TY - JOUR
AU - Huang,, Xiao-Chen
AB - Abstract Detailed phylogenetic relationships, evolutionary histories and phylogeographical hypotheses are still quite rare for the many genera of Apiaceae. One of the reasons for this is that traditional generic circumscriptions based on morphology and anatomy largely failed to recognize the group as monophyletic. This is also true for apioid members of Apiaceae with c. 400 genera. Here we focus on Grammosciadium s.l. centred in Turkey and unravel its evolutionary history. Based on three loci from the plastid genome and nuclear-encoded internal transcribed spacer regions 1 and 2 from the nuclear encoded ribosomal RNA operon, we show that Grammosciadium in its current circumscription is not monophyletic and represents an assemblage of species from four genera (Carum, Chamaesciadium, Fuernrohria, Grammosciadium). Diversification of this group started c. 7 Mya in the late Miocene. Most present-day species arose during the Pliocene with most of the intra-species diversification occurring during the Pleistocene. Plastome type variation does not resolve any of the main clades and there is little spatial structure of the distribution of high plastid genetic variation. The majority of species of pre-Quarternary origin are found east of the Anatolian Diagonal, running diagonally across central and eastern Turkey. This might indicate that the Diagonal acted as a barrier to gene flow and migration during the Pliocene and that eastern Anatolia may have served as a cradle for Pleistocene diversification and speciation processes in Turkey. We also provide a key to the species of the revised genus and introduce several taxonomic changes. Anatolian Diagonal, Carum, evolutionary history, Fuernrohria, Grammosciadium, internal transcribed spacers 1 and 2, plastid DNA, Turkey INTRODUCTION Grammosciadium DC. is a taxonomically difficult genus of 11 taxa in tribe Careae (Apiaceae) (Spalik, Wojewodzka & Downie, 2001; Spalik & Downie 2007; Ajani et al., 2008; Zakharova, Degtjareva & Pimenov, 2012; Bani & Koch, 2015; Bani et al., 2016a;, Bani, Karakaya & Ceter, 2016b). It falls in the ‘apioid superclade’, which is the largest clade in the subfamily Apioideae with its c. 400 genera and 2900 species (Spalik & Downie, 2007; Banasiak et al., 2013). Grammosciadium is closely related to Fuernrohria K.Koch (Downie et al., 2010; Zakharova et al., 2012; Terentieva et al., 2015). Grammosciadium and Fuernrohria were usually placed in tribes Scandieae and Coriandreae, respectively, but this view is not generally accepted (e.g. Vinogradova, 1995) and various preliminary molecular analyses have supported a close relationship between Grammosciadium and Fuernrohria (Katz-Downie et al., 1999; Downie et al., 2000a, b, 2001; Valiejo-Roman et al., 2006; Terentieva et al., 2008). Carum L. has recently been shown to be polyphyletic, with its members being found in different tribes of subfamily Apioideae (Zakharova et al., 2012). A close relationship between Grammosciadium and Carum carvi L., the type species of Carum, has been demonstrated. However, mostly because of limited sampling, the monophyly of Grammosciadium has remained unresolved (Papini, Banci & Nardi, 2007; Zakharova et al., 2012). Other studies included only a few representatives of Careae, making firm conclusions about the monophyly of the genera problematic (e.g. Ajani et al., 2008; Banasiak et al., 2013). The stem group age of Careae is c. 22 Mya and the tribe has a centre of origin in the Irano-Turanian region (Banasiak et al., 2013; Calviño et al., 2016). The onset of diversification in the group of taxa studied here, comprising Grammosciadium and Fuernrohria, can be roughly estimated to 5–6 Mya based on a small taxon sampling of Careae (Banasiak et al., 2013). In general, because morphological characters often display high levels of homoplasy and phylogenetic inference is scarce (e.g. Downie et al., 2010; Zakharova et al., 2012), the taxonomy and systematics of these genera are still unresolved. Grammosciadium has been split into two subgenera and six sections, reflecting previous taxonomic concepts (Table 1). Comprehensive morphometric analyses have built upon this concept and differentiated significantly between the four taxa from section Stenodiptera (Koso-Pol.) Tamamsch. & V.M.Vinogr. [in subgenus Caropodium (Stapf & Wettst.) Tamamsch. & V.M.Vinogr.] and placed G. schischkinii (V.M.Vinogr. & Tamamsch.) V.M.Vinogr. and Caropodium armenum (Bordz.) Schischkin in synonymy of G. pterocarpum Boiss. Two new subspecies have been described in G. pterocarpum (Bani et al., 2016a). Similarly, one new subspecies was described in G. macrodon Boiss. of subgenus Grammosciadium (Bani & Koch, 2015). Table 1. Secondary calibration points (stem group ages) used in BEAST analyses of tribe Careae based on a phylogenetic analyses of 1194 apioid Apiaceae (Banasiak et al., 2013); the shape and log-scale parameters of the lognormal distribution were set to the maximum-likelihood estimates inferred from the posterior distribution of corresponding node ages obtained by Banasiak et al. (2013) Lineage . Median age (Mya) . Shape parameter, μ . Log-scale parameter, σ . Careae 16.68 2.8105663248 0.1405047315 Pyramidoptereae 17.93 2.8096345498 0.1404215425 Selineae 23.67 3.1623673515 0.0848250429 Oenantheae 25.66 3.2448813202 0.0838900841 Lineage . Median age (Mya) . Shape parameter, μ . Log-scale parameter, σ . Careae 16.68 2.8105663248 0.1405047315 Pyramidoptereae 17.93 2.8096345498 0.1404215425 Selineae 23.67 3.1623673515 0.0848250429 Oenantheae 25.66 3.2448813202 0.0838900841 Open in new tab Table 1. Secondary calibration points (stem group ages) used in BEAST analyses of tribe Careae based on a phylogenetic analyses of 1194 apioid Apiaceae (Banasiak et al., 2013); the shape and log-scale parameters of the lognormal distribution were set to the maximum-likelihood estimates inferred from the posterior distribution of corresponding node ages obtained by Banasiak et al. (2013) Lineage . Median age (Mya) . Shape parameter, μ . Log-scale parameter, σ . Careae 16.68 2.8105663248 0.1405047315 Pyramidoptereae 17.93 2.8096345498 0.1404215425 Selineae 23.67 3.1623673515 0.0848250429 Oenantheae 25.66 3.2448813202 0.0838900841 Lineage . Median age (Mya) . Shape parameter, μ . Log-scale parameter, σ . Careae 16.68 2.8105663248 0.1405047315 Pyramidoptereae 17.93 2.8096345498 0.1404215425 Selineae 23.67 3.1623673515 0.0848250429 Oenantheae 25.66 3.2448813202 0.0838900841 Open in new tab With the exception of these studies, there are no reliable quantitative morphometric studies that allow definition of species boundaries in Grammosciadium and comparison with taxa from related genera. Although Zakharova et al. (2012) were not able to prove para- or polyphyly of Grammosciadium, they proposed a ‘clade A’ with ‘Carum carvi and its allies’. Besides C. carvi, this clade was shown to include four other Carum species, one species of Chamaesciadium C.A.Mey. from the Caucasus and adjacent regions, Grammosciadium and Fuernrohria. Of the five investigated Grammosciadium spp. four clustered together, namely G. scabridum Boiss., G. macrodon, G. daucoides DC. and G. platycarpum Boiss. & Hausskn. Grammosciadium pterocarpum was placed in a basal polytomy with Fuernrohria and C. carvi. However, we will show later that in their analysis, G. platycarpum was wrongly determined by Zakharova et al. (2012; AF073551) and actually represents G. macrodon. Close affinities between Chaemaesciadium and Carum were proposed earlier (e.g. Koso-Poljansky, 1914, 1916; Rechinger, 1987) and are indeed supported by morphological characters. Fuernrohria and Grammosciadium are also similar to each other in leaf and flower characters (Vinogradova, 1995). However, phylogenetic placement of these two genera with ‘Carum carvi and its allies’ was commented upon as follows: ‘… the inferred affinity … with Carum is surprising and they could by no means be united with the latter’ (Zakharova et al., 2012). The Irano-Turanian phytogeographical region (IT region) is well known for its floristic species richness (Takhtajan, 1986; Manafzadeh, Staedler & Conti, 2017) and is characterized by levels of endemism exceeding 25% (e.g. Davis, 1971; Zohary, 1971; Takhtajan, 1986; Koch, Karl & German, 2017). Takhtajan (1986) concluded that even the number of endemic genera from the apioid super clade is > 60, therefore representing 16% of all IT endemic genera, with Grammosciadium among them. Grammosciadium is distributed across Inner Anatolia, the East Anatolian mountains, the mountains of Transcaucasica and the Elburz and Zagros ranges in Iran and Iraq, respectively (Vinogradova, 1970). In addition, G. scabridum, which was thought to be restricted to Iran and Iraq, was recently documented in Turkey (Behçet, Kaval & Rüstemoğlu, 2012). However, critical investigations of the specimens collected from this locality (Behçet et al., 2012) showed that they have young fruits only, and the diagnostic character (winged fruits) used to distinguish between G. scabridum and G. platycarpum could not be scored. Therefore, we conclude that G. scabridum is still not unambiguously shown to be distributed in Turkey. Thus, the genus in its traditional circumscription is a western IT element restricted to the Anatolian Plateau. As such, Grammosciadium is highly suited for study of the floristically defined division of Anatolia, the so-called Anatolian ‘Diagonal’ (Davis, 1971; Ekim & Güner, 1986; Bilgin, 2011), in a spatio-temporal context. This diagonal region runs from the north-eastern corner of the Mediterranean Sea to the south-eastern corner of the Black Sea and it divides the Anatolian phytogeographical subregion into western and eastern Anatolia. This pattern was first studied in detail by Davis (1971) and, of 550 species considered, 61% of these followed a western versus eastern distribution pattern. The Diagonal itself represents an important refuge area for many endemic species. Attempts to review and classify spatio-temporal patterns for this divide at intra-specific levels have been made (Bilgin, 2011), but no general pattern was observed, the authors concluding that isolation processes during Pleistocene climatic fluctuations and subsequent genetic differentiation is the most over-arching process of intraspecific evolution in Anatolia. It was also concluded, however, that there is evidence for Anatolia being a large general hotspot of genetic diversity from the continental perspective (Bilgin, 2011). Accordingly, in studies of animals and plants including age calculations, intraspecific diversification processes are placed in the Pliocene or Pleistocene with a 1:1 ratio (referring to reviewed studies cited in Bilgin, 2011). Here, we analyse a comprehensive sampling of Grammosciadium spp. and accessions across the distribution range in Anatolia (western IT region) and sequenced the nuclear encoded internal transcribed spacer (ITS) regions 1 and 2 of ribosomal DNA for subsequent phylogenetic analyses. The results were used to test for monophyly of Grammosciadium, to elaborate on a temporal evolutionary scenario and to estimate divergence times within the framework of previously published studies (Banasiak et al., 2013). Three DNA regions from the plastid genome were sequenced to compare maternally inherited genetic variation with results from ITS-based phylogenetic analysis to infer spatial patterns of genetic variation and to develop a first phylogeographical scenario and introducing Anatolia as a cradle of species biodiversity. MATERIAL AND METHODS Plant material and taxon sampling This study takes advantage of two former phylogenetic, phylogeographical studies. A comprehensive molecular–systematic analysis of the polyphyletic genus Carum was performed recently (Zakharova et al., 2012) and included Grammosciadium spp. Phylogenetic results obtained in their contribution are congruent with recent biogeographical findings presented for Apiaceae (Banasiak et al., 2013). Consequently, we used all accessions and species with available DNA sequence information for ITS (internal transcribed spacers 1 and 2 of nuclear ribosomal RNA) included by Zakharova et al. (2012) for phylogenetic reconstructions and calibrated phylogenetic trees using information provided by Banasiak et al. (2013). During our analyses we identified two potentially erroneously determined accessions from GenBank: AH008898 was published as ‘G. scabridum’ and misidentified. There are two labels on the same voucher sheet. It was initially identified as G. platycarpum, and later as G. scabridum. The specimen has immature fruits, and there are no petals. Thus it is impossible to conclude that it is G. scabridum. However, the specimen resembles perfectly G. macrodon subsp. macrodon or G. cornutum (Nábělek) C.C.Townsend. The second accession, AH008896, clearly resembles G. macrodon. We checked the respective voucher and there is no doubt about the identity of this taxon. In addition, a representative sampling of the various species and subspecies of Grammosciadium was collected in the wild over the past few years and was analysed here (Fig. 1). We also sequenced representatives of newly sampled C. carvi, Falcaria vulgaris Bernh. and Fuernrohria setifolia K.Koch to compare newly obtained DNA sequence variation with already existing database entries (AF077878 and JQ792211, U78378 and U78438, AF008633 and AF009221, respectively) and to check for our data quality. In summary, 88 individuals from 56 populations were analysed. Most of the accessions from Grammosciadium have been studied earlier morphologically (Bani & Koch, 2015; Bani et al., 2016a, b). Detailed accession information is provided in Supporting Information Table S1. All Grammosciadium vouchers have been deposited at ANK under the indicated collection numbers. Figure 1. Open in new tabDownload slide Distribution of analysed Grammosciadium and Fuernrohria accessions from Turkey and the Near East. Accession codes are indicated and correspond to Supporting Information Table S1. Two distribution ranges are indicated. The western range is partly congruent with the Anatolian Diagonal, which runs diagonally across central and eastern Turkey from the north-eastern corner of the Mediterranean Sea to the south-eastern corner of the Black Sea. Figure 1. Open in new tabDownload slide Distribution of analysed Grammosciadium and Fuernrohria accessions from Turkey and the Near East. Accession codes are indicated and correspond to Supporting Information Table S1. Two distribution ranges are indicated. The western range is partly congruent with the Anatolian Diagonal, which runs diagonally across central and eastern Turkey from the north-eastern corner of the Mediterranean Sea to the south-eastern corner of the Black Sea. DNA extraction, amplification and sequencing Total DNA was extracted using the Invisorb Spin Plant Mini Kit (Stratec Biomedical AG, Birkenfeld, Germany). PCR amplification of the markers used (nuclear ITS, plastid trnL intron and trnL-trnF intergenic spacer) was performed in a volume of 25 μL, using 10 μM each primer, respectively, 2.0 mM MgCl2 and 0.5 U Mango-Taq polymerase (Bioline, Luckenwalde, Germany). The primers used for ITS amplification were originally designed by White et al. (1990) with some modifications (18F: 5′-GAAAGGAGAAGTGCTAACAAGA-3′, 25R: 5′-GGGTAATCCCGCCTGACCTGG-3′). For some fragments, PCR quality was further increased to amplify ITS1 and ITS2 separately using internal primers ITS2a (5′-GCTGCGTTCTTCATCGATGC-3′) and ITS3 (5′-GCATCGATGAAGAACGTAGC-3′), respectively. Amplification of the trnL intron and the trnL-trnF intergenic spacer was performed using primers c, d and e of Taberlet et al. (1991) (c: 5′ -CGAAATCGGTAGACGCTACG-3′, d: 5′ -GGGGATAGAGGGACTTGAAC-3′, and e: 5′ -GGTTCAAGTCCCTCTATCATCCC-3′) and a primer designed by Dobeš, Mitchell-Olds & Koch (2004) (5′ -GATTTTCAGTCCTCTGCTCTAC-3′). The plastid rpl16 intron was amplified using primers rpl16-rev (5′-TCTTCCTCTATGTTGACG-3′) and rpl17-for (5′-AATAATCGCTATGCTTAGTG-3′). All primers were extended by the M13 sequence for subsequent sequencing using M13 universal sequencing primers. The amplifications were run on a PTC 200 Peltier thermal cycler (MJ Research, Waltham, MA, USA) under the following conditions for ITS: 3 min initial denaturation at 95 °C; 30 cycles of amplification with 30 s at 95 °C, 30 s at 44 °C and 1 min at 72 °C; and 5 min of final elongation at 72 °C. For all three plastid loci annealing temperature was increased to 50 °C. PCR success was checked with electrophoresis in a 1% agarose gel in TAE-buffer. PCR product clean-up was executed using the Wizard SV Gel and PCR Clean-Up System (Promega, Madison, WI, USA). Custom Sanger-sequencing was performed at GATC Biotech (Konstanz, Germany). The electropherograms were checked and trimmed to the borders of the analysed markers using the program SeqMan DNA-Star Lasergene software package (DNASTAR, Madison, WI, USA). Chromosome number evaluation The seeds which were used for karyological studies were selected from following the accessions: G. daucoides (B.Bani 6982), G. macrodon subsp. macrodon (B.Bani 6887), G. macrodon subsp. nezaketiae B.Bani (B.Bani 6868), G. cornutum (B.Bani 6863), G. confertum Hub.-Mor. & Lamond (B.Bani 6890), G. haussknechtii Boiss. (B.Bani 6903), G. platycarpum (B.Bani 6951) and G. pterocarpum subsp. pterocarpum (B.Bani 6885, B.Bani 6966, B.Bani 6872). Seeds were obtained from as many individuals as possible to prepare bulked seed samples. Seeds were germinated at 4 °C (up to 60 days for germination). Root tips were harvested after the night phase under optimal growth conditions with low starch content and incubated in 8-hydroxyquinoline (0.2 g/L) for 4 h at room temperature and washed afterwards with distilled water. Fixation performed with a 3:1 mixture of absolute ethylalcohol/glacial acetic acid at room temperature for 24 h. After washing with distilled water, root tips were transferred into 1 m HCI solution at 60 °C for 10–12 min and washed again with distilled water. Finally root tips were stained with 2% acetic orcein for 1 h with subsequent washing with water for 5 min. Prepared root tips were squashed, prepared for microscopical inspection (embedding in Entellan), and analysed using a Leica DFC295 camera attached to a Leica DM3000 microscope. At least five metaphase plates were measured for each prepared sample. Analysis of chromosome number evolution was performed with the program ChromEvol vers. 2.0 (Mayrose, Barker & Otto, 2010) and using an unconstrained maximum-likelihood tree of an ITS dataset comprising taxa from Careae (see also paragraph on phylogenetic analyses). Published chromosome numbers were taken from the following records (Retina & Pimenov, 1977; Löve & Löve in Löve, 1982; Pimenov & Vasilieva, 1983; Davlianidze, 1985; Daushkevich, Alexeeva & Pimenov, 1991, 1995; Vasilieva, Alexeeva & Pimenov, 1994; Pimenov et al., 1996; Pimenov, Alexeeva & Kljuykov, 1998; Kiehn et al., 2000; Nazarova & Ghukasyan, 2004; Shner, 2004; Shner et al., 2004). Editing and sequence alignment New ITS DNA sequence data were added to a large and tribal-wide dataset (Zakharova et al., 2012). The alignment was prepared by MAFFT (Katoh & Standley, 2013) using L-INS-I strategy and further manually adjusted in PhyDE (http://www.phyde.de). The final alignment consists of 503 bp (5.8S rDNA excluded) from 225 accessions and is available with the online material and includes GenBank accession codes for already published data (Supporting Information Table S2). For the BEAST analysis the alignment has been reduced and all identical sequences have been removed, resulting in a dataset comprising 173 accessions (ribotypes). Plastid DNA sequences from the trnL intron, trnL-trnF intergenic spacer and rpl16 intron from Grammosciadium and its closest relatives were aligned separately using PhyDE and finally combined into a single concatenated alignment of 1943 bp (finally reduced to 1742 bp because of exclusion of ambiguous or missing sequence data). Alignment lengths of the trnL intron, trnL-trnF intergenic spacer and rpl16 intron were 622 bp (reduced to 566 bp), 396 bp (reduced to 366 bp) and 925 bp (reduced to 808 bp), respectively. The nexus input file is provided with Supporting Information Table S3. Phylogenetic analyses and tree calibration For ITS phylogenetic analyses, we performed maximum-likelihood (ML) and Bayesian inference (BI) analyses. The best-fitting nucleotide substitution model (GTR+I+G) was selected using MrModeltest 2.3 (Nylander, 2004), according to the Akaike information criterion (AIC). ML analyses were performed in RAxML (Stamatakis, 2014) implemented in raxmlGUI (Silvestro & Michalak, 2012), with the search strategy set to rapid bootstrapping. Clade support was evaluated by bootstrap analysis of 1000 replicates. In the Bayesian analyses using MrBayes v. 3.2.6 (Ronquist et al., 2012), four simultaneous runs with four chains each were run for 20 million generations, sampling every 1000 trees. The first 25% of these trees were discarded as burn-in when computing the consensus tree (50% majority rule). For efficient swapping of the chains, the temperature was set to 0.01. Sufficient mixing of the chains was considered to be reached when the average standard deviation of split frequencies was below 0.01. Stationarity of the Markov chains was also confirmed in Tracer (http://tree.bio.ed.ac.uk/software/tracer/) and reliable effective sample size values (> 200) were ensured. From the plastid DNA alignment 1943 bp in length we excluded positions 1–56, 993–1022 and 1831–1947 from subsequent analyses, because of missing sequence information from several accessions and, more importantly, because of numerous polyT-stretches and an accumulation of indels in these regions. ML and Bayesian analyses have been performed as described above using Falcaria vulgaris as an outgroup (a taxon found in an early-branching clade of Careae). However, these analyses did not produce any strongly supported phylogenetic trees (data not shown) and consequently here we show results from a SplitsTree analysis using SplitsTree4 vers. 4.14.4 (Huson & Bryant, 2006). Gaps were treated as missing characters and were not coded separately. A NeighborNet was calculated using uncorrected p-distances. Since F. vulgaris plastid DNA sequence information was distant compared to the other taxa, the analysis was repeated without F. vulgaris and was restricted to C. carvi, F. setifolia and all Grammosciadium plastid DNA sequence types. However, the topology of the networks did not change. Bootstrap analysis was performed running 1000 iterations. Reconstruction of a dated tree is problematic because macrofossil evidence for primary node calibration is missing. This issue has been intensively discussed for Apiaceae by Banasiak et al. (2013). Here we followed their approach using remnants of microfossil pollen for tree calibration. Pollen attributed to Pleurospermeae was documented from the Priabonian (38.0–33.9 Mya; Gruas-Cavagnetto & Cerceau-Larrival, 1984). Therefore, the first calibration point was placed at the stem node of Pleurospermeae and constrained to a log-normal distribution with a lower bound (offset) of 33.9 Mya. The upper bound was set to 55.8 Mya using the confidence intervals and posterior density shown by Banasiak et al. (2013). Secondary calibration points (median) were set at the stem nodes of Careae of 16.68 Mya, Pyramidoptereae of 17.93 Mya, Pimpinelleae/Selineae of 23.67 Mya and Oenantheae (excluding Trocdaris Raf.) of 25.66 Mya (Table 1). The standard deviation of the lognormal distribution was set to 0.5. The shape and log-scale parameters of the lognormal distribution were set to the ML estimates inferred from the posterior distribution of ages of corresponding nodes obtained by Banasiak et al. (2013). The results presented by Banasiak et al. (2013) are also based on ITS analysis, and we compared all posterior densities for the calibration points between the studies, which makes it a good control for estimating divergence times in general. The shape and log-scale parameters of these distributions were inferred through fitting to the posterior distributions for the chosen nodes obtained in the previous analyses (Banasiak et al., 2013). These calculations were performed using SPSS vers. 22 (Released 2013, IBM SPSS Statistics for Windows, Version 22.0; IBM Corp., Armonk, NY, USA). To test the significance of monophyly of Grammsciadium we ran an ML analysis with a subset of the ITS alignment (Careae) with and without constraining Grammosciadium into monophyly and compared respective analyses using CONSEL software (Shimodaira & Hasegawa, 2001). RESULTS Taxon identity and monophyletic groups in Careae Phylogenetic analyses of the family-wide ITS-based datasets on ML and BI are fully congruent with each other (Supporting Information Figs S1, S2) and are also fully in congruence with previous phylogenetic studies (e.g. Zakharova et al., 2012; Banasiak et al., 2013). The redrawn results from ML analysis (based on Supporting Information Fig. S1) are shown in Figure 2 with both ML bootstrap and Bayesian posterior values indicated. In Careae, we were not able to resolve any deeper nodes significantly, but the monophyly of Grammosciadium was not supported. Grammosciadium confertum grouped consistently with Fuernrohria setifolia, and Carum carvi was nested between the two other major Grammosciadium clades. In ML analyses of Careae only and comparing results from constrained (H2) (monophyly of Grammsciadium) and unconstrained (H1) analyses the topology of the unconstrained phylogenetic tree is identical to the family-wide ML tree. In the constrained (monophyly of Grammosciadium) ML tree the entire structure of relationships among Grammosciadium spp., C. carvi, Fuernrohria and other Carum spp. disappeared, and we revealed a large and increased polytomy; generally bootstrap values are c. 30% or less. The tree structure of the unconstrained analysis (H1) has a higher likelihood (−lnL = 3004.176) than that of the constrained analysis (H2) (−lnL = 3007.533). This result is consistent with an approximate Bayesian posterior probability test (H1/H2: P = 0.967/0.033), which also rejects the hypothesis of H2 with P < 0.05. Figure 2. Open in new tabDownload slide Phylogenetic tree based on ITS data of tribe Careae. Redrawn from the apioid-wide ML analysis shown with Supporting Information Figure S1 (indicated in red with the cartoon in the upper left). ML bootstrap (left) and Bayesian posterior values (right) are provided if > 75%. Chromosome numbers have been collected from the literature and are indicated for respective taxa. Colour codes of taxa follow Figure 1. Figure 2. Open in new tabDownload slide Phylogenetic tree based on ITS data of tribe Careae. Redrawn from the apioid-wide ML analysis shown with Supporting Information Figure S1 (indicated in red with the cartoon in the upper left). ML bootstrap (left) and Bayesian posterior values (right) are provided if > 75%. Chromosome numbers have been collected from the literature and are indicated for respective taxa. Colour codes of taxa follow Figure 1. In all phylogenetic trees one clade consists of G. daucoides (including G. scabridum), G. macrodon and G. cornutum. ITS is able to separate the recently described subspecies G. macrodon subsp. nezaketiae along a basal polytomy from G. macrodon subsp. macrodon (Bani & Koch, 2015) and this might indicate that these can be also treated at the species level. BEAST analysis even recognizes G. macrodon subsp. nezaketiae as a monophyletic group (Fig. 4), but the Bayesian posterior value is < 50%. The second clade consists of G. platycarpum and a group of taxa combining G. pterocarpum, G. haussknechtii and G. schischkinii (accessions 6820, 6925, 6929 and 6932; Fig. 1), which has been shown to be largely indistinguishable from G. pterocarpum and has been synonymized accordingly (Bani et al., 2016a). Accessions of G. pterocarpum subsp. sivasicum (accession 7255) analysed here have shared ITS and plastid types, supporting previous morphological results. Also G. pterocarpum subsp. bilgilii (one accession analysed, accession 6965) has a unique ITS type, therefore supporting morphological evidence. Genetic variation among accessions within the various species or subspecies is extremely low and ITS was the only marker which could serve as a suitable DNA barcode of high efficiency. This is congruent with an Apiaceae-wide evaluation of DNA barcodes showing a 73.3% identification efficiency of ITS (Liu et al., 2014). In agreement with morphological data, ITS data fail to differentiate between G. schischkinii and G. pterocarpum; G. scabridum (accession 8631) shares identical ITS sequences with some G. daucoides accessions (Fig. 2). Plastid DNA sequence data support some of the results from the ITS analysis (Fig. 3). Fuernrohria setifolia is placed closest to G. confertum, and C. carvi is nested in Grammosciadium closer to G. platycarpum and the G. cornutum/daucoides/macrodon accessions. Although G. schischkinii is morphologically challenging (considering its type specimen), it is not distinct from G. pterocarpum as a whole (Bani et al., 2016a). We found that some populations with haplotypes E2–E6 and E8 (Fig. 3) are separated from other plastid types of G. pterocarpum accessions. These haplotypes show a strong biogeographical signal (Fig. 3, north-eastern group) and are from the same region as the type specimen of G. schischkinii (39°31′11″N, 42°48′15″E). Grammosciadium pterocarpum subsp. sivasicum has its own distinct haplotype, but all remaining plastid types cannot be attributed significantly to certain species or subspecies except for single accessions analysed here (G. haussknechtii with unique haplotype F17). Consistent with the ITS analysis, G. scabridum (accession 8631) shares a plastid haplotype with G. daucoides. However, the entire DNA sequence data set (ITS and plastid data) contains also conflicting results and all of them affect G. pterocarpum/G. schischkinii. Grammosciadium schischkinii (accession 6929) is placed in the plastid DNA network with G. platycarpum (although with its ITS it is placed with other G. schischkinii) and G. pterocarpum accession 6914 carries another unique plastid type (A5; unrelated to G. pterocarpum/G. schischkinii plastid types). This supports the conclusion that in its previous circumscription G. schischkinii does not exist as a distinct taxon. However, we also do not have any further significant evidence, for example for introgression and hybridization among different species, that might account for this finding and future detailed population-based genetic studies might unravel these aspects. Figure 3. Open in new tabDownload slide Parsimony SplitsTree graph for plastid DNA variation in Grammosciadium, Fuernrohria and Carum carvi. Colour code of accessions follows Figure 1. Bootstrap values from 1000 replicates are all < 75% and are not shown. Figure 3. Open in new tabDownload slide Parsimony SplitsTree graph for plastid DNA variation in Grammosciadium, Fuernrohria and Carum carvi. Colour code of accessions follows Figure 1. Bootstrap values from 1000 replicates are all < 75% and are not shown. The genetic data do not show any obvious geographical patterns (data not shown), and there is only limited biogeographical substructure following genetically defined taxonomic groups (Fig. 1). G. macrodon subsp. nezaketiae is restricted to the eastern range (largely congruent to the Anatolian Diagonal). In addition, a genetically confirmed subclade of G. pterocarpum accessions (plastid types) is found here. Endemic taxa for the eastern range are Fuernrohria setifolia and G. scabridum. In the western range, we found only genetically defined G. confertum and the endemic G. haussknechtii. The most meaningful phylogeographical splits are between (1) G. confertum (West) versus its sister-species Fuernrohria setifolia (East), and (2) G. macrodon subsp. macrodon (West) versus G. macrodon subsp. nezaketiae (East), which might not be sister taxa. The remaining taxa, G. daucoides and G. platycarpum, show a wide distribution, but differing from each other, because G. platycarpum is the genetically better defined taxon (ITS and chloroplast DNA). Divergence time estimates Our divergence time estimates are fully congruent with those estimated earlier at the family level (Banasiak et al., 2013). Divergence time estimates of tribe Careae are presented in Figure 4 (full information is shown with Supporting Information Fig. S3); and we also show the graphs comparing posterior densities obtained in this study for secondary calibration points (blue bars) plotted against the posterior densities obtained in the original large-scale and fossil-calibrated data (red bars; Banasiak et al., 2013). The approach we use has been adopted from Spalik et al. (2014). Comparison of the analyses indicates not only nearly identical median split time values, but also that the distribution of posterior densities is congruent. We take this as evidence that divergence time estimates presented with our study for Grammosciadium diversification are consistent with previous studies. The phylogenetic hypothesis based on BEAST analysis is also congruent with ML and Bayesian analysis with one exception, and the deepest split among the various Grammosciadum clades/groups is calculated to 7.77 Mya. Stem group (split between Careae and Pyramidoptereae) and crown group ages of Careae are 24.43 and 18.97 Mya, respectively. Stem group ages of the Grammosciadium clades (1) G. daucoides/scabridum (2.77 Mya), (2) G. cornutum (2.32 Mya), (3) G. macrodon (3.21 Mya), (4) G. platycarpum (5.33 Mya), (5) G. pterocarpum (5.33 Mya) and (6) G. confertum (4.73 Mya) are all in the Pliocene. Among these splits the geographically meaningful phylogeographical splits of (i) G. confertum (west) versus Fuernrohria setifolia (east), and (ii) G. macrodon subsp. macrodon (west) versus G. macrodon subsp. nezaketiae (east) are at c. 4.73 and 3.21 Mya, respectively. Consequently all splits between species/subspecies date back to the Pliocene, and all further splits dated to the Pleistocene are found within species-specific ITS lineages. The aforementioned difference in phylogenetic signal among BEAST and ML/BI indicates that the Fuernrohria/G. confertum clade is placed sister to Grammosciadium including C. carvi, some other Carum spp. and Chamaesciadium. Figure 4. Open in new tabDownload slide Maximum clade credibility tree resulting from BEAST analysis under an uncorrelated lognormal molecular clock. The figure shows part of the tree as indicated in red lines with the cartoon in the lower left (entire tree is presented in Supporting Information Fig. S3). Node ages are represented by median heights and the 95% posterior density interval is shown. Circles with numbers (1–4) denote secondary calibration points, and circle X denotes the microfossil pollen used as the primary calibration. Posterior densities of calibration points obtained in this study (blue line) are plotted against posterior densities obtained in the original calibration using this and additional fossil data (red bars and red line; Banasiak et al., 2013). Figure 4. Open in new tabDownload slide Maximum clade credibility tree resulting from BEAST analysis under an uncorrelated lognormal molecular clock. The figure shows part of the tree as indicated in red lines with the cartoon in the lower left (entire tree is presented in Supporting Information Fig. S3). Node ages are represented by median heights and the 95% posterior density interval is shown. Circles with numbers (1–4) denote secondary calibration points, and circle X denotes the microfossil pollen used as the primary calibration. Posterior densities of calibration points obtained in this study (blue line) are plotted against posterior densities obtained in the original calibration using this and additional fossil data (red bars and red line; Banasiak et al., 2013). Chromosome number variation Chromosome number reports are scarce among Careae. However, there is convincing evidence that the group of taxa analysed here (Carum, Chamaesciadium, Fuernrohria, Grammosciadium) that started to radiate (crown group age) c. 19 Mya is characterized by an ancestral shift from 2n = 22 to 2n = 20 chromosomes (Fig. 2) (results from ChromEvol analysis, new chromosome counts provided with Supporting Information Table S1). It is only Fuernrohria that has been reported with 2n = 22 chromosomes, but with good evidence for a sister-relationship with G. confertum. The results of the BEAST analysis are congruent with the chromosome number data, because Fuernrohria/G. confertum is placed as sister to all 2n = 20 taxa (Fig. 4). All remaining taxa of Careae show the ancestral karyotype of Careae with 2n = 22 with two known exceptions, Aegopodium komarovii (Karjagin) Pimenov & Zakharova and A. decumbens (Thunb. ex J.A.Murray) Pimenov & Zakharova. (Fig. 2). The results from our own ongoing chromosome number counts differ from the literature for G. platycarpum and G. pterocarpum (Table 2) and indicate parallel chromosome number reduction even below 2n = 20 in the majority of Grammosciadium spp. Table 2. Traditional taxonomy and systematics of Grammosciadium s.l. and Fuernrohria; diploid chromosome numbers are indicated with literature data given in parentheses Taxon . . Chromosome number . Grammsciadium DC. Subgenus Grammosciadium Section Grammosciadium G. daucoides DC. 20 (20) G. scabridum Boiss. Section Macrodon Koso-Pol. G. macrodon Boiss. subsp. macrodon 18 G. macrodon subsp. nezaketiae Bani 18 Section Ceratodon Tamamsch. & V.M.Vinogr. G. cornutum (Nábělek) C.C.Towns. 18 Section Heterocarpum V.M.Vinogr. G. confertum Hub.-Mor. & Lamond 18 Subgenus Caropodium (Stapf & Wettst.) Tamamsch. & V.M.Vinogr. Section Caropodium G. platycarpum Boiss. & Hausskn. 18 (20) Section Stenodiptera (Koso-Pol.) Tamamsch. & V.M.Vinogr. G. haussknechtii Boiss. 16 G. pterocarpum Boiss. subsp. pterocarpum [including Caropodium armenum (Bordz.) Schischkin, G. schischkinii (V.M.Vinogr. & Tamamsch) V.M.Vinogr.] 18 (20) G. pterocarpum subsp. bilgilii Bani G. pterocarpum subsp. sivasicum Bani 18 Fuernrohria K.Koch F. setifolia K.Koch 22 Taxon . . Chromosome number . Grammsciadium DC. Subgenus Grammosciadium Section Grammosciadium G. daucoides DC. 20 (20) G. scabridum Boiss. Section Macrodon Koso-Pol. G. macrodon Boiss. subsp. macrodon 18 G. macrodon subsp. nezaketiae Bani 18 Section Ceratodon Tamamsch. & V.M.Vinogr. G. cornutum (Nábělek) C.C.Towns. 18 Section Heterocarpum V.M.Vinogr. G. confertum Hub.-Mor. & Lamond 18 Subgenus Caropodium (Stapf & Wettst.) Tamamsch. & V.M.Vinogr. Section Caropodium G. platycarpum Boiss. & Hausskn. 18 (20) Section Stenodiptera (Koso-Pol.) Tamamsch. & V.M.Vinogr. G. haussknechtii Boiss. 16 G. pterocarpum Boiss. subsp. pterocarpum [including Caropodium armenum (Bordz.) Schischkin, G. schischkinii (V.M.Vinogr. & Tamamsch) V.M.Vinogr.] 18 (20) G. pterocarpum subsp. bilgilii Bani G. pterocarpum subsp. sivasicum Bani 18 Fuernrohria K.Koch F. setifolia K.Koch 22 Open in new tab Table 2. Traditional taxonomy and systematics of Grammosciadium s.l. and Fuernrohria; diploid chromosome numbers are indicated with literature data given in parentheses Taxon . . Chromosome number . Grammsciadium DC. Subgenus Grammosciadium Section Grammosciadium G. daucoides DC. 20 (20) G. scabridum Boiss. Section Macrodon Koso-Pol. G. macrodon Boiss. subsp. macrodon 18 G. macrodon subsp. nezaketiae Bani 18 Section Ceratodon Tamamsch. & V.M.Vinogr. G. cornutum (Nábělek) C.C.Towns. 18 Section Heterocarpum V.M.Vinogr. G. confertum Hub.-Mor. & Lamond 18 Subgenus Caropodium (Stapf & Wettst.) Tamamsch. & V.M.Vinogr. Section Caropodium G. platycarpum Boiss. & Hausskn. 18 (20) Section Stenodiptera (Koso-Pol.) Tamamsch. & V.M.Vinogr. G. haussknechtii Boiss. 16 G. pterocarpum Boiss. subsp. pterocarpum [including Caropodium armenum (Bordz.) Schischkin, G. schischkinii (V.M.Vinogr. & Tamamsch) V.M.Vinogr.] 18 (20) G. pterocarpum subsp. bilgilii Bani G. pterocarpum subsp. sivasicum Bani 18 Fuernrohria K.Koch F. setifolia K.Koch 22 Taxon . . Chromosome number . Grammsciadium DC. Subgenus Grammosciadium Section Grammosciadium G. daucoides DC. 20 (20) G. scabridum Boiss. Section Macrodon Koso-Pol. G. macrodon Boiss. subsp. macrodon 18 G. macrodon subsp. nezaketiae Bani 18 Section Ceratodon Tamamsch. & V.M.Vinogr. G. cornutum (Nábělek) C.C.Towns. 18 Section Heterocarpum V.M.Vinogr. G. confertum Hub.-Mor. & Lamond 18 Subgenus Caropodium (Stapf & Wettst.) Tamamsch. & V.M.Vinogr. Section Caropodium G. platycarpum Boiss. & Hausskn. 18 (20) Section Stenodiptera (Koso-Pol.) Tamamsch. & V.M.Vinogr. G. haussknechtii Boiss. 16 G. pterocarpum Boiss. subsp. pterocarpum [including Caropodium armenum (Bordz.) Schischkin, G. schischkinii (V.M.Vinogr. & Tamamsch) V.M.Vinogr.] 18 (20) G. pterocarpum subsp. bilgilii Bani G. pterocarpum subsp. sivasicum Bani 18 Fuernrohria K.Koch F. setifolia K.Koch 22 Open in new tab DISCUSSION Phylogenetics, taxonomy and morphology Although the phylogenetic trees presented here do not significantly resolve any deeper nodes among Grammosciadium evolutionary lineages (Figs 2, 4), we can conclude that there are three strongly supported lineages comprising (i) G. daucoides, G. scabridum, G. cornutum and G. macrodon, (ii) G. pterocarpum, G. haussknechtii and G. platycarpum, with C. carvi less significantly joined with this second clade, and (iii) G. confertum and Fuernrohria. This makes Grammosciadium a polyphyletic group and indicates the need for taxonomic changes rendering Grammosciadium monophyletic. Since C. carvi is the type of Carum and since Fuernrohria is a monotypic and morphologically defined genus, we suggest considering lineage (i) as a newly circumscribed Grammosciadium s.s. Grammosciadium taxa from lineage (ii) have been previously grouped into subgenus Caropodium (Table 2), the name initially introduced as a genus long ago (Stapf, 1886) with some binominals already available (C. haussknechtii, C. platycarpum, C. pterocarpum). To reflect the sister-relationship of Fuernrohria and G. confertum in lineage (iii), G. confertum has to be newly defined generically and we thus propose the genus Vinogradovia based on Grammosciadium section Heterocarpum. The name Heterocarpum has been used earlier for the sectional classification of this single species (Vinogradova, 1995) but it cannot be used as a name of a genus for a nomenclatural reason (homonymy). The details are given in the ‘Taxonomic conclusions’ paragraph below. In light of this new generic classification, it is worth discussing cytological, morphological and anatomical characters. Chromosome numbers do not provide any congruent result. Respective members of Caropodium have diploid chromosome numbers of 16, 18 and 20 (considering also data from the literature). The newly circumscribed genus Grammosciadium in this study comprises representatives containing 2n = 18 and 20. Also, Vinogradovia (G. confertum) with 2n = 18 differs from its sister-species Fuernrohria with 2n = 22. The most diagnostic characters for distinguishing taxa of Grammosciadium and Fuernrohria concern fruit characters (morphological and anatomical) and flower characteristics. These characters are summarized in Supporting Information Table S4. The close relationship between G. confertum and Fuernrohria is supported by an anatomical character, namely topology and arrangement of vascular bundles in transverse sections of fruit mericarps. In contrast to all other species, the bundles are continuous and hard to separate from each other (arranged as a ring: Vinogradova, 1995; Bani, Mavi & Adigüzel, 2011; Bani & Koch, 2015). The phylogenetic hypotheses derived from ML/Bayesian and BEAST analyses provide, with little support, two alternative relationships for C. carvi with respect to Grammosciadium and Caropodium. ML/Bayesian analyses suggest a relationship with (Grammosciadium(Carum carvi, Caropodium)) (scenario 1), whereas the BEAST analysis indicates a relationship with (Carum carvi(Grammosciadium, Caropodium)) (scenario 2). The second scenario is more consistent with traditional classifications, but based on morphological characters used for delimitation of the various species of Grammosciadium (fully supported by molecular analysis), neither of the hypotheses is more intuitive. Discrete characters such as number of vascular bundles in fruit mericarps (except G. cornutum), mericarp surface structure and occurrence of winged fruits and continuous characters such as fruit width and presence of sepals distinguish Grammosciadium and Caropodium (e.g. Bani & Adigüzel, 2010; Bani & Koch, 2015; Bani et al., 2016a, b), but they do not allow us to make any firm conclusion about the phylogenetic–systematic position of Carum carvi. Additional morphological characters such as ray number and length, sepal length, stylopodium length, style length and fruit dimensions (Bani et al., 2016a, b) and leaf and bract characters (e.g. Bani & Koch, 2015) have been used successfully in morphometric analyses to differentiate the various taxa introduced here at species and subspecies rank, but these characters do not differentiate taxa at higher taxonomic levels. Phylogeographical implications The crown group age of Careae was assumed to be c. 18 Mya. This estimate is largely congruent with previous studies (e.g. Banasiak et al., 2013; Calviño, Teruel & Downie, 2016). With a number of other tribes such as Selineae or Pimpinelleae, Careae belong to a clade confined to the IT floristic region as an ancestral area since the Oligocene 31 Mya (Banasiak et al., 2013). This long-term persistence in the IT region might explain the extraordinarily high number of genera and endemic taxa of Apioideae found there. The onset of diversification in Grammosciadium and related taxa was placed in the Miocene c. 7.5 Mya and gave rise to the various species of Carum, Chamaesciadium, Fuernrhoria and Grammosciadium until the end of the Oligocene. In Apioideae as a whole, the increase of evolutionary lineages over 40 My is relatively constant (log-scale) (Banasiak et al., 2013), and consequently diversification events in Careae cannot be attributed to any hypothetical shift in speciation rates in Apioideae. With our data the majority of species or groups of species are of pre-Quarternary origin and are restricted to defined regions in Anatolia (Fig. 5). Most of them are found east of the Anatolian Diagonal, except G. confertum west of the Diagonal. This might indicate that the Diagonal acted as barrier to gene flow and migration during the Pliocene, but demonstrating high permeability throughout the Pleistocene. The three widespread Anatolian species (Fig. 5) evolved (stem group age) either earlier (G. platycarpum, G. pterocarpum; late Miocene) or later during the Pleistocene (G. daucoides). This might suggest that the Pleistocene, with its cycling environmental conditions, drastically affected humidity and, thereby, the distribution and extent of open steppe vegetation, providing a period in time for new diversification. In this regard, G. pterocarpum might be one such example. This taxon further differentiated during the Pleistocene and shows a distribution pattern (Figs 3, 5) with defined maternally inherited plastid types at the north-eastern corner of its distribution range. Similarly, these distribution patterns at the margin of the total distribution range of putative ancestral species are found with G. pterocarpum subsp. bilgilii, G. haussknechtii and G. pterocarpum subsp. sivasicum (Fig. 5b) and G. scabridum (Fig. 5f). Similar spatio-temporal diversification patterns in Anatolia have been also observed in other plant species such as Arabis alpina L. and its relatives (Koch et al., 2006; Ansell et al., 2011; Karl et al., 2012) and Aubrieta Adans. (Koch et al., 2017), indicating that the Pleistocene was an important epoch for speciation processes in Anatolia. There are only a few examples studying vicariance patterns characterized by the Anatolian Diagonal and unravelling a temporal evolutionary scenario. Bilgin (2011) provided a review of studies of intraspecific genetic diversity in Anatolia as a whole with particular reference to divergence times. However, among the discussed studies there is only one from plants (Bittkau et al., 2005) and this study focuses on Aegean Island diversity also established during the Pleistocene (Nigella arvensis L. alliance). All other dated results reviewed by Bilgin (2011) are from animals and exemplify a diverse and complex amalgamation of spatiotemporal patterns, and it was concluded that the multiple chains of mountains including the Anatolian Diagonal do not seem to have been impermeable boundaries for post- (or peri-)glacial expansion (e.g. Ciplak, Demirsoy & Bozcuk, 1993; Demirsoy, 1996). This is fully supported by our results from spatial plastid DNA type variation. It is not possible to structure the identified haplotypes as indicated with the network analysis (Fig. 3) in a spatial context and to define geographically characterized groups (details not shown), except the example of G. pterocarpum. Plastid DNA-based gene diversity calculations also support this result, and in most of our cases high haplotype diversity (h) and moderate to high nucleotide diversity (π) are found in those species that originated prior to the Pleistocene (Table 3). High h and high π can be considered as an expected signature for long-term stable populations with large long-term effective population sizes. Alternatively, such a pattern could also be the result from admixture of individuals from historically sundered populations, but for this scenario our data do not provide further evidence. There are only a few studies in Apiaceae providing plastid DNA-based genetic diversity statistics on the species level. In Pastinaca sativa L. haplotype diversity (h) and nucleotide diversity (π) estimated from 114 individuals were 0.55–0.94 and 0.0067–0.0139, respectively (Jogesh et al., 2015). In Mulinum spinosum Pers. high h and π were estimated (0.757 and 0.00245, respectively), on the basis of 71 individuals (Sede et al., 2012). In both cases intraspecies differentiation processes were placed also into the Pleistocene. Figure 5. Open in new tabDownload slide Cartoons of the distribution of the various Grammosciadium taxa and Fuernrohria (b–f). The schematic phylogenetic tree (a) has been redrawn from Figure 2. Colour codes correspond to Figures 1–3. Figure 5. Open in new tabDownload slide Cartoons of the distribution of the various Grammosciadium taxa and Fuernrohria (b–f). The schematic phylogenetic tree (a) has been redrawn from Figure 2. Colour codes correspond to Figures 1–3. Table 3. Haplotype (h) and nucleotide (π) diversity calculations for the phylogenetically defined taxa of Grammosciadium; analyses were performed with DnaSP version 5.10.1 (Librado & Rozas, 2009), and standard deviations are given in parentheses Taxon . Number of sequences . Number of haplotypes . h . π . G. pterocarpum 17 10 0.868 (0.070) 0.0034 (0.0005) G. platycarpum 10 6 0.778 (0.137) 0.0021 (0.0006) G. daucoides 26 12 0.911 (0.035) 0.0039 (0.0003) G. macrodon 5 4 0.900 (0.161) 0.0024 (0.0011) G. macrodon subsp. nezaketiae 5 4 0.900 (0.161) 0.0025 (0.0006) G. cornutum 7 4 0.810 (0.130) 0.0045 (0.0009) G. confertum 3 1 0.000 0.0000 Taxon . Number of sequences . Number of haplotypes . h . π . G. pterocarpum 17 10 0.868 (0.070) 0.0034 (0.0005) G. platycarpum 10 6 0.778 (0.137) 0.0021 (0.0006) G. daucoides 26 12 0.911 (0.035) 0.0039 (0.0003) G. macrodon 5 4 0.900 (0.161) 0.0024 (0.0011) G. macrodon subsp. nezaketiae 5 4 0.900 (0.161) 0.0025 (0.0006) G. cornutum 7 4 0.810 (0.130) 0.0045 (0.0009) G. confertum 3 1 0.000 0.0000 Open in new tab Table 3. Haplotype (h) and nucleotide (π) diversity calculations for the phylogenetically defined taxa of Grammosciadium; analyses were performed with DnaSP version 5.10.1 (Librado & Rozas, 2009), and standard deviations are given in parentheses Taxon . Number of sequences . Number of haplotypes . h . π . G. pterocarpum 17 10 0.868 (0.070) 0.0034 (0.0005) G. platycarpum 10 6 0.778 (0.137) 0.0021 (0.0006) G. daucoides 26 12 0.911 (0.035) 0.0039 (0.0003) G. macrodon 5 4 0.900 (0.161) 0.0024 (0.0011) G. macrodon subsp. nezaketiae 5 4 0.900 (0.161) 0.0025 (0.0006) G. cornutum 7 4 0.810 (0.130) 0.0045 (0.0009) G. confertum 3 1 0.000 0.0000 Taxon . Number of sequences . Number of haplotypes . h . π . G. pterocarpum 17 10 0.868 (0.070) 0.0034 (0.0005) G. platycarpum 10 6 0.778 (0.137) 0.0021 (0.0006) G. daucoides 26 12 0.911 (0.035) 0.0039 (0.0003) G. macrodon 5 4 0.900 (0.161) 0.0024 (0.0011) G. macrodon subsp. nezaketiae 5 4 0.900 (0.161) 0.0025 (0.0006) G. cornutum 7 4 0.810 (0.130) 0.0045 (0.0009) G. confertum 3 1 0.000 0.0000 Open in new tab Taxonomic conclusions The first attempt to group the members of Grammosciadium s.l. was made by Boissier (1872). Schischkin (1923) properly classified the winged members in a separate genus by resurrecting Caropodium, which was described in 1886 (Stapf, 1886). Finally Tamamschian & Vinogradova (1969b, 1970) reduced Caropodium to the rank of subgenus. However, our molecular phylogenetic hypothesis strongly supports the concept of Schischkin (1923). We designate the clades of G. pterocarpum/G. haussknechtii and G. platycarpum as a genus, Caropodium. Vinogradova (1995) discussed the possibility of recognizing G. confertum as a separate genus based on peculiar fruit morphology and anatomy, but favoured placing it into a monotypic section Heterocarpum of Grammosciadium subgenus Grammosciadium sensu Tamamschian & Vinogradova (1969b, 1970). Here we raise this section to the genus level. The new genus Vinogradovia has unique characters including strongly conferted, thickened and erect rays, sessile central umbels, strongly conferted fruits and heteromorphic mericarps, as discussed by Vinogradova (1995). Another character (reticulate-striate mericarp ornamentation), which supports the distinctness of the genus, was presented in a recent study (Bani et al. 2016b). Summarizing, we divide the polyphyletic genus Grammosciadium s.l. into three genera: Grammosciadium s.s., the restored Caropodium and a newly described Vinogradovia. The morpho-anatomical characters which support the molecular phylogenetic hypothesis are presented here. In addition to fruit wing characters, the genera of Grammosciadium s.s. and Caropodium can be distinguished traditionally by calyx length (Vinogradova 1995; Bani et al., 2015, 2016b). However, detailed morphological examinations have shown that shorter calyces were also observed in samples of G. daucoides (0.12–1.70 mm). Thus, this character is no longer diagnostic. Bani et al. (2016b) indicated that fruit surface characters are useful for discrimination of these genera (rugose-favulariate and tuberculate-striate in Grammosciadium s.s.; reticulate-tuberculate and ribbed-striate in Caropodium). The presence of sclereids (stone cells) in transverse sections of the roots was found to be a good character defining the genus Caropodium (Ulusoy et al., 2016). We observed irregularly arranged sclerenchyma tissue (not stone cells) in transverse sections of roots in Grammosciadium s.s. and Vinogradovia. The number of vascular bundles in transverse sections of mericarps of Grammosciadium s.s. is five (except for G. cornutum with nine vascular bundles) as in all members of Caropodium. Grammosciadium cornutum has an intermediate position between these genera; Vinogradovia and Fuernrohria have continuous vascular bundles in transverse sections of mericarps. Grammosciadium s.s. This genus comprises four species with five taxa in total. The most discriminative characters are from fruits and flowers. However, form and shape of oil-ducts in valleculae divide these five taxa into two groups. The first group contains G. daucoides and G. scabridum, which have large and elliptical vallecular oil-ducts in transverse sections of the mericarps. The second group consists of G. macrodon subspp. macrodon and nezaketiae and G. cornutum, which share the character of small and orbicular oil-ducts in vallecular regions of the mericarps. The sister-relationship of G. macrodon subspp. macrodon and nezaketiae remains unresolved, but at least BEAST analysis indicate that G. macrodon subsp. nezaketiae might be treated in future as a separate species with more significant phylogenetic information at hand. Oil-ducts in the petals of G. daucoides are characteristically short and confined to a notch. All other taxa have long and linear oil-ducts in the petals. The number of vascular bundles also clearly separates G. cornutum (nine bundles) from the other species (five bundles), which is also the case for the presence of sclerenchymatous tissue in valleculae in G. cornutum. Large and elliptical vascular bundles in transverse sections of the mericarps and also dorsally compressed mericarps are characteristic for G. macrodon subspp. macrodon and nezaketiae. All other species have nearly isodiametric and relatively small vascular bundles and laterally compressed mericarps. Fruit primary ridges are generally prominent. One exception is found with G. macrodon subsp. macrodon, which has obscure primary fruit ridges. Long and recurved calyx teeth were observed in G. cornutum and G. macrodon subsp. nezaketiae (> 1.9 mm), clearly separating these two taxa from G. daucoides and G. scabridum. (calyx teeth straight and < 1.6 mm); G. macrodon subsp. macrodon has an intermediate position in terms of calyx teeth characteristics (straight and 1.6–6.0 mm). Caropodium As a result of the phylogenetic analysis presented herein, G. platycarpum, G. haussknechtii, G. pterocarpum subspp. pterocarpum, sivasicum (accessions BS3825, 7255) and bilgilii (6965) are transferred from Grammosciadium s.l. to Caropodium, which now comprises five taxa. Morpho-anatomical affinities and important characteristics of Caropodium were previously discussed in detail (Bani et al., 2016a). Long pedicellate stipular segments are characteristic for C. platycarpum. All other taxa in this genus have nearly sessile stipular segments (Bani et al., 2016a). Spreading fruits, and thickened and straight umbel rays (in fruiting stage) separate C. platycarpum and C. haussknechtii from the C. pterocarpum complex. Width of fruit wings was previously used as one of the most discriminative characters (Boissier, 1872; Bordzilowski 1915; Tamamschian and Vinogradova, 1969a, b, 1970; Vinogradova, 1995). The C. pterocarpum complex, in particular, shows high morphological variation as indicated earlier (Bani et al., 2016a). Width of fruit wings gradually increases from west to east. Recently, an isolated population in western Anatolia has been discovered and described as a new subspecies (subsp. bilgilii) based on narrower wings and shorter fruits (Bani et al., 2016a). Although the widest winged populations were observed in areas adjacent to the type locality of previously synonymized species G. schischkinii (north-east of the distribution range of G. pterocarpum, Fig. 5), it was not possible to find any character to support the distinctness of these populations when considering all populations within the distribution area of the C. pterocarpum complex, with continuous characters always overlapping. We provide below identification keys to the genera and infrageneric taxa, followed by a description of Grammosciadium, Vinogradovia and Caropodium including infrageneric classifications. However, infrageneric classification is presented here only to avoid future confusion with earlier taxonomic concepts. Generally, we do not see any necessity for further subgeneric or sectional classification. Key to Carum, Caropodium, Fuenrrohria, Grammosciadium s.s. and Vinogradovia 1. Fruits less than three times longer than broad 2. Bracteoles present………Fuernrohria (F. setifolia) *2. Bracteoles absent…………………Carum (C. carvi) *1. Fruits more than three times longer than broad 3. Lateral ridges of mericarps unwinged 4. Central umbels sessile; fruiting rays conferted………………………………..Vinogradovia *4. Central umbels pedicellate, fruiting rays not conferted…………………………..Grammosciadium *3. Lateral ridges of mericarps winged……Caropodium Grammosciadium, key to the taxa 1. Petal oil-duct short, confined to notch…………………………………………………..…….G. daucoides *1. Petal oil-duct linear 2.  Nine vascular bundles in each mericarp………………………………………………………G. cornutum *2. Five vascular bundles in each mericarp 3. Wing-like stria present on dorsal side of primary ridges of mericarps; sepals recurved ….………………….G. macrodon subsp. nezaketiae *3. Wing-like stria absent on dorsal side of primary ridges of mericarps; sepals erect or patent 4. Vallecular vittae smaller and orbicular; vascular bundles elliptical in each mericarp ….……………………G. macrodon subsp. macrodon *4. Vallecular vittae larger and elliptical; vascular bundles isodiametric in each mericarp …….……………………………………G. scabridum Caropodium, key to the taxa 1. Fruiting rays thickened; fruits spreading 2. Stipular segments of cauline leaves long pedicellate; two funicular oil-ducts present in each mericarp……………………………..C. platycarpum *2. Stipular segments of cauline leaves nearly sessile; funicular oil-duct absent in each mericarp……………………………C. haussknechtii *1.  Fruiting rays not thickened; fruits erect 3. Two funicular oil-ducts present in each mericarp……………..C. pterocarpum subsp. sivasicum *3.  Funicular oil-duct absent in each mericarp 4. Wings of mericarp > 1 mm broad………………………………….C. pterocarpum subsp. pterocarpum *4. Wings of mericarp < 1 mm broad………………………………………….C. pterocarpum subsp. bilgilii Synopses of Grammosciadium, Carodium and Vinogradovia Grammosciadium DC., Coll. Mem. 5: 62 (1829) Type [LT, designated by Pimenov & Tikhomirov, Novosti Sist. Vyssh. Rast. 16: 163 (1979)]: G. daucoides DC. Perennial plants with stout tap roots crowned with a fibrous collar. Leaves three- to five-pinnatisect; segments linear and mucronate. Cauline leaves with leaf-like stipules. Bracts and bracteoles present. Umbels polygamous. Sepals present. Petals white with central oil-duct, outer petals somewhat radiate. Mericarps glabrous, ± terete, linear to oblong, vittate, primary ridges five. Vascular bundles five or nine in each mericarp. Grammosciadium section Grammosciadium 1. G. daucoides DC., Coll. Mem. 5: 62 (1829). ≡Prionitis daucoides (DC.) Koso-Pol. in Bull. Soc. Imp. Naturalistes Moscou 29: 140 (1916). =G. szovitsii Boiss. in Ann. Sci. Nat. ser. 3(2): 67 (1844). =G. aucheri Boiss. in Ann. Sci. Nat. ser. 3(2): 67 (1844). ≡P. aucheri (Boiss.) Koso-Pol. in Bull. Soc. Imp. Naturalistes Moscou 29: 140 (1916). =G. aucheri subsp. pauciradiatum Freyn & Sint. in Oesterr. Bot. Z. 42: 128 (1892). 2. G. scabridum Boiss. in Ann. Sci. Nat. ser. 3(2): 66 (1844). ≡Prionitis scabrida (Boiss.) Koso-Pol. in Bull. Soc. Imp. Naturalistes Moscou 29: 140 (1916). =G. longilobum Boiss. & Hausskn. in Boiss., Fl. Orient. 2: 900 (1872). ≡Prionitis longiloba (Boiss.) Koso-Pol. in Bull. Soc. Imp. Naturalistes Moscou 29: 140 (1916). Grammosciadium section Macrodon Koso-Pol. in Journ, Russ. Bot. 1915 (1–2): 12 (1915). 3. G. macrodon Boiss. in Ann. Sci. Nat. ser. 3(2): 67 (1844). 3a. G. macrodon subsp. macrodon ≡Prionitis macrodon (Boiss.) Koso-Pol. in Bull. Soc. Imp. Naturalistes Moscou 29: 137 (1916). 3b. G. macrodon subsp. nezaketiae Bani in Phytotaxa 224 (3): 271 (2015). Grammosciadium section Ceratodon Tamamsch. & V.M.Vinogr. [in Vinogradova, Monogr. Review of Grammosciadium DC.: abstract of PhD thesis. 14 (1971), nom. inval.], sect. nov. Mericarps with nine vascular bundles in transverse section developed both in primary ribs and valleculae (vs. five in primary ribs only in other Grammosciadium spp.). Type: G. cornutum (Nábělek) C.C.Towns. The name of this section first appeared in the abstract of a PhD thesis (Vinogradova, 1971) with neither a description nor diagnosis and was therefore not effectively or validly published. Similarly, it is also a nomen nudum in Vinogradova (1995), apparently the only paper where it was subsequently used. As long as one of the assumed authors repeatedly used this name in a given sense clearly demonstrating the same intention, it is validated here with original authorship. 5. G. cornutum (Nábělek) C.C.Towns. in Kew Bull. 20: 83 (1966). ≡G. macrodon var. cornutum Nábělek in Publ. Fac. Sci. Univ. Masaryk, Brno 35: 124 (1923). Vinogradovia Bani, D.A.German & M.A.Koch, nom. et stat. nov. Type: V. conferta (Hub.-Mor. & Lamond) Bani, D.A.German & M.A.Koch Based on: Grammosciadium subgenus Caropodium section Heterocarpum V.M.Vinogr. in Bot. Zhurn. 80 (1): 94 (1995). Biennial plants with fusiform roots without a fibrous collar. Leaves three- (four-)pinnatisect; segments linear and mucronate. Cauline leaves with leaf-like stipules. Rays strongly thickened and conferted. Bracts and bracteoles present. Umbels polygamous, central umbel sessile. Sepals present. Petals white with central oil-duct, outer petals somewhat radiate. Mericarps glabrous, ± terete, heteromorphic, linear to oblong, vittate, primary ridges five. Vascular bundles continuous as a ring in each mericarp. The new generic name is introduced in order to avoid creation of ‘Heterocarpum’, which apparently would be treated as a later homonym of a name of the lichen genus Heterocarpon Müll. Arg. (1885). Vinogradovia commemorates Vera Mikhaylovna Vinogradova (1937–2008) who contributed greatly to the systematics of Grammosciadium s.l. and, in particular, emphasized the distinctiveness of G. confertum by placing it in a separate section of Grammosciadium. 1. V. conferta (Hub.-Mor. & Lamond) Bani, D.A.German & M.A.Koch comb. nov. Basionym: G. confertum Hub.-Mor. & Lamond in Notes Roy. Bot. Gard. Edinburgh 31: 75 (1971). Caropodium Stapf et Wettst. in Denkschr. Kaiserl. Akad. Wiss., Wien. Math.-Naturwiss. Kl. 51: 317 (1886). Type: Caropodium meoides Stapf & Wettst. =Stenodiptera Koso-Pol. in Journ. Russ. Bot. 1915 (1–2): 12 (1915). Perennial plants with stout tap roots crowned with a fibrous collar. Leaves three- to six-pinnatisect; segments linear and mucronate. Cauline leaves with leaf-like stipules. Bracts and bracteoles present. Umbels polygamous. Sepals present. Petals white with central oil-duct, outer petals somewhat radiate. Mericarps glabrous, ± terete, linear to oblong, vittate, primary ridges five, laterals winged. Vascular bundles nine in each mericarp. Caropodium section Caropodium 1. C. platycarpum (Boiss. & Hausskn.) Schischkin in Not. Syst. Herb. Hort. Bot. Petrop. 4: 30 (1923). Basionym: G. platycarpum Boiss. & Hausskn. in Boiss., Fl. Orient. 2: 901 (1872). ≡Stenodiptera platycarpa (Boiss. & Hausskn.) Koso-Pol. in Journ. Russ. Bot. 1915 (1–2): 13 (1915). =Caropodium meodies Stapf et Wettst. in Denkschr. Kaiserl. Akad. Wiss., Wien. Math.-Naturwiss. Kl. 51: 317 (1886). Caropodium section Stenodiptera (Koso-Pol.) Bani & M.A.Koch comb. nov. Type [LT designated here]: Caropodium pterocarpum (Boiss.) Schischkin Basionym: Stenodiptera Koso-Pol. in Journ. Russ. Bot. 1915 (1–2): 12 (1915). 2. C. haussknechtii (Boiss.) Schischkin in Not. Syst. Herb. Hort. Bot. Petrop. 4: 30 (1923). Basionym: Grammosciadium haussknechtii Boiss., Fl. Orient. 2: 901 (1872). ≡Stenodiptera haussknechtii (Boiss.) Koso-Pol. in Journ. Russ. Bot. 1915 (1–2): 13 (1915). 3. C. pterocarpum (Boiss.) Schischkin in Not. Syst. Herb. Hort. Bot. Petrop. 4: 30 (1923). 3a. C. pterocarpum subsp. pterocarpum Basionym: Grammosciadium pterocarpum Boiss. in Ann. Sci. Nat. ser. 3(2): 68 (1844). ≡Stenodiptera pterocarpa (Boiss.) Koso-Pol. in Journ. Russ. Bot. 1915 (1–2): 13 (1915). =Grammosciadium pterocarpum subsp. longipes Freyn in Bull. Herb. Boiss. sér. 2, 1: 269 (1901). =Stenodiptera armena Bordz. in Mem. Soc. Nat. Kiev 25 (1): 96 (1915). =Caropodium pterocarpum var. schischkinii V.M.Vinogr. & Tamamsch. in Notes Roy. Bot. Gard. Edinburgh 28: 203 (1968). ≡Grammosciadium schischkinii (V.M.Vinogr. & Tamamsch.) V.M.Vinogr. in Bot. Zhurn. 80 (1): 94 (1995). 3b. C. pterocarpum subsp. bilgilii (Bani) Bani & M.A.Koch comb. nov. Basionym: Grammosciadium pterocarpum subsp. bilgilii Bani in PhytoKeys 68: 80 (2016). 3c. C. pterocarpum subsp. sivasicum (Bani) Bani & M.A.Koch comb. nov. Basionym: Grammosciadium pterocarpum subsp. sivasicum Bani in PhytoKeys 68: 81 (2016). CONCLUSIONS Grammosciadium has been shown to be paraphyletic in its traditional circumscription. Morphological and anatomical characters are not able to unravel the relationships with Carum and Fuernrohria unambiguously, but in light of molecular data, morphological characters can be interpreted accordingly to present a new generic concept. Phylogenetic analyses demonstrate that most species started to evolve during the late Miocene and Pliocene in Anatolia, highlighting the importance of this region as a cradle of biodiversity of the western IT region. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Table S1. Detailed accession list and GenBank accession numbers. Table S2. Nexus file of ITS sequence data. Table S3. Nexus file of plastid sequence data. Table S4. Summary of diagnostic morphological and anatomical characters. Figure S1. Phylogenetic tree based on nuclear ITS data (maximum-likelihood inference of tribal alignment). Figure S2. Phylogenetic tree based on nuclear ITS data (Bayesian inference of tribal alignment). Figure S3. Results from BEAST analysis based on nuclear ITS data (*.tre input file). ACKNOWLEDGEMENTS We thank Peter Sack, Lisa Kretz and especially Xenia Augsten for their help in the lab. Bilal Şahin, Hüseyin Eroğlu, Ebru Doğan Güner and Bilgehan Bilgili helped during the field studies. Nezaket Adıgüzel, Özlem Mavi İdman, Muhammet Ali Karakaya, Fatma Ulusoy, Cihat Ölçücü and Emre Erez contributed to anatomical and karyological studies. We also thank the curators of Berlin herbarium for providing voucher material. Moreover, we are grateful to Stephen Downie, Krzysztof Spalik and Lukasz Banasiak for their discussion and helpful comments and generous support providing detailed and updated results from their earlier work crucial for the divergence time estimates presented herein. We thank Graham Muir for his editorial help and further general discussions. This work was supported by the TUBİTAK, under grant number 114Z094. REFERENCES Ajani Y , Ajani A , Cordes JM , Watson MF , Downie SR . 2008 . Phylogenetic analysis of nrDNA ITS sequence reveals relationships within five groups of Iranian Apiaceae subfamily Apioideae . Taxon 57 : 383 – 401 . OpenURL Placeholder Text WorldCat Ansell SW , Stenøien HK , Grundmann M , Russell SJ , Koch MA , Schneider H , Vogel JC . 2011 . The importance of Anatolian mountains as the cradle of global diversity in Arabis alpina, a key arctic-alpine species . Annals of Botany 108 : 241 – 252 . Google Scholar Crossref Search ADS PubMed WorldCat Banasiak L , Piwczyński M , Uliński T , Downie SR , Watson MF , Shakya R , Spalik K . 2013 . Dispersal patterns in space and time: a case study of Apiaceae subfamily Apioideae . Journal of Biogeography 40 : 1324 – 1335 . Google Scholar Crossref Search ADS WorldCat Bani B , Adigüzel N . 2010 . Seed surface analysis of some threatened endemic plants from Tahtali Mountains (Adana-Kayseri/Turkey) . Biological Diversity and Conservation 3 : 87 – 92 . OpenURL Placeholder Text WorldCat Bani B , Mavi Ö , Adigüzel N . 2011 . Morphological and anatomical notes on a local endemic: Grammosciadium confertum Hub.-Mor. & Lamond (Umbelliferae) . Biological Diversity and Conservation 4 : 1 – 6 . Google Scholar Crossref Search ADS WorldCat Bani B , Koch M . 2015 . A new subspecies of Grammosciadium macrodon Boiss. (Apiaceae) from Turkey . Phytotaxa 224 : 267 – 275 . Google Scholar Crossref Search ADS WorldCat Bani B , Ulusoy F , Karakaya MA , Koch MA . 2016a . Taxonomical implications from morphological and anatomical studies in section Stenodiptera from genus Grammosciadium (Apiaceae) . PhytoKeys 68 : 73 – 89 . Google Scholar Crossref Search ADS WorldCat Bani B , Karakaya MA , Ceter T . 2016b . Fruit micromorphological characters of the genus Grammosciadium DC. (Apiaceae) in Turkey . Phytotaxa 246 : 184 – 191 . Google Scholar Crossref Search ADS WorldCat Behçet L , Kaval I , Rüstemoğlu M . 2012 . Three new records for Turkey: Allium giganteum (Liliaceae), Grammosciadium scabridum, and Ferula angulata subsp. carduchorum (Apiaceae) . Turkish Journal of Botany 36 : 637 – 643 . OpenURL Placeholder Text WorldCat Bilgin R . 2011 . Back to the suture: the distribution of intraspecific genetic diversity in and around Anatolia . International Journal of Molecular Sciences 12 : 4080 – 4103 . Google Scholar Crossref Search ADS PubMed WorldCat Bittkau C , Comes HP . 2005 . Evolutionary processes in a continental island system: molecular pylogeography of the Aegean Nigella arvensis alliance (Ranunculaceae) inferred from chloroplast DNA . Molecular Ecology 14 : 4065 – 4083 . Google Scholar Crossref Search ADS PubMed WorldCat Boissier PE . 1872 . Flora Orientalis, Vol. 2 . Geneva : H. Georg . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Boissier PE . 1844 . Plantae Aucherianae adjunctis nonnullis e regionibus Mediterraneis et Orientalibus aliis cum novarum specierum descriptione; Umbelliferae . Annales des Sciences naturelles, Botanique 3 : 66 – 68 . OpenURL Placeholder Text WorldCat Bordzilowski EI . 1915 . Contribution to the flora of Caucasus . Mémoires de la Société des Naturalistes de Kiew 25 : 96 (in Russian). OpenURL Placeholder Text WorldCat Candolle AP . 1829 . Mémoire sur la famille des Ombellifères . In: Candolle AP , ed. Collection de mémoires pour servir à l’histoire du régne vegetal 5 . Paris , 1 – 84 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Calviño CI , Teruel FE , Downie SR . 2016 . The role of the Southern Hemisphere in the evolutionary history of Apiaceae, a mostly north temperate plant family . Journal of Biogeography 43 : 398 – 409 . Google Scholar Crossref Search ADS WorldCat Ciplak B , Demirsoy A , Bozcuk AN . 1993 . Distribution of Orthoptera in relation to the Anatolian Diagonal in Turkey . Articulata 8 : 1 – 20 . OpenURL Placeholder Text WorldCat Daushkevich JV , Alexeeva TV , Pimenov MG . 1991 . IOPB chromosome data 3 . International Organization of Plant Biosystematists Newsletter 17 : 8 – 9 . OpenURL Placeholder Text WorldCat Daushkevich JV , Alexeeva TV , Pimenov MG . 1995 . IOPB chromosome data 10 . International Organization of Plant Biosystematists Newsletter 25 : 7 – 8 . OpenURL Placeholder Text WorldCat Davlianidze MT . 1985 . Chromosome numbers in the representatives of the flora from Georgia . Botanicheskii Zhurnal (Moscow & Leningrad) 70 : 698 – 700 (in Russian). OpenURL Placeholder Text WorldCat Davis PH . 1971 . Distribution patterns in Anatolia with particular reference to endemism . In: Davis PH , Harper PC , Hedge IC , eds. Plant life of South West Asia . Edinburgh : Botanical Society of Edinburgh , 15 – 27 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Demirsoy A . 1996 . Genel ve Türkiye Zoocograjyasi . Ankara : Meteksan AS . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Dobeš CH , Mitchell-Olds T , Koch MA . 2004 . Extensive chloroplast haplotype variation indicates Pleistocene hybridization and radiation of North American Arabis drummondii, A. × divaricarpa, and A. holboellii (Brassicaceae) . Molecular Ecology 13 : 349 – 370 . Google Scholar Crossref Search ADS PubMed WorldCat Downie SR , Spalik K , Katz-Downie DS , Reduron JP . 2010 . Major clades within Apiceae subfamily Apioideae as inferred by phylogenetic analysis of nrDNA ITS sequences . Plant Diversity and Evolution 128 : 111 – 136 . Google Scholar Crossref Search ADS WorldCat Ekim T , Güner A . 1986 . The Anatolian Diagonal: fact or fiction? Proceedings of the Royal Society of Edingburgh. Section B. Biological Sciences 89 : 69 – 77 . Google Scholar Crossref Search ADS WorldCat Freyn J . 1901 . Über neue und bemerkenswerte orientalische Pflanzenarten, 5 . Bulletin de l’Herbier Boissier, Série 2 , 1 : 245 – 289 . OpenURL Placeholder Text WorldCat Gruas-Cavagnetto C , Cerceau-Larrival MT . 1984 . Apports des pollens fossiles d’Ombellifères à la connaissance paléoécologique et paléoclimatique de l’Éocène français . Review of Palaeobotany and Palynology 40 : 317 – 345 . Google Scholar Crossref Search ADS WorldCat Huson DH , Bryant D . 2006 . Application of phylogenetic networks in evolutionary studies . Molecular Biology and Evolution 23 : 254 – 267 . Google Scholar Crossref Search ADS PubMed WorldCat Jogesh T , Peery R , Downie SR , Berenbaum MR . 2015 . Patterns of diversity in the globally invasive species wild parsnip (Pastinaca sativa) . Invasive Plant Science and Management 8 : 415 – 429 . Google Scholar Crossref Search ADS WorldCat Karl R , Kiefer C , Ansell SW , Koch MA . 2012 . Systematics and evolution of Arctic-Alpine Arabis alpina (Brassicaceae) and its closest relatives in the eastern Mediterranean . American Journal of Botany 99 : 778 – 794 . Google Scholar Crossref Search ADS PubMed WorldCat Katoh K , Standley DM . 2013 . MAFFT multiple sequence alignment software version 7: improvements in performance and usability . Molecular Biology and Evolution 30 : 772 – 780 . Google Scholar Crossref Search ADS PubMed WorldCat Katz-Downie DS , Valiejo-Roman CM , Terentieva EI , Troitsky AV , Pimenov MG , Lee B , Downie SR . 1999 . Towards a molecular phylogeny of Apiaceae subfamily Apioideae: additional information from nuclear ribosomal DNA ITS sequences . Plant Systematics and Evolution 216 : 167 – 195 . Google Scholar Crossref Search ADS WorldCat Kiehn M , Vitek E , Dobes C . 2000 . New counts in different genera . In: Dobes C , Vitek E , eds. Documented chromosome number checklist of Austrian vascular plants . Vienna : Verlag des Naturhistorischen Museums Wien , 41 – 82 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Koch MA , Kiefer C , Ehrich D , Vogel J , Brochmann C , Mummenhoff K . 2006 . Three times out of Asia Minor: the phylogeography of Arabis alpina L. (Brassicaceae) . Molecular Ecology 15 : 825 – 839 . Google Scholar Crossref Search ADS PubMed WorldCat Koch MA , Karl R , German DA . 2017 . Underexplored biodiversity of Eastern Mediterranean biota: systematics and evolutionary history of the genus Aubrieta (Brassicaceae) . Annals of Botany 119 : 39 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat Koso-Poljansky B . 1914 . O filogenii rodov Umbelliferae Kavkaza. I. Apioideae [On phylogeny of the Umbelliferae genera of Caucasus. I. Apioideae] . Trudy Tiflisskogo Botanicheskogo Sada 16 : 172 – 229 . OpenURL Placeholder Text WorldCat Koso-Poljansky BM . 1915 . A revision of the oriental genus Grammosciadium DC . Journal Russe de Botanique 1915 : 1 – 22 . OpenURL Placeholder Text WorldCat Koso-Poljansky B . 1916 . Sciadophytorum sistematis lineamenta . Bulletin de la Société Impériale des Naturalistes de Moscou, Nouvelle Série 29 : 93 – 221 . OpenURL Placeholder Text WorldCat Librado P , Rozas J . 2009 . DnaSP v5: a software for comprehensive analysis of DNA polymorphism data . Bioinformatics 25 : 1451 – 1452 . Google Scholar Crossref Search ADS PubMed WorldCat Liu J , Shi L , Han J , Li G , Lu H , Hou J , Zhou X , Meng F , Downie SR . 2014 . Identification of species in the angiosperm family Apiaceae . Molecular Ecology Resources 14 : 1231 – 1238 . Google Scholar Crossref Search ADS PubMed WorldCat Löve Á . 1982 . IOPB chromosome number reports LXXVI . Taxon 31 : 574 – 598 . OpenURL Placeholder Text WorldCat Manafzadeh S , Staedler Y , Conti E . 2017 . Visions of the past and dreams of the future in the Orient: the Irano-Turanian region from classical botany to evolutionary studies . Biological Reviews 92 : 1365 – 1388 . Google Scholar Crossref Search ADS PubMed WorldCat Mayrose I , Barker MS , Otto SP . 2010 . Probabilistic models of chromosome number evolution and the inference of polyploidy . Systematic Biology 59 : 132 – 144 . Google Scholar Crossref Search ADS PubMed WorldCat Nazarova E , Ghukasyan A . 2004 . In: Gabrielian E , ed. Chromosome numbers of flowering plants of Armenian Flora . In: Gabrielian E, ed. Yerevan : Institute of Botany, National Academy of Sciences of Armenia, 171 pp . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Nylander JAA . 2004 . MrModeltest v2 . Program distributed by the author. Uppsala : Evolutionary Biology Centre, Uppsala University . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Papini A , Banci F , Nardi E . 2007 . Molecular evidence of polyphyletism in the plant genus Carum L. (Apiaceae) . Genetics and Molecular Biology 30 : 475 – 482 . Google Scholar Crossref Search ADS WorldCat Pimenov MG , Dauschkevich JV , Vasilieva MG , Kljuykov EV . 1996 . Mediterranean chromosome number reports 6 (716–748) . Flora Mediterranea 6 : 288 – 307 . OpenURL Placeholder Text WorldCat Pimenov MG , Alexeeva TV , Kljuykov EV . 1998 . Mediterranean chromosome number reports 8 (905–935) . Flora Mediterranea 8 : 221 – 245 . OpenURL Placeholder Text WorldCat Pimenov MG , Vasilieva MG . 1983 . IOPB chromosome number reports LXXXI . Taxon 32 : 663 – 664 . OpenURL Placeholder Text WorldCat Rechinger KH . 1987 . Chamaesciadium, Carum, Aegopodium . In: Rechinger KH , ed. Flora Iranica 162 . Graz : Akademische Druck- und Verlagsanstalt , 262 – 268 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Retina TA , Pimenov MG . 1977 . Chromosome numbers in species of the family Umbelliferae of middle Asia . Nauchnye Doklady Vysshei Shkoly (Biologicheskie Nauki) 7 : 90 – 95 . OpenURL Placeholder Text WorldCat Ronquist F , Teslenko M , van der Mark P , Ayres DL , Darling A , Höhna S , Larget B , Liu L , Suchard MA , Huelsenbeck JP . 2012 . MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space . Systematic Biology 61 : 539 – 542 . Google Scholar Crossref Search ADS PubMed WorldCat Schischkin BA . 1923 . On the genus Stenodiptera Kos.-Pol . Botanicheskie Materialy Gerbariya Glavnogo Botanicheskogo Sada RSFSR [Notulae Systematicae ex Herbario Horti Botanici Petropolitani] 4 : 29 – 30 . OpenURL Placeholder Text WorldCat Sede SM , Nicola MV , Pozner R , Johnson LA . 2012 . Phylogeography and palaeodistribution modelling in Patagonian steppe: the case of Mulinum spinosum (Apiaceae) . Journal of Biogeography 39 : 1041 – 1057 . Google Scholar Crossref Search ADS WorldCat Shimodaira H , Hasegawa M . 2001 . CONSEL: for assessing the confidence of phylogenetic tree selection . Bioinformatics 17 : 1246 – 1247 . Google Scholar Crossref Search ADS PubMed WorldCat Shner JV . 2004 . Mediterranean chromosome number reports 14 (1391–1409) . Flora Mediterranea 14 : 435 – 447 . OpenURL Placeholder Text WorldCat Shner JV , Pimenov MG , Kljuykov EV , Alexeeva TV , Ghahremani-Nejad F , Mozaffarian V . 2004 . Chromosome numbers in the Iranian Umbelliferae . Chromosome Science 8 : 1 – 9 . OpenURL Placeholder Text WorldCat Silvestro D , Michalak I . 2012 . raxmlGUI: a graphical front-end for RAxML . Organisms Diversity and Evolution 12 : 335 – 337 . Google Scholar Crossref Search ADS WorldCat Spalik K , Downie SR . 2007 . Intercontinental disjunctions in Cryptotaenia (Apiaceae, Oenantheae): an appraisal using molecular data . Journal of Biogeography 34 : 2039 – 2054 Google Scholar Crossref Search ADS WorldCat Spalik K , Wojewodzka A , Downie SR . 2001 . The evolution of fruit in Scandiceae tribe Scandicinae (Apiaceae) . Canadian Journal of Botany 79 : 1358 – 1374 . Google Scholar Crossref Search ADS WorldCat Spalik K , Banasiak L , Feist MAE , Downie SR . 2014 . Recurrent short-distance dispersal explains wide distributions of hydrophytic umbellifers (Apiaceae tribe Oenantheae) . Journal of Biogeography 41 : 1559 – 1571 . Google Scholar Crossref Search ADS WorldCat Stamatakis A . 2014 . RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies . Bioinformatics 30 : 1312 – 1313 . Google Scholar Crossref Search ADS PubMed WorldCat Stapf O . 1886 . Die botanischen Ergebnisse der Polak’schen expedition nach Persien in Jahre 1882. Plantae collectae a Dre. J. E. Polak et Th. Pichler. II Theil . Denkschriften der Kaiserlichen Akademie der Wissenschaften, Wien. Mathematisch-naturwissenschaftliche Klasse 51 : 271 – 346 . OpenURL Placeholder Text WorldCat Taberlet P , Gielly L , Pautou G , Bouvet J . 1991 . Universal primers for amplification of three non-coding regions of chloroplast DNA . Plant Molecular Biology 17 : 1105 – 1109 . Google Scholar Crossref Search ADS PubMed WorldCat Takhtajan A . 1986 . The floristic regions of the world . Berkeley : University of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tamamschian SG , Vinogradova VM . 1969a . A contribution to the taxonomy of the genus Grammosciadium DC. Umbelliferae . Botanicheskii Zhurnal (Leningrad) 54 : 1197 – 1212 . OpenURL Placeholder Text WorldCat Tamamschian SG , Vinogradova VM . 1969b . Grammosciadium and Caropodium (Umbelliferae-Apioideae) . Taxon 18 : 546 – 548 . Google Scholar Crossref Search ADS WorldCat Tamamschian SG , Vinogradova VM . 1970 . Grammosciadium and Caropodium . Taxon 19 : 652 . Google Scholar Crossref Search ADS WorldCat Terentieva EI , Valiejo-Roman CM , Pimenov MG , Kljuykov EV . 2008 . Phylogenetic analysis of nuclear rDNA ITS and morphological data reveal relationships among some endemic Umbelliferae taxa of Middle Asia, Kazakhstan and Siberia . In: Anonymous , ed. Apiales 2008. Programme and proceedings of the 6th International Symposium on Apiales . Moscow : KMK Scientific Press , 134 – 137 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Terentieva EI , Valiejo-Roman CM , Samigullin TH , Pimenov MG , Tilney PM . 2015 . Molecular phylogenetic and morphological analyses of the traditional tribe Coriandreae (Umbelliferae-Apioideae) . Phytotaxa 195 : 251 – 271 . Google Scholar Crossref Search ADS WorldCat Ulusoy F , İdman DÖM , Karakaya MA , Bani B . 2016 . A new diagnostic character in the roots of the genus Grammosciadium DC. (Apiaceae) . Phytotaxa 292 : 150 – 160 . Google Scholar Crossref Search ADS WorldCat Valiejo-Roman CM , Terentieva EI , Samigullin TH , Pimenov MG , Ghahremani-Nejad F , Mozaffarian V . 2006 . Molecular data (nrITS-sequencing) reveal relationships among Iranian endemic taxa of the Umbelliferae . Feddes Repertorium 117 : 367 – 388 . Google Scholar Crossref Search ADS WorldCat Vasilieva MG , Alexeeva GV , Pimenov MG . 1994 . Geographical variation of chromosome numbers in thin-rhizomatous species of Aegopodium (Umbelliferae) . Botanicheskii Zhurnal 79 : 27 – 31 (in Russian). OpenURL Placeholder Text WorldCat Vinogradova VM . 1970 . Grammosciadium DC. (Umbelliferae) an endemic Antasitan genus . Botanicheskii Zhurnal 55 : 54 – 60 (in Russian). OpenURL Placeholder Text WorldCat Vinogradova VM . 1971 . Monographic review of the genus Grammosciadium DC. (Apiaceae) . Abstract of PhD thesis, V. L. Komarov Botanical Institute , Leningrad (in Russian). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Vinogradova VM . 1995 . The new data on the genus Grammosciadium and the systematic position of Fuernrohria setifolia (Apiaceae) . Botanicheskii Zhurnal SSSR 80 : 91 – 99 (in Russian). OpenURL Placeholder Text WorldCat White TJ , Burns T , Lee S , Taylor J . 1990 . Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics . In: Innis MA , Gelfand DH , Sninsky JJ , White TJ . eds. PCR protocols: a guide to methods and applications . San Diego : Academic Press , 315 – 322 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Zakharova EA , Degtjareva GV , Pimenov MG . 2012 . Redefined generic limits of Carum (Umbelliferae, Apioideae) and new systematic placement of some of its taxa . Willdenowia 42 : 149 – 168 . Google Scholar Crossref Search ADS WorldCat Zohary M . 1971 . The phytogeographical foundations of the Middle East . In: Davis PH , Harper PC , Hedge IC , eds. Plant life of South West Asia . Edinburgh : Botanical Society of Edinburgh , 43 – 52 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC © 2017 The Linnean Society of London, Botanical Journal of the Linnean Society
TI - Phylogenetics, phylogeography and vicariance of polyphyletic Grammosciadium (Apiaceae: Careae) in Anatolia
JF - Botanical Journal of the Linnean Society
DO - 10.1093/botlinnean/box051
DA - 2017-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/phylogenetics-phylogeography-and-vicariance-of-polyphyletic-ELZ0rkkIHe
SP - 168
VL - 185
IS - 2
DP - DeepDyve
ER -