CSIRO PUBLISHING www.publish.csiro.au/journals/asb Australian Systematic Botany, 21, 301–311 On the conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae): a combined nuclear and plastid molecular approach Núria Garcia-Jacas A,C, Mercè Galbany-Casals A, Kostyantyn Romashchenko B and Alfonso Susanna A A Botanic Institute of Barcelona (CSIC-ICUB), Passeig del Migdia s.n., E-08038 Barcelona, Spain. Institute of Botany M. G. Kholodny, Tereshchenkovska 2, 01601 Kiev, Ukraine. C Corresponding author. Email: ngarciajacas@ibb.csic.es B Abstract. The limits of the genera that compose the Onopordum group of the Cardueae–Carduinae are difficult to establish. There are two main life forms; one is exemplified in the genus Onopordum, which includes only biennial colonisers in the Mediterranean region and temperate Eurasia; the second life form is exemplified in the group of perennial herbs of the genera Alfredia, Ancathia, Lamyropappus, Olgaea, Synurus, Syreitschikovia and Xanthopappus, all of them growing in the mountains of central Asia. We explored relationships among the genera of the complex by using Bayesian and parsimony analyses of a combined dataset of nuclear and plastid DNA sequences. Our results confirmed that the group is natural and the two life forms correspond to well defined entities. Generic limits within the eight central Asian genera are, however, very difficult to establish. Our results suggested that the present genus circumscription is artificial, especially for the largest genus, Olgaea, which appears paraphyletic. Some solutions are suggested. The most preferable might be lumping all small genera together in a broadly redefined genus Alfredia, and assigning sectional rank to the natural groups that result from correlating morphology with our molecular results. However, none of the possible solutions is free of problems because morphological characters and molecular phylogeny are not fully congruent. Some considerations on the origin and peculiar adaptations for becoming a successful coloniser shown by Onopordum are also offered, finding parallels to these adaptations in other examples of biennial colonisers within subtribe Carduinae. Introduction The Onopordum group (Cardueae–Carduinae) is composed of two well defined groups. The first one is constituted by the following eight small genera of perennial herbs with a narrow central and eastern Asian distribution: Alfredia Cass. (four species), Ancathia DC. (one species), Lamyropappus Knorring & Tamamsch. (one species), Olgaea Iljin (16 species), Syreitschikovia Pavlov (two species), Synurus Iljin (four species), Takeikadzuchia Kitag. & Kitam. (one species) and Xanthopappus C.Winkl. (one species). The second group is formed by a single large genus of widespread biennials, Onopordum L. (60 species), native to the Irano-Turanian and Mediterranean regions and introduced as noxious weeds in Australia, California and South America (Susanna and GarciaJacas 2007). The existence of this group and its limits are a recent finding and, indeed, the narrow relationships of ubiquitous Onopordum and the small Asian genera were only hinted at until recent molecular analyses were carried out. In these, Häffner and Hellwig (1999) reported a connection between Onopordum and some Asian genera; and Garcia-Jacas et al. (2002) and Susanna et al. (2006) confirmed the relationships of most of the genera of the group. Morphological relationships were
CSIRO 22 October 2008 explored by Häffner (2000) and Smirnov (2001) who pointed out morphological relationships between Alfredia and Olgaea. The systematics of the group has always been confusing and complicated, no doubt because collections of many of them are extremely scarce. For example, Syreitschikovia spinulosa (Franch.) Pavlov has been collected only three times, one of the collections by our team (Susanna et al. 2002). To illustrate the difficulties posed by the group, we shall briefly describe its taxonomic history. Some of the small middle Asian genera were originally described either as Cirsium Mill. (Ancathia was described as Cirsium igniarium Spreng. and Lamyropappus as Cirsium schakaptaricum O.Fedtsch. & B.Fedtsch.), or as Carduus L. (many species of Olgaea). Synurus was described as Onopordum by Aiton (1789) in a keen approximation to its true relationships, and Syreitschikovia was first placed in Serratula L. and then in Jurinea Cass. (Susanna et al. 2002). Even a genus described as independent, namely Alfredia, was later included in Carduus (Kazmi 1963, 1964), just as well as Olgaea. Finally, one species of Olgaea, first described in Carduus, was redescribed as a new genus Takeikadzuchia Kitag. & Kitam (Kitamura 1934). Not only was the genus classification complicated, but even their subtribal placement was problematic. Synurus and 10.1071/SB08018 1030-1887/08/040301 302 Australian Systematic Botany Syreitschikovia were included in subtribe Centaureinae (Dittrich 1977; Bremer 1994) and the rest of the genera of the group were placed in the subtribe Carduinae. Synurus was moved to the Carduinae by Häffner and Hellwig (1999). Syreitschikovia was placed in the Carduinae by Pavlov (1933), then Iljin and Semidel (1963) moved Syreitschikovia to the Centaureinae and this placement was followed by Dittrich (1977) and Bremer (1994). Susanna et al. (2002) placed it in the Carduineae on the basis of achene characters. The absence of receptacular bracts is the usual defining character for the Onopordum group. A pitted, naked receptacle is rare in the tribe. However, at least two species of Alfredia, A. acantholepis Kar. & Kir. and A. nivea Kar. & Kir., show paleate receptacle (Susanna and Garcia-Jacas 2007) and the taxonomic value of this character is diminished because it appears randomly in some rare and obviously unrelated genera across three of the five subtribes in Cardueae. Apart from the Onopordum group, naked receptacles appear also in Tugarinovia Iljin from the Carlininae, in Dolomiaea DC. from the Carduinae and in Russowia C.Winkl. and Myopordon Boiss. from the Centaureinae (Susanna and Garcia-Jacas 2007). In fact, Myopordon owes its name to its epaleate receptacle, which led Boissier (1846) to relate it to Onopordum. In a humoresque tone, Boissier chose the humble name of mice (myos in Greek) for the new genus because of its very small size, in contrast to the more imposing name of donkey (onos) accorded by Linnaeus (1753) to the tall Onopordum. Other useful characters for defining the Onopordum group are leaves, very often entire or subentire, usually spiny-dentate with parallel margins, green above and white beneath, and often arcuate (Ancathia, Syreitchikovia, part of Olgaea). The bracts of the capitulum are also very characteristic. These are always narrow, ending in a spine often arcuate in the outermost and middle bracts; the innermost and middle ones are usually dark red, and the outer ones are green. Achenes are also peculiar and very different from those of the Carduus–Cirsium model (Susanna and Garcia-Jacas 2007): an apical rim is often present, the apical nectary present in Carduus–Cirsium is absent and the pericarp is diversely pitted, wrinkled, rugulose or sulcate, rarely smooth (some species of Olgaea and Syreitschikovia). Nevertheless, all of these characters show examples of deviation and are individually misleading. On the basis of achene characters alone, Dittrich (1977) related Onopordum to Arctium L., Cousinia Cass., Tiarocarpus Rech. f. and Myopordon; and placed Alfredia, Ancathia and Olgaea in another group with Cynara L., Ptilostemon Cass. and Lamyropsis (Kharadze) Dittrich. In fact, most of the species of Onopordum are visually so different from the Asian genera that it is a complicated task to find morphological connections between Onopordum and the rest of the complex. Only in some species of Onopordum are involucral bracts, which are probably the best shared character, similar to those of the Asian genera. The anatomy of the pericarp and a set of other microanatomical features of the achenes, which are very difficult to observe and evaluate, are the unique characters that could indicate this relationship (Häffner 2000). In a phylogenetic analysis of morphological characters, the complex was placed at the base of Carduinae (Häffner 2000); however, it did not form a monophyletic group. This basal position among thistles was also supported by N. Garcia-Jacas et al. phytoserological components found in Onopordum (Fischer and Jensen 1990). However, ITS and plastid phylogenetic analyses that included some genera of the Onopordum complex suggested the monophyly of the group (Häffner and Hellwig 1999; Garcia-Jacas et al. 2002; Susanna et al. 2006), which was consistently placed near the base of the subtribe Carduinae, in a clade separate from the Carduus–Cirsium complex. The basic chromosome number is unknown in the smaller genera, with only Alfredia (Zhukova 1967; Rostovtseva 1983; Malakhova 1990), Ancathia (Krasnikov et al. 2003) and Synurus (Arano 1964; Gurzenkov 1973; Taniguchi et al. 1975) with 2n = 26 having been counted. In Alfredia, some counts of 2n = 24 have also been reported (Chuksanova et al. 1968; Rostovtseva 1979). For the rest of the genera, there are virtually no counts available in the literature. We have studied one species of Olgaea, O. pectinata Iljin, which also has 2n = 26, and Syreitschikovia spinulosa with 2n = 24 (LópezVinyallonga, unpubl. data). Both of these basic chromosome numbers are infrequent in the subtribe Carduinae and contrast with the x = 17 of Onopordum, which is much more widespread. The basic chromosome number x = 13 is found in the subtribe only in Cousinia, Lamyropsis and Saussurea DC. In contrast, other genera of the Carduinae such as Cirsium, Cynara, Jurinea, Notobasis Cass., Onopordum, Staehelina L., Silybum Vaill. or Tyrimnus Cass. have a higher basic chromosome number x = 17. Many molecular studies have been reported in the tribe Cardueae, focused on the tribe level (Susanna et al. 1995, 2006; Häffner and Hellwig 1999; Garcia-Jacas et al. 2001, 2002) or on the genus level (Susanna et al. 1999, 2003; Garcia-Jacas et al. 2000, 2006; Vilatersana et al. 2000; Kelch and Baldwin 2003; von Raab-Straube 2003; Kita et al. 2004; Garnatje et al. 2005; Martins and Hellwig 2005; Hidalgo et al. 2006; Suárez-Santiago et al. 2007; Wang et al. 2007). In some of the works cited above, the combined analysis of the nrDNA region ITS and the plastid DNA trnL50 –trnF non-coding region has been used as a good tool at the genus level. We have carried out a molecular survey of both regions of these scarcely studied genera from central Asia, in order to (1) verify the monophyly of the Onopordum complex, including all the genera of the complex not included previously in phylogenetic studies, (2) examine the relationships between the genera of the complex, (3) contribute to the genus delineation within the group, and (4) unveil some parallelisms in phylogenetical pattern and biogeographical features of Onopordum group and some other Cardueae. Materials and methods Plant material Sampling comprised representatives of all the genera of the Onopordum group accepted in the latest revision (Susanna and Garcia-Jacas 2007), including 10 of the 16 species of Olgaea, three of the four species of Alfredia, the two species of Syreitschikovia, one of the four species of Synurus, 12 of the 60 species of Onopordum, and the only species of the monotypic genera Ancathia, Xanthopappus and Lamyropappus. The number of species of Onopordum was limited to 12 because sequence Conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae) differences in the studied regions are virtually non-existent among the genus. We have also included Staehelina dubia L., S. fruticosa L. and Berardia subacaulis Vill. as outgroups according to previous sequence analyses (Garcia-Jacas et al. 2002; Susanna et al. 2006). In total, we have included in the analyses 35 ITS sequences, from which 21 are new, and 25 trnL– trnF sequences, from which 12 are new. Voucher data, source of material and GenBank sequence accession numbers are given in Table 1. DNA extraction, amplification and sequencing Total genomic DNA was extracted following the CTAB method of Doyle and Doyle (1987) as modified by Cullings (1992) from silica-gel-dried leaves collected in the field, fresh plants cultivated at the Botanic Institute of Barcelona and herbarium material. For Australian Systematic Botany 303 difficult material the commercial kits NucleoSpin Plant (Macherey-Nagel GmbH & Co. KG, Düren, Germany) and the DNeasy extraction kit (Qiagen Inc., Hilden, Germany) were used, following the manufacturer’s instructions. nrDNA ITS region strategies Double-stranded DNA of the ITS region was amplified by using the 17SE forward and the 26SE reverse primers (Sun et al. 1994) or the ITS1 forward and ITS4 reverse primers (White et al. 1990). The profile used for amplification with the 17SE and 26SE primers was as described in Galbany-Casals et al. (2004). The profile used with the primers ITS1 and ITS4 was as described in Garcia-Jacas et al. (2002). Double-stranded PCR products were purified with either QIAquick Purification Kit (Qiagen Inc., Valencia, CA, USA) or DNA Clean & Concentrator-5 (Zymo Table 1. Origin of the materials, herbaria where the vouchers are deposited and GenBank accession numbers (the new sequences are boldfaced) Species Voucher trnL–trnF accession ITS accession Alfredia acantholepis Kar. & Kir. Alfredia cernua (L.) Cass. Alfredia nivea Kar. & Kir. Ancathia igniaria (Spreng.) DC. Berardia subacaulis Vill. Lamyropappus schakaptaricus (O.Fedtsch & B.Fedtsch.) Knorr. & Tamamsch. Olgaea baldschuanica (C.Winkl.) Iljin Olgaea chodshamuminensis B.A.Sharipova Olgaea eriocephala (C.Winkl.) Iljin Kazakhstan, Susanna 2092 et al. (BC) Denmark, University of Copenhagen Bot. G. (BC) Kazakhstan, Susanna 2090 et al. (BC) Dagestan, Menitsky, T. Popova & Gorlina 8.8.1981 (LE) France, Garnatje 27 & Luque (BC) Kyrgyzstan, Poljakov 29.8.1953 (LE) FJ007860 AY772269 AY772270 FJ007861 AY772278 AY772335 AY826224 AY826225 AY826226 FJ007872 AY826234 AY826296 Tajikistan, Kamelin et al. (LE) Tajikistan, Chukavina & Chevtaeva 12.10.1979 (TAD) Tajikistan, Ovczinnicov & Zaprjagaeva 8.1923 (LE) 1 Tajikistan, Kudratov Romashchenko & Susanna 2506 (BC) 2 Mongolia, Darambin, Dariimaa, Tsooj & Vallès 4.9.2004 (BC) Mongolia, Potanin & Soldatov 1899 (LE) Kyrgyzstan, P. Polyakov 27.8.1954 (LE) Kyrgyzstan, Konnov, Kochkareva, Shibkova & Ovczinnicov 25.7.1965 (LE) Kazakhstan, Susanna 2187 et al. (BC) Tajikistan, Kudratov, Romashchenko & Susanna 2539 (BC) Kyrgyzstan, G. A. Lazkov & N. V. Kehtabaeva 19.7.2003 (LE) Spain, Garcia-Jacas & Susanna 1874 (BC) Spain, M. Font & Susanna 1815 (BC) Turkey, P. Hein 4371 (Berlin-Dahlem Bot. G.) AY772342 FJ007862 FJ007863 FJ007864 FJ007865 FJ007866 FJ007867 FJ007868 AY826304 FJ007873 FJ007874 FJ007875 FJ007875 FJ007877 FJ007878 FJ007879 AY772343 FJ007869 FJ007870 Not seq. Not seq. Not seq. AY826305 FJ007880 FJ007881 FJ007882 FJ007883 FJ007884 Turkey, P. Hein 4353a (Berlin-Dahlem Bot. G.) Not seq. FJ007885 Turkey, P. Hein 4279 (Berlin-Dahlem Bot. G.) Dijon Bot. G. (BC) Iran, Susanna 1631 et al. (BC) Not seq. Not seq. Not seq. Dijon Bot. G. (BC) Spain, Canarias, Viera y Clavijo Bot. G. (BC) Tajikistan, Kudratov, Romashchenko & Susanna 2523 (BC) Turkey, P. Hein 4439 (Berlin-Dahlem Bot. G.) Turkey, P. Hein 4403 (Berlin-Dahlem Bot. G.) France, Garnatje 25 & Luque (BC) Greece, Kriti, Garnatje 147 & Luque (BC) Japan, Kyoto, Nippon Shinyaku Institute for Botanical Research Kazakhstan, Susanna 2200 et al. (BC) Kyrgyzstan, Botschantzev 13.8. 1974 (LE) China, Qinghai [Tibet], Liu 050 (PE) AY772346 Not seq. Not seq. AY772347 Not seq. AY772365 AY772366 AY772373 FJ007886 FJ007887 AF319086, AF319140 AY826308 FJ007888 FJ007890 AY826309 FJ007889 AY826330 AY826331 AY826338 AY772374 FJ007871 AY914870 AY826339 FJ007891 AY914832 Olgaea leucophylla (Turcz.) Iljin Olgaea lomonossowii (Trautv.) Iljin Olgaea longifolia (C.Winkl.) Iljin Olgaea nidulans (Rupr.) Iljin Olgaea pectinata Iljin Olgaea petri-primi B.A.Sharipova Olgaea vvedenskyi Iljin Onopordum acanthium L. Onopordum acaulon L. Onopordum anatolicum Boiss. & Heldr. Onopordum carduchorum Bornm. & Beauverd Onopordum caricum Hub.-Mor. Onopordum illyricum L. Onopordum leptolepis DC. Onopordum nervosum Boiss. Onopordum nogalesii Svent. Onopordum seravschanicum Tamamsch. Onopordum tauricum Willd. Onopordum turcicum Danin Staehelina dubia L. Staehelina fruticosa (L.) L. Synurus palmatopinnatifidus Kitam. Syreitschikovia spinulosa (Franch.) Pavlov Syreitschikovia tenuifolia (Bong.) Pavlov Xanthopappus subacaulis C.Winkl. 304 Australian Systematic Botany Research, Orange, CA, USA), depending on the DNA quantity, and sequenced with 17SE or ITS1 as the forward primer, and 26SE or ITS4 as the reverse primer. Direct sequencing of the amplified DNA segments was performed with a ‘Big Dye Terminator v3.1 kit’ (Applied Biosystems, Foster City, CA, USA), following the protocol recommended by the manufacturer. Nucleotide sequencing was carried out at the ‘Serveis Científico-Tècnics’ of the University of Barcelona on an ABI PRISM 3700 DNA analyser (Applied Biosystems, Foster City, CA, USA). cpDNA trnL–trnF region strategies Double-stranded DNA of the trnL50 –trnF region was amplified with the trnL-c forward and trnL-f reverse primers (Taberlet et al. 1991). In some cases, trnL-d reverse and trnL-e forward were also used. The profile used for amplification of this region was as described in Susanna et al. (2006). Purification of the PCR products and direct sequencing of the amplified DNA segments were performed as for the ITS region. Analyses Nucleotide sequences were edited with Chromas 2.0 (Technelysium Pty Ltd, Tewantin, Australia) and aligned visually by sequential pairwise comparison (Swofford and Olsen 1990). Data matrices are available on request from the corresponding author. Parsimony analyses Parsimony analyses of the ITS alone, of the trnL-F alone and of the combined ITS–trnL-F dataset involved heuristic searches conducted with PAUP version 4.0b10 (Swofford 2002) by using TBR branch swapping, with character states specified as unordered and unweighted. The indels were coded as missing data. All most-parsimonious trees (MPTs) were saved. To locate other potential islands of MPTs (Maddison 1991), we performed 1000 replications with random taxon addition, also with TBR branch swapping. Bootstrap analyses (BS) (Felsenstein 1985) were performed using 100 replicates of heuristic search with the default options. For the strict consensus tree, consistency index (CI), retention index (RI) and homoplasy index (HI) are given, excluding uninformative characters. Bayesian inference Bayesian inference estimation of the ITS alone, of the trnL-F alone and of the combined ITS–trnL-F dataset was calculated with MrBayes 3.1.2 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). The best available model of molecular evolution required for Bayesian estimations of phylogeny was selected by using hierarchical likelihood ratio tests (hLRT) and Akaike information criteria (AIC) as implemented in the software MrModelltest 2.2 (Nylander 2004), which considers only nucleotide substitution models that are currently implemented in PAUP and MrBayes 3.1.2. The symmetrical model, with equal base frequencies and with variable sites assumed to follow a discrete gamma distribution SYM+G (Zharkikh 1994), was selected as the best-fit model of nucleotide substitution for the ITS dataset. The general time-reversible model (Rodríguez et al. N. Garcia-Jacas et al. 1990), with some sites assumed to be invariable (GTR+I; Hasegawa et al. 1985), was selected as the best-fit model of nucleotide substitution for the trnL-F dataset. Consequently, for the ITS–trnL-F combined dataset, two partitions were defined to use the SYM+G model for the ITS part and the GTR+I for the trnL-F part. The Bayesian inference analyses were initiated with random starting trees and were run for 106 generations. Four Markov chains were run by using the Markov Chain Monte Carlo (MCMC) principles to sample trees. We saved one of every 100 generations, resulting in 10 000 sample trees. The first 1000 trees (burnin) were excluded to avoid trees that might have been sampled before the convergence of the Markov chains. A majority-rule consensus tree was calculated with PAUP version 4.0b10 (Swofford 2002) from the last 9000 of the 10 000 trees sampled. Internodes with posterior probabilities
95% were considered statistically significant. A partition homogeneity test (ILD; Farris et al. 1995) was carried out to test the congruence of ITS and trnL-F datasets. This test (implemented in PAUP ver. 4.0b10; Swofford 2002) was accomplished excluding uninformative characters and by the use of a heuristic search and a simple addition of taxa for 1000 random partitions of the data. Results The numerical results of the analyses from the ITS and the ITS– trnL-F combined dataset are given in Table 2 (the results for the trnL-F alone are not reflected in the table). The result of the ILD test (P = 0.966; significant incongruence adjusted at P < 0.01) indicated congruence between nuclear and chloroplastic data, which allowed combining both regions. Both parsimony and Bayesian inference analyses showed highly congruent topologies. Therefore, for each dataset, results of both analyses were commented together on a unique phylogram obtained from the Bayesian analyses, where bootstrap values (BS) from the parsimony analysis are shown below branches, and Bayesian posterior probabilities (PP) above branches. Figure 1 corresponds to ITS analyses and Fig. 2 to the combined ITS–trnL-F analyses. The result of the analysis of the trnL–trnF region alone resulted in a large polytomy and thereafter it is not illustrated. All of the performed analyses confirmed the monophyly of the ingroup with high support values (ITS: BS = 99%, Table 2. Comparison of results from the ITS and ITS–trnL-F datasets The consistency, retention and homoplasy indices are calculated excluding uninformative characters Analysis Parsimony Number of taxa Total characters Informative characters Number of MPTs Consistency index (CI) Retention index (RI) Homoplasy index (HI) Bayesian inference Model of molecular evolution ITS ITS + trnL-F 35 652 141 802 0.5389 0.7820 0.4611 25 1478 158 20 0.5799 0.6948 0.4201 SYM+G ITS: SYM+G; trnL–F: GTR+I Conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae) Australian Systematic Botany 305 Olgaea eriocephala 2 0.97 Olgaea eriocephala 1 Olgaea nidulans 1 96 Olgaea vvedenskyi 1 80 Olgaea longifolia Alfredia acantholepis Alfredia nivea 1 99 Alfredia cernua Olgaea pectinata 1 95 1 1 100 89 Olgaea baldschuanica Olgaea petri-primi Olgaea chodshamuminensis Syreitschikovia spinulosa 1 83 Syreitschikovia tenuifolia Olgaea lomonossowii 1 99 Olgaea leucophylla Ancathia igniaria Xanthopappus subacaulis Synurus palmatopinnatifidus Lamyropappus schakaptaricus 1 99 Onopordum acanthium 1 86 Onopordum seravschanicum Onopordum acaulon Onopordum nogalesii Onopordum turcicum Onopordum tauricum Onopordum nervosum Onopordum leptolepis 1 100 Onopordum illyricum Onopordum carduchorum Onopordum anatolicum Onopordum caricum Staehelina dubia 1 100 Staehelina fruticosa Berardia subacaulis 0.1 Fig. 1. Phylogram obtained from the Bayesian analysis of ITS sequences. Bayesian posterior probabilities
0.94 are shown above branches. Bootstrap values from the parsimony analysis
80% are shown below branches. PP = 1.00; ITS-trnL-F: BS = 100%, PP = 1.00), as already suggested on a much broader sample of the tribe by Susanna et al. (2006). The ITS analyses showed two main clades with high support: one that contains all the Onopordum representatives (BS = 100%, PP = 1.00), and another that contains all the other genera of the 306 Australian Systematic Botany N. Garcia-Jacas et al. 1 100 1 86 Olgaea baldschuanica Olgaea petri-primi Olgaea chodshamuminensis 1 - A Ancathia igniaria 0.94 - Xanthopappus subacaulis Syreitschikovia spinulosa 1 90 B Syreitschikovia tenuifolia Lamyropappus schakaptaricus 0.94 - 1 100 Olgaea vvedenskyi Olgaea eriocephala 1 1 95 C Olgaea eriocephala 2 ASIAN GENERA Olgaea nidulans 1 87 Olgaea longifolia Alfredia nivea Alfredia acantholepis 1 100 D Alfredia cernua Olgaea pectinata 1 98 Synurus palmatopinnatifidus Olgaea lomonossowii 1 100 Onopordum nervosum ONOP. 1 100 E Olgaea leucophylla Onopordum tauricum Staehelina dubia 1 100 Staehelina fruticosa OUTGR. 1 100 Berardia subacaulis 0.1 Fig. 2. Phylogram obtained from the Bayesian analysis of the combined ITS–trnL–trnF dataset. Bayesian posterior probabilities
0.94 are shown above branches. Bootstrap values from the parsimony analysis
80% are shown below branches. ONOP: Onopordum. OUTGR: outgroup taxa. ‘Onopordum group’ (BS = 95%, PP = 1.00), including Alfredia, Ancathia, Xanthopappus, Lamyropappus, Olgaea, Synurus and Syreitschikovia. The same two clades also appeared with high support values in the combined analyses, although only two Onopordum representatives were included in this case. The position of Synurus palmatopinnatifidus Kitam. and the monotypic genera Ancathia, Xanthopappus and Lamyropappus was not well resolved in any analyses, although it is clear that they constitute a monophyletic group with Olgaea, Alfredia and Syreitschikovia. Some of the genera included in the ingroup are clearly monophyletic. The two species of Syreitschikovia were grouped together with strong support in all the analyses performed (ITS: BS = 83%, PP = 1.00; ITS–trnL-F: BS = 90%, PP = 1.00). The same occurred with the three species of Alfredia, which also constituted a well supported monophyletic group in all Conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae) analyses (ITS: BS = 99%, PP = 1.00; ITS–trnL-F: BS = 100%, PP = 1.00). In contrast, Olgaea appeared to be paraphyletic (if we accept Ancathia and Xanthopappus as different genera), although both the ITS and the combined analyses resulted in three well supported clades. The first clade contains four species, including Olgaea eriocephala (C.Winkl.) Iljin, O. nidulans (Rupr.) Iljin, O. vvedenskyi Iljin and O. longifolia (C.Winkl.) Iljin (ITS: BS = 96%, PP = 1.00; ITS–trnL-F: BS = 100%, PP = 1.00). The second one contains three species, including Olgaea baldschuanica (C.Winkl.) Iljin, O. petri-primi B.A.Sharipova and O. chodshamuminensis B.A.Sharipova (ITS: BS = 89%, PP = 1.00; ITS–trnL-F: BS = 86%, PP = 1.00). In the combined ITS–trnL-F analyses, Xanthopappus and Ancathia formed a weakly supported clade that is sister to the second group of Olgaea, without parsimony support but with strong Bayesian support (PP = 1.00). The third clade of Olgaea is composed by O. lomonossowii (Turcz.) Iljin and O. leucophylla (Trautv.) Iljin (ITS: BS = 99%, PP = 1.00; ITS–trnL-F: BS = 100%, PP = 1.00). However, the relationships between these groups of Olgaea and also with other genera were not well resolved. The clade formed by O. lomonossowii and O. leucophylla came out as sister to the remaining species and genera in the combined ITS–trnL-F Bayesian analysis, but without statistic support. Finally, the position of Olgaea pectinata was not well resolved in the parsimony analyses, whereas it appeared as sister of Alfredia clade in the Bayesian analyses, although without statistic support. Discussion The fact that the Onopordum group forms a monophyletic complex of genera was previously demonstrated by the combined phylogeny by Susanna et al. (2006). Present molecular results confirmed the existence of the following two main groups which can be defined within the complex on morphological, karyological and biogeographical grounds: the genus Onopordum on one hand, and the complex of Asian genera on the other. Onopordum All the species included in the study formed a polytomy in a sister clade to the species belonging to the Asian genera. In the phylogram we see that the branch length is very short. Everything indicates that despite morphological differences between the species of the genus, speciation in Onopordum has been rapid and explosive, a process probably linked to the development of the Mediterranean climate 3.2–2.8 million years ago during the Pliocene (Suc 1984). None of the markers used showed variability considerable enough to suggest anything other than generic uniformity. The Asian genera This complex of genera concentrates the overwhelming majority of taxonomic problems within the Onopordum group. The main difficulty for establishing a natural generic classification is that Olgaea, the genus with the greatest number of species, is not monophyletic according to the analyses of the combined two regions. The species of Olgaea included in the study appeared in Australian Systematic Botany 307 the polytomy with other Asian genera, and in separate well supported clades. Once established that the phylogeny failed to support the present taxonomic treatment of these exclusively Asian genera, we attempted to verify whether or not there were morphological or biogeographical relationships in the associations of the distinct clades. To do this, we followed the phylogeny of the combined dataset (Fig. 2). The first clade comprised Olgaea baldschuanica (type species of the genus), O. petri-primi and O. chodshamuminensis (Fig. 2, Clade A, upper branch). These three species of Olgaea have obvious morphological and biogeographical affinities. All of them are plants capable of attaining 100 cm or more in height, with long slender glabrous stems branched in the upper part and with bracts near the capitula. The basal leaves are long (24– 40 cm), not very wide (5.5–7 cm) and entire. O. petri-primi and O. chodshamuminensis are Tadjikistan endemics, whereas O. baldschuanica has a wider distribution across the Pamir Alai from Kazakhstan as far as the limits of the Pamir range (Iljin 1963). The Bayesian analysis of the combined dataset strongly supported that these three species of Olgaea have as sister group the representatives of two monotypic genera, Ancathia igniaria (Spreng.) DC. and Xanthopappus subacaulis C.Winkl. (Fig. 2; Clade A). There is no clear morphological relationship between these two genera, neither is there with the three species. A. igniaria is a species with a not very robust habit, 20–30-cm height and with linear or linear-lanceolate, entirespinulose leaves. It is the Asian genus with the greatest range, from the Caucasus to Mongolia. X. subacaulis is a QinghaiTibetan endemic, almost stemless, with a basal rosette of petiolate oblong-lanceolate leaves and yellow flowers (Wang et al. 2007). The only two species of the genus Syreitschikovia formed a monophyletic clade (Fig. 2; Clade B). These plants exhibit a more delicate habit than those of the other genera (5–40 cm in height). There is a clear morphological relationship between these two species, with slight variations in leaf form: in the case of S. tenuifolia (Bong.) Pavlov the leaves are linear-lanceolate, whereas in S. spinulosa they are oblong-lanceolate. We could say that the former is a more reduced form of S. spinulosa. The two species occur in the Tian Shan. In the following clade we have another group of Olgaea species, including O. nidulans, O. eriocephala, O. longifolia and O. vvedenskyi (Fig. 2; Clade C). The first three species are very similar in habit, size, indumentum, scape, involucral bract and capitula morphology. The leaves are spiny and variable in outline, but in most cases O. nidulans and O. eriocephala present less-divided leaves than does O. longifolia (spiny-pinnatifid). Within the inflorescence, a graduation from lax to dense corymbs or to agglomerate clusters of capitula (O. nidulans) can be observed, according to the species. O. vvedenskyi has more entire leaves with less spiny margins than the aforementioned species; the capitula are oblong-oval whereas in the other species of Olgaea they are hemispherical; the external involucral bracts end in a recurved spine, the next rows are upright, the terminal spine is reddish at the apex, and the innermost bracts are red and spineless. Also striking is the different indumentum between the bracts of O. vvedenskyi, which are tomentose, and the densely 308 Australian Systematic Botany lanate bracts of O. nidulans, O. eriocephala and O. longifolia. The four species are endemic to central Asia. O. eriocephala and O. longifolia are found in the Pamir Alai, and O. nidulans along with O. vvedenskyi occur in the Tian Shan. Regarding the three species of Alfredia, these grouped together with high support in the two analyses (ITS and combined ITS plus trnL-F, Fig. 2; Clade D). In the combined analysis, also O. pectinata appeared in this group without support (Fig. 2), although there is no morphological relationship between the species of Alfredia and O. pectinata. O. pectinata is a plant up to 30–70 cm high, with spiny pinnately lobed leaves; external and middle involucral bracts have a large terminal spine, and the internal ones a scarious appendage; capitula are oblong-oval. It is endemic to the western Tian Shan. The species of Alfredia range between 40 and ~150 cm in height and possess entire basal leaves with the margin denticulate to spiny, hastate, lobed and with a long petiole. The involucral bracts are clearly differentiated, with the external ones ending in a more or less long spine, the middle ones with a more or less membranous appendage with prickly apical and lateral fimbria, and the innermost being narrow-linear with a scarious appendage. The capitula are often nodding at anthesis. Alfredia is morphologically similar to Synurus owing to the leaf outline and great size, with its hastate shape and denticulate margin, and also the nodding capitula. The distribution area is central Asia, including Tian Shan and Dzungar Alatau. A. cernua (L.) Cass. can also be found in western Siberia. The last clade was formed by Olgaea lomonossowii and O. leucophylla (Fig. 2; Clade E), separated from the rest of Olgaea species. Morphologically, O. lomonossowii is a species with a less vigorous habit than for the majority of Olgaea species. The leaves are herbaceous, flat, with a ciliatespiny margin; the involucral bracts reddish and hardly spiny. O. leucophylla has wavy coriaceous leaves with a spiny margin, external involucral bracts green with a recurved spine, and middle and innermost bracts slightly reddish ending in a thin spine. However, there is a biogeographical relationship between the two species, with both growing in Mongolia. O. lomonossowii was considered to be a different genus, Takeikadzuchia; however, according to results of the ITS and ITS–trnL-F regions this new genus cannot be maintained, and in fact, it groups very closely with O. leucophylla, with very high support (BS = 100%, PP = 1.00). Also other authors had already suggested that Takeikadzuchia is very close to Olgaea (Bremer 1994; Grubov 2001). The morphological characteristics fall within the range of variability of the genus Olgaea and of the perennial species of the Onopordum group. Lamyropappus and Synurus appear isolated within the complex. In the analyses of ITS region and ITS–trnL-F (Figs 1, 2), their relationship with the other Asian genera remained undefined. However, morphologically they fall clearly within the Onopordum group. Lamyropappus is a monotypic genus endemic to central Asia and is found in the south-western Lake Balkhash area, eastern Syr Darya and southern Tian Shan. Whereas Synurus is the only genus in the group with a distribution area extending eastwards beyond the mountains of central Asia, from Mongolia to Japan. The ITS– N. Garcia-Jacas et al. trnL-F phylogeny suggested a basal position with respect to the other Asian genera, except the two Mongolian Olgaea species (with low support, PP = 0.94). This could indicate that species with a more oriental distribution show greater differentiation from the rest of the group. Taxonomic implications in the Asian genera Generally speaking, we can say that although it is true that in some groups we find a correlation between morphological and molecular evidence, in others, the biogeographical relationship is what predominates. However, even within the clades with sound morphological support, we found exceptions: e.g. in the group formed by O. nidulans, O. eriocephala, O. longifolia and O. vvedensky (Fig. 2, Clade C), the last species showed morphological differences in regard to the other three. These exceptions do not allow us to state that there is a clear correlation in all cases. The results suggest that the generic classification needs a revision; in particular, present delineation of Olgaea is not supported by molecular evidence. Segregating Ancathia or Xanthopapus would leave Olgaea paraphyletic (Fig. 2, Clade A). To arrive at a more natural classification we could consider two alternatives. First, a broad proposal, in which all Asian genera were to be reduced to one genus with sectional or subgeneric divisions. If this proposal were taken up, the valid name for the genus would be Alfredia. The second proposal would be to make more analytic taxonomic divisions, where the genus Olgaea would end up dispersed into various other genera. We see the first option in a more favourable light, but there are obvious difficulties in creating a clear sectional or subgeneric division following the phylogeny ITS–trnL-F. The existing genera Alfredia, probably Synurus and Syreitschikovia would each form a natural section, and the genera Ancathia, Lamyropappus, Olgaea pectinata and Xanthopappus would in each case form a monotypic section. However, it is difficult to propose a sectional classification for the remaining Olgaea species because, as we have seen above, in some cases there is no morphological correlation. The more detailed morphological study we are carrying out, including new characters such as pollen types, may help find a new boundary for the genera, but the solution is very complicated. This is a case parallel to that of the related group Arctium, in which the natural classification of the genera Arctium, Cousinia, Hypacanthium Juz. and Schmalhausenia C.Winkl. has turned out to be an almost insoluble problem because of the lack of correlation between morphological and molecular data (López-Vinyallonga et al. In press). Biogeographical considerations and plant habit of the Onopordum group: a comparison with some other Cardueae of similar biogeographical and life strategies The existence of an Onopordum group made up of Asian genera on one hand (Alfredia, Ancathia, Lamyropappus, Olgaea, Synurus and Syreitschikovia) and Onopordum on the other is now confirmed as much on molecular as on morphological grounds, but the three main differences between the two groups call for attention. Conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae) The area of distribution is the first difference that stands out. The distribution of the perennial genera is exclusively central and eastern Asia, whereas Onopordum has a Mediterranean distribution (and in the second place cosmopolitan, with the genus comprising colonising species). The relatively young age of the genus Onopordum, on the basis of very low molecular variability, together with the exclusively Asian circumscription of its relatives, suggests a central Asian or Irano-Turanian origin for the genus. The second difference is the habit, or life form, with the Asian genera being perennial, whereas Onopordum is biennial. The third difference has to do with the chromosome number. In the Asian genera, the basal numbers are x = 12 and 13, according to the few counts encountered in the literature and made by us, whereas in Onopordum the number is higher at x = 17 (Watanabe 2008). This panorama of biogeographical, karyological and behavioural differences between the Asian genera and Onopordum can also be observed in the Arctium–Cousinia complex which, as we have stated, presents some very similar taxonomic problems. Like the Asian genera, Cousinia has a central Asian and Irano-Turanian distribution; the species are mainly perennial, there being very few annuals (16 of ~600 species); and the basic numbers are x = 9, 10, 11, 12 and 13. Arctium, in contrast, exhibits a cosmopolitan distribution as a coloniser, in the same way as Onopordum; its habit is biennial and the basic chromosome number is x = 18 (Watanabe 2008). The superior colonisation ability of some species, as compared with others, could be attributed to two factors: polyploidy and habit. Both Arctium and Onopordum are ancient polyploids as suggested by their high basic numbers (x = 17 and 18; cf.; Grant 1963, 1981; Soltis et al. 2003). This polyploid origin, as traditionally sustained, grants them a superior level of adaptability (Levin 1983; Thompson and Lumaret 1992; Soltis and Soltis 2000). Regarding habit, three of the most successful colonising genera in the Cardueae (Arctium, Onopordum and Silybum) are biennials (Susanna and Garcia-Jacas 2007). It would be worth verifying whether biennial life form is particularly well adapted to disturbed habitats. Acknowledgements Financial support from the Spanish Ministery of Education and Science (Projects BOS2001-3041-C02-02 and CGL2004-04563-C02-01/BOS) and the Generalitat de Catalunya (Ajuts a Grups de Recerca Consolidats 2005/ SGR/00344) is gratefully acknowledged. Authors thank the Komarov Botanical Institute, the Berlin-Dahlem, Copenhagen, Dijon and Kyoto Botanical Gardens for providing plant material; collectors (as listed in Table 1) for their help in collecting the material; N. Montes for technical assistance; and S. Pyke for kindly improving the English. Three anonymous reviewers made very useful suggestions for improving the manuscript. References Aiton W (1789) ‘Hortus Kewensis; or, a catalogue of the plants cultivated in the Royal botanic garden at Kew, 3.’ (G. Nicol: London) Arano H (1964) Cytotaxonomic studies in subfam. Carduoideae of Japanese Compositae. X. The karyotype analysis in some species of Saussurea and its related genera. Kromosomo 57–59, 1869–1875. Boissier E (1846) ‘Diagnoses plantarum orientalium novarum, ser. 1, 6. B. ’ (Herrmann: Leipzig) Australian Systematic Botany 309 Bremer K (1994) ‘Asteraceae. Cladistics & classification.’ (Timber Press: Portland, OR) Chuksanova PA, Sveshnikova LT, Alexandrova TV (1968) A new evidence on chromosome numbers in species of the family Compositae Giseke. Citologija 10, 381–386. Cullings KW (1992) Design and testing of a plant-specific PCR primer from ecological and evolutionary studies. Molecular Ecology 1, 233–240. doi: 10.1111/j.1365-294X.1992.tb00182.x Dittrich M (1977) Cynareae—systematic review. In ‘The biology and chemistry of the Compositae’. (Eds VH Heywood, JB Harborne, BL Turner) pp. 999–1015. (Academic Press: London) Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19, 11–15. Farris JS, Källersjö M, Kluge AG, Bult C (1995) Testing significance of incongruence. Cladistics 10, 315–319. doi: 10.1111/j.1096-0031.1994.tb00181.x Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791. doi: 10.2307/2408678 Fischer H, Jensen U (1990) Phytoserological investigation of the tribe Cardueae s.l. (Compositae). Plant Systematics and Evolution. Supplementum 4, 99–111. Galbany-Casals M, Garcia-Jacas N, Susanna A, Sáez L, Benedí C (2004) Phylogenetic relationships in the Mediterranean Helichrysum (Asteraceae, Gnaphalieae) based on nuclear rDNA ITS sequence data. Australian Systematic Botany 17, 241–253. doi: 10.1071/SB03031 Garcia-Jacas N, Susanna A, Mozaffarian V, Ilarslan R (2000) The natural delimitation of Centaurea (Asteraceae: Cardueae): ITS sequence analysis of the Centaurea jacea group. Plant Systematics and Evolution 223, 185–199. doi: 10.1007/BF00985278 Garcia-Jacas N, Susanna A, Garnatje T, Vilatersana R (2001) Generic delimitation of phylogeny of the subtribe Centaureinae (Asteraceae): a combined nuclear and chloroplastic DNA analysis. Annals of Botany 87, 503–515. doi: 10.1006/anbo.2000.1364 Garcia-Jacas N, Garnatje T, Susanna A, Vilatersana R (2002) Tribal and subtribal delimitation and phylogeny of the Cardueae (Asteraceae): a combined nuclear and chloroplast DNA analysis. Molecular Phylogenetics and Evolution 22, 51–64. doi: 10.1006/mpev.2001.1038 Garcia-Jacas N, Uysal T, Romashchenko K, Suárez-Santiago VN, Ertugrul K, Susanna A (2006) Centaurea revisited: a molecular survey of the Jacea group. Annals of Botany 98, 741–753. doi: 10.1093/aob/mcl157 Garnatje T, Susanna A, Garcia-Jacas N, Vilatersana R, Vallès J (2005) A first approach to the molecular phylogeny of the genus Echinops L. (Asteraceae): sectional delimitation and relationships with the genus Acantholepis Less. Folia Geobotanica 40, 407–419. doi: 10.1007/BF02804288 Grant V (1963) ‘The origin of adaptations.’ (Columbia University Press: New York) Grant V (1981) ‘Plant speciation.’ (Columbia University Press: New York) Grubov VI (2001) ‘Key to the vascular plants of Mongolia, 2.’ (Science Publishers: Plymouth) Gurzenkov NN (1973) Studies of chromosome number of plants from the south of the Soviet Far East. Komarov Lectures 20, 47–61. Häffner E (2000) On the phylogeny of the subtribe Carduinae (tribe Cardueae, Compositae). Englera 21, 1–209. Häffner E, Hellwig FH (1999) Phylogeny of the tribe Cardueae (Compositae) with emphasis on the subtribe Carduinae: an analysis based on ITS sequence data. Willdenowia 29, 27–39. Hasegawa M, Kishino H, Yano T (1985) Dating the human–ape split by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22, 160–174. doi: 10.1007/BF02101694 310 Australian Systematic Botany Hidalgo O, Garcia-Jacas N, Garnatje T, Susanna A (2006) Phylogeny of Rhaponticum (Asteraceae, Cardueae–Centaureinae) and related genera inferred from nuclear and chloroplast DNA sequence data: taxonomic and biogeographic implications. Annals of Botany 97, 705–714. doi: 10.1093/aob/mcl029 Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogenenetic trees. Bioinformatics 17, 754–755. doi: 10.1093/bioinformatics/17.8.754 Iljin MM (1963) Rod 1591. Olgaea Iljin. In ‘Flora SSSR. Vol. 28’. (Eds EG Bobrov, SK Czerepanov) pp. 53–63. (Bishen Singh Mahendra Pal Singh & Koeltz Scientific Books: Dehra Dun, India) Iljin MM, Semidel GA (1963) Rod 1609. Syreitschikovia Pavlov. In ‘Flora SSSR. Vol. 28’. (Eds EG Bobrov, SK Czerepanov) pp. 376– 379. (Bishen Singh Mahendra Pal Singh & Koeltz Scientific Books: Dehra Dun, India) Kazmi SMA (1963) Revision der Gattung Carduus (Compositae). Teil I. Mitteilungen aus der Botanischen Staatssammlung München 5, 139–198. Kazmi SMA (1964) Revision der Gattung Carduus (Compositae). Teil II. Mitteilungen aus der Botanischen Staatssammlung München 5, 279–550. Kelch DG, Baldwin BG (2003) Phylogeny and ecological radiation of New World thistles (Cirsium, Cardueae–Compositae) based on ITS and ETS rDNA sequence data. Molecular Ecology 12, 141–151. doi: 10.1046/j.1365-294X.2003.01710.x Kita Y, Fujikawa K, Ito M, Ohba H, Kato M (2004) Molecular phylogenetic analyses and systematics of the genus Saussurea and related genera (Asteraceae, Cardueae). Taxon 53, 679–690. doi: 10.2307/4135443 Kitamura S (1934) Compositae novae japonicae VII. Acta Phytotaxonomica et Geobotanica 3, 97–111. Krasnikov AA, Zhirova OS, Lomonosova MN, Smirnov SV (2003) Chromosome numbers of Asteraceae from the southern Siberia and Kazakstan. Botanicheskii Zhurnal 88, 151–153. Levin DA (1983) Polyploidy and novelty in flowering plants. American Naturalist 122, 1–25. doi: 10.1086/284115 Linnaeus C (1753) ‘Species plantarum II.’ (Holmiae Impensis Laurentii Salvii: Stockholm) López-Vinyallonga S, Mehregan I, Garcia-Jacas N, Tscherneva O, Susanna A, Kadereit JW (In press) Phylogeny and evolution of the Arctium-Cousinia complex (Compositae, Cardueae–Carduinae). Taxon, in press. Maddison DR (1991) The discovery and importance of multiple islands of most-parsimonious trees. Systematic Zoology 40, 315–328. doi: 10.2307/2992325 Malakhova LA (1990) Kariologocheskij analiz primodnykh populjacij redkich i ischezajushikh rastenij na juge Tomskoj Oblasti. Byulleten Glavnogo Botanicheskogo Sada 155, 60–66. [Karyological analysis of natural populations of rare or endangered plants from the south of the province of Tomsk] Martins L, Hellwig FH (2005) Systematic position of the genera Serratula and Klasea (Cardueae, Asteraceae) inferred from ETS and ITS sequence data and new combinations in Klasea. Taxon 54, 632–638. Nylander JAA (2004) ‘MrModeltest ver. 2.’ Program distributed by the author. Evolutionary Biology Centre, Uppsala University, Uppsala. Pavlov NV (1933) Syreitschikovia, genus novum Compositarum ex Media Asia. Feddes Repertorium Specierum Novarum Regni Vegetabilis 31, 192–193. von Raab-Straube E (2003) Phylogenetic relationships in Saussurea (Compositae, Cardueae) sensu lato, inferred from morphological, ITS and trnL–trnF sequence data, with a synopsis of Himalaiella gen. nov., Lipschitziella and Frolovia. Willdenowia 33, 379–402. Rodríguez F, Oliver JL, Marín A, Medina JR (1990) The general stochastic model of nucleotide substitution. Journal of Theoretical Biology 142, 485–501. doi: 10.1016/S0022-5193(05)80104-3 N. Garcia-Jacas et al. Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574. doi: 10.1093/bioinformatics/btg180 Rostovtseva TS (1979) Chromosome numbers of some species of the family Asteraceae Dumort. Botanicheskii Zhurnal 64, 582–589. Rostovtseva TS (1983) Chromosome numbers of some species of the family Asteraceae II. Botanicheskii Zhurnal 68, 1538–1543. Smirnov SV (2001) Uto takoe Olgaea altaica (Asteraceae)? Turczaninowia 4, 18–22. [What is Olgaea altaica (Asteraceae)?] Soltis DE, Soltis PS (2000) The role of genetic and genomic attributes in the success of polyploids. Proceedings of the National Academy of Sciences, USA 97, 7051–7057. doi: 10.1073/pnas.97.13.7051 Soltis DE, Soltis PS, Tate JA (2003) Advances in the study of polyploidy since plant speciation. New Phytologist 161, 173–191. Suárez-Santiago VN, Salinas MJ, Garcia-Jacas N, Soltis PS, Soltis DE, Blanca G (2007) Evolution by reticulation of the Acrolophus subgroup (Centaurea L., Compositae) in the occidental Mediterranean: origin and diversification of the section Willkommia Blanca. Molecular Phylogenetics and Evolution 43, 156–172. doi: 10.1016/j.ympev.2006.08.006 Suc JP (1984) Origin and evolution of the Mediterranean vegetation and climate in Europe. Nature 307, 429–432. doi: 10.1038/307429a0 Sun Y, Skinner DZ, Liang GH, Hulbert SH (1994) Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacers of nuclear ribosomal DNA. Theoretical and Applied Genetics 89, 26–32. doi: 10.1007/BF00226978 Susanna A, Garcia-Jacas N (2007) Tribe Cardueae. In ‘Flowering plants. Eudicots. Asterales. Vol. 8’. (Eds JW Kadereit, C Jeffrey) pp. 123–146. In ‘The families and genera of vascular plants’. (Ed. K Kubitzki) (Springer Verlag: Berlin) Susanna A, Garcia-Jacas N, Soltis DE, Soltis PS (1995) Phylogenetic relationships in tribe Cardueae (Asteraceae) based on ITS sequences. American Journal of Botany 82, 1056–1068. doi: 10.2307/2446236 Susanna A, Garnatje T, Garcia-Jacas N (1999) Molecular phylogeny of Cheirolophus (Asteraceae: Cardueae–Centaureinae) based on ITS sequences of nuclear ribosomal DNA. Plant Systematics and Evolution 214, 147–160. doi: 10.1007/BF00985736 Susanna A, Garnatje T, Garcia-Jacas N, Vilatersana R (2002) On the correct subtribal placement of the genera Syreitschikovia and Nikitinia (Asteraceae, Cardueae): Carduinae or Centaureinae? Botanical Journal of the Linnean Society 140, 313–319. doi: 10.1046/j.1095-8339.2002.00104.x Susanna A, Garcia-Jacas N, Vilatersana R, Garnatje T (2003) Generic boundaries and evolution of characters in the Arctium group: a nuclear and chloroplast DNA analysis. Collectanea Botanica (Barcelona) 26, 101–118. Susanna A, Garcia-Jacas N, Hidalgo O, Vilatersana R, Garnatje T (2006) The Cardueae (Compositae) revisited: insights from ITS, trnL–trnF, and matK nuclear and chloroplast DNA analysis. Annals of the Missouri Botanical Garden 93, 150–171. doi: 10.3417/0026-6493(2006)93[150:TCCRIF]2.0.CO;2 Swofford DL (2002) ‘PAUP*. Phylogenetic analysis using parsimony (*and other methods), ver. 4.0b10.’ (Sinauer: Sunderland, MA) Swofford DL, Olsen GJ (1990) Phylogeny reconstruction. In ‘Molecular systematics’. (Eds D Hillis, C Moritz) pp. 411–501. (Sinauer: Sunderland, MA) 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. doi: 10.1007/BF00037152 Taniguchi K, Tanaka R, Yonezawa Y, Komatsu H (1975) Types of banding patterns of plant chromosomes by modified BSG method. Kromosomo 100, 3123–3135. Conflicting generic delineation in the Onopordum group (Compositae, Cardueae–Carduinae) Thompson JD, Lumaret R (1992) The evolutionary dynamics of polyploid plants: origins, establishment and persistence. Trends in Ecology & Evolution 7, 302–307. doi: 10.1016/0169-5347(92)90228-4 Vilatersana R, Susanna A, Garcia-Jacas N, Garnatje T (2000) Generic delimitation and phylogeny of the Carduncellus–Carthamus complex (Asteraceae) based on ITS sequences. Plant Systematics and Evolution 221, 89–105. doi: 10.1007/BF01086383 Wang Y-J, Liu J-Q, Miehe G (2007) Phylogenetic origins of the Himalayan endemic Dolomiaea, Diplazoptilon and Xanthopappus (Asteraceae: Cardueae) based on three DNA Regions. Annals of Botany 99, 311–322. doi: 10.1093/aob/mcl259 Watanabe K (2008). Index to chromosome numbers in Asteraceae. http:// www-asteraceae.cla.kobe-u.ac.jp/index.html [Accessed 1 April 2008]. Australian Systematic Botany 311 White TJ, Bruns T, Lee S, Taylor J (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In ‘PCR protocols: a guide to methods and applications’. (Eds MA Innis, DH Gelfand, JJ Sninsky, TJ White) pp. 315–322. (Academic Press: San Diego, CA) Zharkikh A (1994) Estimation of evolutionary distances between nucleotide sequences. Journal of Molecular Evolution 39, 315–329. doi: 10.1007/BF00160155 Zhukova PG (1967) Karyology of some plants, cultivated in the Arctic–Alpine Botanical Garden. In ‘Plantarum in Zonam Polarem Transportatio II’. (Ed. NA Avrorin) pp. 139–149. (Akademiia Nauk SSSR: Leningrad) Manuscript received 8 April 2008, accepted 21 August 2008 http://www.publish.csiro.au/journals/asb