Melampodium cinereum and M. leucanthum (Asteraceae, Heliantheae). Am J Bot 91: 889–898.
- 36. Kolář F, Štech M, Trávníček P, Rauchová J, Urfus T, et al. (2009) Towards resolving the Knautia arvensis agg. (Dipsacaceae) puzzle: primary and secondary contact zones and ploidy segregation at landscape and microgeographic scales. Ann Bot-London 103: 963–974.
- 37. Schönswetter P, Lachmayer M, Lettner C, Prehsler D, Rechnitzer S, et al. (2007) Sympatric diploid and hexaploid cytotypes of Senecio carniolicus (Asteraceae) in the Eastern Alps are separated along an altitudinal gradient. J Plant Res 120: 721–725.
- 38. Willis KJ (1996) Where did all the flowers go? The fate of temperate European flora during glacial periods. Endeavour 20: 110–114.
- 39. Petit RJ, Brewer S, Bordács S, Burg K, Cheddadi R, et al. (2002) Identification of refugia and post-glacial colonisation routes of European white oaks based on chloroplast DNA and fossil pollen evidence. Forest Ecol Manag 156: 49–74.
- 40. Heuertz M, Fineschi S, Anzidei M, Pastorelli R, Salvini D, et al. (2004) Chloroplast DNA variation and postglacial recolonization of common ash (Fraxinus excelsior L.) in Europe. Mol Ecol 13: 3437–3452.
- 41. Hewitt GM (1999) Post-glacial re-colonization of European biota. Biol J Linn Soc 68: 87–112.
- 42. Hewitt G (2000) The genetic legacy of the Quaternary ice ages. Nature 405: 907–913.
- 43. Magri D, Vendramin GG, Comps B, Dupanloup I, Geburek T, et al. (2006) A new scenario for the quaternary history of European beech populations: palaeobotanical evidence and genetic consequences. New Phytol 171: 199–221.
- 44. Ronikier M, Cieślak E, Korbecka G (2008) High genetic differentiation in the alpine plant Campanula alpina Jacq. (Campanulaceae): evidence for glacial survival in several Carpathian regions and long-term isolation between the Carpathians and the Alps. Mol Ecol 17: 1763–1775.
- 45. Frenzel B (1964) Über die offene Vegetation der letzten Eiszeit am Ostrande der Alpen. Vhdlg. Zool.-Botan. Ges. Wien Band 103/104: 110–143.
- 46. Birks HJB, Willis KJ (2008) Alpines, trees, and refugia in Europe. Plant Ecol Div 1: 147–160.
- 47. Pawłowski B (1970) Remarques sur l'endemisme dans la flore des Alpes et des Carpates. Vegetatio 21: 181–243.
- 48. Tribsch A, Schönswetter P (2003) Patterns of endemism and comparative phylogeography confirm palaeoenvironmental evidence for Pleistocene refugia in the Eastern Alps. Taxon 52: 477–497.
- 49. Niklfeld H (1972) Der niederösterreichische Alpenostrand – ein Glazialrefugium montaner Pflanzensippen. Jahrbuch des Vereins zum Schutze der Alpenpflanzen und -tiere 37: 1–54.
- 50. Zimmermann A (1972) Pflanzenareale am niederösterreichischen Alpenostrand und ihre florengeschichtliche Deutung. Dissertationes Botanicae Band 18: 1–146 (excl. Beilagen).
- 51. Schönswetter P, Paun O, Tribsch A, Niklfeld H (2003) Out of the Alps: colonization of northern Europe by east alpine populations of the glacier buttercup Ranunculus glacialis L. (Ranunculaceae). Mol Ecol 12: 3373–3381.
- 52. Fér T, Vašák P, Vojta J, Marhold K (2007) Out of the Alps or Carpathians? Origin of Central European populations of Rosa pendulina. Preslia 79: 367–376.
- 53. Koch M, Dobeš C, Bernhardt KG, Kochjarová J (2003) Cochlearia macrorrhiza (Brassicaceae): a bridging species between Cochlearia taxa from the Eastern Alps and the Carpathians? Plant Syst Evol 242: 137–147.
- 54. Cieślak E, Ronikier M, Koch MA (2007) Western Ukrainian Cochlearia (Brassicaceae) – the identity of an isolated edge population. Taxon 56: 112–118.
- 55. Hey J, Nielsen R (2007) Integration within the Felsenstein equation for improved Markov chain Monte Carlo methods in population genetics. Proc Natl Acad Sci U S A 104: 2785–2790.
- 56. Hey J (2010) Isolation with migration models for more than two populations. Mol Biol Evol 27: 905–920.