Molecular Phylogenetics and Evolution 61 (2011) 515–520
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
On the origin of the woody buckwheat Fagopyrum tibeticum (=Parapteropyrum tibeticum) in the Qinghai–Tibetan Plateau ⇑ Xinmin Tian a, Jian Luo b, Ailan Wang c, Kangshan Mao a, Jianquan Liu a, a Molecular Ecology Group, Key Laboratory of Arid and Grassland Ecology, Lanzhou University, Lanzhou 730000, Gansu, PR China b College of Tibet Agriculture and Animal Husbandry, Nyingchi 860000, Tibet, PR China c School of Life Science, Ludong University, Yantai 264000, Shandong, PR China article info abstract
Article history: Here we tested whether ‘insular woodiness’, a striking evolutionary pattern that commonly occurs on Received 31 March 2011 islands, has also appeared in QTP continental endemics. Parapteropyrum, a monotypic shrubby genus Revised 20 June 2011 occurring in the central QTP, has been previously placed in the tribe Atraphaxideae of the family Polyg- Accepted 5 July 2011 onaceae, while all the other woody species of this tribe mainly occur in western and central Asia. We Available online 21 July 2011 studied sequence variations of nuclear ITS (internal transcribed spacer) and cp (chloroplast) DNA (rbcL and accD) of this genus and the other ten genera. The constructed phylogenies based on ITS, cpDNA or Keywords: a combination of both datasets, suggest that the woody Parapteropyrum is nested within and most likely Parapteropyrum evolved from the herbaceous Fagopyrum. We propose that the large-scale uplift of the QTP not only pro- Fagopyrum Atraphaxideae moted continental species radiation, but also the secondary feature of woodiness in a few herbaceous lin- Phylogeny eages in response to strong selection pressures, similar to those acting on island flora. In addition, the Insular woodiness confirmation of Parapteropyrum within Fagopyrum highlights its potential use as a new, perennial source of buckwheat. Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction creation of new, unoccupied habitats for plants (Hughes and Eastwood, 2006). Nevertheless, it is not known whether insular Since habitats on islands are largely unoccupied initially, colo- woodiness also appeared in continental regions as a result of rapid nizing plants have frequently diversified extensively, resulting in evolution of plants in isolated habitats. radiative speciation, convergent morphological evolution and the Parapteropyrum, a genus endemic to the QTP, has been previ- appearance (inter alia) of woody species in predominantly herba- ously placed in the woody tribe Atraphaxideae, which includes ceous genera (Darwin, 1859; Carlquist, 1970, 1974; Givnish, three other woody genera (Atraphaxis, Pteropyrum and Calligonum) 1998; Ballard and Sytsma, 2000). This phenomenon of ‘insular (Fig. 1). Their distributions have been suggested to represent typi- woodiness’ has attracted interest from numerous evolutionary cal eastern–western Eurasian disjunctions (Sun, 2002; Wu, 2004). biologists since Darwin (1859), and all recent molecular studies Parapteropyrum contains only one species, Parapteropyrum tibeti- indicate that various woody species originated independently from cum, which occurs sparsely in dry valleys along the Yarlung Zangbo herbaceous ancestors (e.g. Böhle et al., 1996; Givnish, 1998; Panero river in Tibet, at elevations between 3000 and 3500 m (Li, 1981). Li et al., 1999). Island-like radiation and subsequent convergent evo- (1981) suggested that this recently established genus may be clo- lution has also been repeatedly observed in continental biodiver- sely related to Pteropyrum because of their shared woody charac- sity hotspots (such as the Qinghai–Tibetan Plateau, QTP), where teristics, and similar leaf clusters at each woody node. However, extensive geological episodes (for example, mountain uplift) or cli- Hong (1995, 1998) examined pollen traits of the Atraphaxideae matic changes have occurred (Richardson et al., 2001; Hughes and and found none that united these two genera. Furthermore, Eastwood, 2006; Liu et al., 2006; Wang et al., 2005, 2009; Givnish Parapteropyrum differs from Pteropyrum in both inflorescences et al., 2009; Tu et al., 2010; Zhang et al., 2011). This is because and stamen composition (Bao and Li, 1993), and while Parapteropy- (analogously to events following island formation) major geologi- rum has 48 (2n) chromosomes, Calligonum and Atraphaxis have 18 cal or climatic changes in continental regions often result in the and 22, respectively (Tian et al., 2009). In addition, while other spe- cies of the Polygonaceae that occur at high altitudes on the QTP are herbaceous (Li, 1981), the other three genera of the Atraphaxideae ⇑ Corresponding author. Address: Key Laboratory of Arid and Grassland Ecology, College of Life Science, Lanzhou University, Lanzhou 730000, Gansu, PR China. Fax: are mainly distributed in arid deserts of central Asia at elevations +86 931 8914288. below 1000 m (Fig. 1). All this evidence casts doubt on the validity E-mail addresses: [email protected], [email protected] (J. Liu). of including Parapteropyrum in the Atraphaxideae.
1055-7903/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2011.07.001 516 X. Tian et al. / Molecular Phylogenetics and Evolution 61 (2011) 515–520
Fig. 1. The distribution and morphology of Fagopyum tibeticum and closely-related genera. (A) Distributions of five genera, and photographs of: (B) Fagopyum cymosum, (C) F. cymosum, (D and E) F. urophyllum, and (F and G) Fagopyum (Parapteropyrum) tibeticum.
Molecular evidence has raised similar doubts. Although a cla- material that they sampled. Hence, to clarify these issues there is a distic analysis of morphological traits supported the monophyly need to re-examine the phylogenetic relationships of these genera of the Atraphaxideae (Tavakkoli et al., 2008), molecular analyses based on plant material that was not used in either of these two based on cp (chloroplast) DNA (ndhF) and ITS (internal transcribed previous studies. spacers) sequences suggested weak support for a lineage compris- Around 12–16 species have been recorded for Fagopyrum, most ing Calligonum, Pteropyrum and Pteroxygonum was weakly sup- of which occur on the QTP and in adjacent regions (Ohnishi and ported (Sanchez et al., 2009). A further phylogenetic analysis of Matsuoka, 1996; Li, 1998; Ohnishi, 1998). Previous phylogenetic Calligonum and Atraphaxis based on ITS and cpDNA trnL-F se- analyses (e.g. Liu et al., 2000, 2006; Friesen et al., 2000) of the quences, however, rejected the monophyly of this tribe and sug- QTP endemic genera or species suggest that most of them origi- gested that Parapteropyrum grouped with two herbaceous nated in situ. Therefore, it seems highly likely that Parapteropyrum Fagopyrum species (Tavakkoli et al., 2010). In addition, the results is not a natural member of the Atraphaxideae and may, instead, of the later study also suggested that the ITS and chloroplast have originated from the local herbaceous lineages (for example, DNA sequences of Calligonum and Pteropyrum reported by Sanchez Fagopyrum) of Polygonaceae on the QTP. In the study presented et al. (2009) may be erroneous, due to the misidentification of the here we tested this hypothesis, and used Parapteropyrum as a X. Tian et al. / Molecular Phylogenetics and Evolution 61 (2011) 515–520 517 model system to test whether insular woodiness has occurred (in gaps were treated as missing characters in all phylogenetic analy- an analogous fashion to its occurrence on islands) in QTP continen- ses we performed. We used PAUP version 4.010b (Swofford, 2002) tal endemics. for maximum parsimony (MP) and maximum likelihood (ML) anal- yses. MP trees were constructed using 100 repetitions of random sequence additions of the taxa. Starting trees were obtained by 2. Material and methods stepwise addition, and branch-swapping was performed using the TBR option. The best-fit ML substitution model for each dataset 2.1. Taxon sampling, DNA extraction, amplification and sequencing was selected by MODELTEST (Posada and Crandall, 1998) and ML trees were constructed using the simple addition, TBR branch- We included 11 genera of the Polygonaceae occurring on the swapping, MULTREES and COLLAPSE options. Support for branches QTP and from central Asia in our phylogenetic analyses including: was tested by bootstrap analysis with 1000 replicates (Felsenstein, Parapteropyrum, Pteropyrum, Atraphaxis, Calligonum, Rheum, Rumex, 1985; Guindon and Cascuel, 2003). Bayesian trees with posterior Oxyria, Fallopia, Polygonum, Konegia and Fagopyrum, with Limonium probabilities (PP) were inferred using MrBayes version 3.12 aureum (Plumbaginaceae) as an outgroup (Liedo et al., 1998). We (Huelsenbeck and Ronquist, 2001) and the same model of DNA included most described Fagopyrum species in our final analyses. evolution as for the ML analysis. Four simultaneous Monte Carlo All species of Fagopyrum are annual or perennial herb, while Markov Chain (MCMC) analyses were run for 2,000,000 genera- Parapteropyrum is a woody shrub (Fig. 1). Of the sequences used tions, saving one tree every 100 generations. The last 16,000 trees in this study, 49 were newly generated, including those for all spe- were used to calculate posterior possibilities (PP). Based on the cies of Parapteropyrum, Calligonum, Pteropyrum and Atraphaxis (for consensus tree (with branch length) from the Bayesian analysis accession numbers see Appendix A). For Fagopyrum, we initially se- of the combined nuclear and chloroplast datasets, we further con- quenced four species representing the two major clades identified structed the evolutionary history of the herbaceous versus woody by Yasui and Ohnishi (1998a,b) and Ohsako and Ohnishi (2000). habit. We inferred the likelihood ancestral states using Mesquite Since the obtained Fagopyrum sequences and constructed phylog- (Maddison and Maddison, 2004) based on the Mk1 Model (Markov eny were consistent with findings of previous studies (see Yasui 1 parameter). and Ohnishi, 1998a,b; Ohsako and Ohnishi, 2000), we included all Fagopyrum species reported by these researchers in our final analyses. Therefore, all sequences used for the analyses presented 2.3. Divergence estimation here are independent of those used by Sanchez et al. (2009) and Tavakkoli et al. (2010). Voucher species and provenances of all spe- We only used the ITS dataset to estimate divergence times be- cies sampled for the first time are listed in Appendix A. tween the recovered clades for two reasons. First, the correspond- We extracted DNA using a modified CTAB method (Doyle and ing cpDNA sequences of Pteropyrum have not been obtained Doyle, 1987). In the following amplification and sequencing, we despite repeated attempts. Second, in the cpDNA phylogeny tree, used the primers designed for cpDNA rbcL and accD and for nrITS we found several long branches, which may increase errors when by the previous researchers (Saiki et al., 1988; Zurawski and Clegg, estimating divergence times. We used a likelihood-ratio test 1987; White et al., 1990; Fay et al., 1997; Pryer et al., 2001) (LRT: Huelsenbeck and Rannala, 1997) implemented in PAUP ver- (Appendix B). The PCR amplification was conducted with 25 lL sion 4.010b (Swofford, 2002) to test the hypothesis of a molecular PCR reaction mixtures including 10–40 ng template DNA, 2.5 lLof clock (comparing the log likelihood (ln L) of the ML trees with and
10Â Taq polymerase reaction buffer, 0.525 lL of 50 mmol/L MgCl2, without assuming a molecular clock). Since the hypothesis of a 0.5 lL of 10 mmol/L dNTP solution, 1 lL of each primer (5 mol/lL), molecular clock was rejected (GTR + G, for the ITS data set: 1.0 unit rTaq DNA Polymerase (TAKARA, Dalian, China) and sterile 2Dln L = 203.07, d.f. = 24, P < 0.001), we used BEAST version 1.5.4 double-distilled water. The PCR amplification program consisted of (Drummond et al., 2007) to estimate the genetic divergences based an initial template denaturation step at 95 °C for 5 min, followed on an uncorrelated lognormal relaxed molecular clock tree. Follow- by 38 cycles of 94 °C for 1 min, 52 °C for 50 s, 72 °C for 1.20 min, ing a burn-in of 1,000,000 steps, all parameters were sampled once and a final extension step of 72 °C for 8 min. We then purified every 1000 steps from 40,000,000 MCMC steps. We checked con- the PCR products with a CASpure PCR Purification Kit (CAS ARRAY, vergence of the chains to the stationary distribution by visual Shanghai, China) and sequenced them (at Lanzhou University) inspection of plotted posterior estimates using the program TRA- using an ABI3130SL Automated DNA Sequencer. ITS products of CER. The effective sample size for each parameter sampled from three species could not be sequenced directly. They were therefore the MCMC analysis was found to exceed 500, usually by an order ligated and cloned with a pGEM-T vector system, Promega, of magnitude. The analysis was repeated and the samples from Madison, Wisconsin, USA). Since the sequenced clones from the the two runs were combined. We then estimated the genetic diver- same species always clustered together as a single group, only gences between the clades. We then estimated the genetic diver- one of each of these sets of clones was chosen to represent each gence times between the clades, using ITS substitution rates of À9 species in the final phylogenetic analyses that included all species 3.3 and 7.9 Â 10 substitutions per site per year (Richardson and genera. et al., 2001), the reported range for most perennial herb or shrub genera, to estimate ages of crown and stem lineages of Fagopyrum as well as the divergence of P. tibeticum from its sister clade. 2.2. Phylogenetic analyses
All finally selected sequences were aligned by CLUSTALX ver- 3. Results sion 1.84 (Thompson et al., 1997), followed by manual adjustments in MEGA4 (Tamura et al., 2007). Three datasets (ITS, cpDNA (rbcL The aligned ITS dataset contained a total of 742 bp. Of these and accD and combined ITS and cpDNA) were used for the phylo- characters, 534 were variable (182 parsimony-uninformative and genetic analyses. We employed an incongruence length difference 352 parsimony-informative) while 208 were constant. Most of (ILD) test (Farris et al., 1995) to evaluate the congruence between the Calligonum and Pteropyrum sequences recovered here, based the cpDNA and ITS datasets. The ILD test was conducted with 1000 on independent collections, are largely identical or similar to those replicates of the heuristic searches using TBR branch-swapping reported by Tavakkoli et al. (2010), but different from those re- with 100 random sequence additions in PAUP version 4.010b; all ported by Sanchez et al. (2009). For ITS sequences, the topologies 518 X. Tian et al. / Molecular Phylogenetics and Evolution 61 (2011) 515–520
100/100 Fagopyrum leptopodum --/58 Fagopyrum statice
/ 100/100 93 100 Fagopyrum lineare Fagopyrum capillatum 100/99 Fagopyrum gracilipes 100/100 Fagopyrum urophyllum
99/99 Parapteropyrum tibeticum
90/100 Fagopyrum esculentum
100/100 Fagopyrum homotropicum Fagopyrum tataricum 92/100 --/85 Fagopyrum cymosum
93/99 Atraphaxis decipiens --/98 Atraphaxis spinosa 100/100 Atraphaxis pungens --/87 100/100 Atraphaxis bracteata Fallopia aubertii
98/99 Polygonum hookeri / 56 89 98/99 Koenigia islandica Polygonum viviparum
58/66 Rumex crispus Oxyria digyna 85/100 Rheum officinale
100/100 Calligonum arborescens 100/100 Calligonum rubicundum Calligonum caput -medusae Limonium aureum 0.01 substitutions/site Fig. 3. Maximum-likelihood reconstruction of habit evolution (herbaceous, white; Fig. 2. Maximum likelihood tree based on combined ITS and cpDNA datasets. black, woody) in the Polygonaceae. Circle size is proportional to the probability Numbers before slashes are MP bootstrap values and the numbers after slashes frequency. indicate Bayes posterior probabilities. Values <50% not shown. were completely congruent in the MP, ML and Bayesian analyses icum and Fagopyrum esculentum) and two wild species, also distrib- (Appendix 3). Four genera of the tribe Atraphaxideae grouped with uted on the QTP. These two clades and the inter-clade relationships other genera and Parapteropyrum nested within Fagopyrum, show- with the Parapteropyrum clade received high statistical support ing no any affinity with Pteropyrum. In addition to Parapteropyrum, (Fig. 2). The ancestral state reconstruction based on the Bayesian two distinct clades were identified among the sampled Fagopyrum tree of the combined dataset suggested that the woody habit species. P. tibeticum grouped with one clade with high support may have evolved repeatedly in the Polygonaceae and that the her- (BP = 100, PP = 100). The divergence between Parapteropyrum and baceous habit seems to be the mostly likely ancestral state of the its sister subclade in Fagopyrum occurred between 6.4 and Fagopyrum lineage (89% herbaceous versus 11% woody) (Fig. 3). 14.8 Mya, based on the slow and fast rates of ITS substitution as- sumed. Age estimations for the stem lineage and crown lineage 4. Discussion of Fagopyrum were between 19.5 and 46.3 Mya and between 16.99 and 39.4 Mya, respectively. Our phylogenetic analyses unambiguously placed P. tibeticum The aligned cpDNA dataset consisted of 2670 characters; 407 within Fagopyrum (Fig. 2, Appendices 3 and 4). More specifically, variable but parsimony-uninformative, 396 potentially parsi- two major clades were identified under the Fagopyrum as previ- mony-informative and 1867 constant. Phylogenetic analyses by ously found (see Yasui and Ohnishi, 1998a,b; Ohsako and Ohnishi, the MP, ML and Bayesian methods produced completely congruent 2000) and one clade was closely-related to the only species of topologies (Appendix 4). Similarly, the tribe Atraphaxideae was not Parapteropyrum, P. tibeticum (BP = 100; PP = 100). Further examina- monophyletic and Parapteropyrum grouped with one of two Fago- tion of morphological and karyological data corroborated this pyrum clades with high support (BP = 100, PP = 100). We retained molecular inference. Both P. tibeticum and all species of Fagopyrum data for all species when both ITS and cpDNA sequences were share racemous inflorescences, five deeply lobed perianths, eight available. Of the combined 3412 characters, 2076 characters were stamens arranged in two wheels (five outer and three inner) and constant, 592 variable characters were parsimony-uninformative three styles with head-like stigmas (Li, 1998). Further, P. tibeticum and 744 characters were parsimony-informative. These two data- has 24 pairs of chromosomes, and is probably a hexaploid with the sets are largely congruent because the ILD test revealed only a same basic chromosome number (x =8;Tian et al., 2009) found in marginally significant difference (P > 0.01). Phylogenetic analyses all species of Fagopyrum (2n = 16 or 32; Chen, 1999). In contrast, of this combined dataset generated identical topologies with MP, basic chromosome numbers of x = 9 or 11 have been reported for ML and Bayesian methods. As in the separate analyses of each data the other two genera (Atraphaxis or Calligonum) of the Atraphaxi- set, Parapteropyrum nested within Fagopyrum (Fig. 2) when both deae (Mao et al., 1983; Tian et al., 2009). All this evidence suggests datasets were combined. Its sole species comprised a separate that the woody P. tibeticum evolved from and should be taxonom- clade, sister to one Fagopyrum clade, which included six wild sam- ically plated in the herbaceous genus Fagopyrum (=F. tibeticum (A.J. pled species, almost exclusively occurring in the QTP. The other Li) Adr. Sanchez & Jan. Burke) (Sanchez et al., 2011). Our results clade comprised two widely cultivated species (Fagopyrum tatarc- corroborate the hypothesis by Carlquist (2003) that woody X. Tian et al. / Molecular Phylogenetics and Evolution 61 (2011) 515–520 519 lineages of the Polygonaceae might have originated from the her- Our results confirmed a recent study that reduced the taxo- baceous lineages because most species of this family are perenni- nomic rank of Parapteropyrum and considered its sole species to als. In fact, the woody character probably evolved multiple times be a member of the genus Fagopyrum (Sanchez et al., 2011). In in this family as suggested by Lamb-Frye and Kron (2003) and addition to a separate clade consisting of this species, our analyses our reconstruction of the habit evolution along the phylogenetic confirmed two other species clades within this genus (Fig. 2), as re- history (Fig. 3). These finding therefore obviously support the ported by previous studies (Yasui and Ohnishi, 1998a,b; Ohsako new taxonomic treatment of the subfamily Polygonoideae, which and Ohnishi, 2000). Further work is required to confirm whether placed the woody species in the different tribes (Sanchez et al., these three clades should be ranked as different infrageneric sec- 2011). tions. All wild species of this genus have been used as buckwheat For various island plants, the evolution of woodiness has prob- resources for artificial domestication and the creation of new ably contributed strongly to avoidance of ‘overtopping’ by their breeds through hybridization (Ohnishi and Matsuoka, 1996; Li, respective communities, and hence substantial reduction of inter- 1998; Ohnishi, 1998). Like those of the other species (Ohnishi, specific competition (Darwin, 1859). Woodiness may be particu- 1998), achenes of F. tibeticum are also collected and eaten as the larly important for insect-pollinated species colonizing new wild ‘food’ by the local dwellers according to our field surveys. Fur- habitats, in which insects are likely to be initially rare, and for sel- ther examination of the nutritional quality and production of ach- fing species, since it can help to counter inbreeding depression enes in this species is required to determine its domestication arising from founder effects in geographically isolated habitats potential. As a possible woody buckwheat species, it has high po- (Wallace, 1878). Due to its considerable adaptive advantages, insu- tential value, since its cultivation would avoid annual sowing, re- lar woodiness appears to have repeatedly evolved from ancestral duce labor costs and increase habitat protection. herbs in numerous genera and families (e.g. Böhle et al., 1996; Givnish, 1998; Panero et al., 1999). Analogously to events follow- Acknowledgments ing the formation of islands, uplift of the QTP resulted in the crea- tion of unoccupied habitats for plants, and (hence) probably We are highly grateful for Prof. Thomas J. Givnish for his invalu- triggered island-like species radiation in numerous groups (Wang able comments on the previous version of this manuscript and et al., 2005, 2009; Liu et al., 2006). It also probably led to the devel- excellent suggestions for the explanation of the woodiness origin opment of insular woodiness, as indicated by our results, which in the Qinghai–Tibetan Plateau. We also thank Amy Litt and anon- suggest that F. tibeticum originated from the herbaceous genus Fag- ymous reviewers for their constructive and thoughtful comments. opyrum between 6.4 and 14.8 Mya. This estimate should be treated This research was supported by the National Natural Science Foun- cautiously and improved when better calibrations based on fossils dation of China (Grant Numbers 30725004 to Jianquan Liu and are available. However, during this period, geological evidence sug- 30860026 to Jian Luo), the Key Project of International Collabora- gests that the QTP was extensive uplifted, especially between 15– tion Program, the Ministry of Science and Technology of China 13 Mya and 8–7 Mya (Harrison et al., 1992; Li et al., 1995; Shi et al., (2010DFB63500) and the International Collaboration ‘111’ Project. 1998; Spicer et al., 2003). In addition, case studies suggest that radiative evolution of several species-rich genera occurred during this period (Wang et al., 2005, 2009; Liu et al., 2006; Tu et al., Appendix A. Supplementary material 2010; Zhang et al., 2011). Therefore, rapid speciation in small, iso- lated populations in the new niches created by the extensive uplift Supplementary data associated with this article can be found, in probably favored the fixation of unique or rare morphological char- the online version, at doi:10.1016/j.ympev.2011.07.001. acters in the newly evolved species (Kadereit, 1994). Further, the appearance of these characters seems to have misled taxonomists, References resulting in species being elevated to higher taxonomic positions. This may explain why numerous monotypic genera that occur in Ballard, H.E., Sytsma, K.J., 2000. Evolution and biogeography of the woody Hawaiian violets (Viola, Violaceae): arctic origins, herbaceous ancestry and bird dispersal. the QTP nest within other genera represented by other species in Evolution 54, 1521–1532. this region (see, for instance, Liu et al., 2000; Friesen et al., 2000). Bao, B.J., Li, A.R., 1993. A study of the genus Atraphaxis in China and the system of The woodiness of F. tibeticum presumably evolved in a similar Atraphaxideae (Polygonaceae). Acta Phytotax. Sin. 31 (2), 127–139. Böhle, U.R., Hilger, H.H., Martin, W.F., 1996. Island colonization and evolution of the manner. However, it remains unclear why the other tree species insular woody habit in Echium L. (Boraginaceae). Proc. Natl. Acad. Sci. USA 93, did not colonize the same habitats where F. tibeticum grows. As 11740–11745. suggested by Givnish (1998) and Givnish et al. (2009), most tree Carlquist, S., 1970. Hawaii, A Natural History: Geology, Climate, Native Flora and Fauna Above the Shoreline. Pacific Tropical Botanical Garden, Lawai, HI. species have heavy seed and could not establish their initial seed- Carlquist, S., 1974. Island Biology. Columbia Univ. Press, New York. lings without the deep litters in the islands. The dry valleys of Tibet Carlquist, S., 2003. Wood anatomy of Polygonaceae: analysis of a family with may similarly be unfavorable for the heavy seeds of the tree spe- exceptional wood diversity. Bot. J. Linn. Soc. 141, 25–51. Chen, Q.F., 1999. A study of resources of Fagopyrum (Polygonaceae) native to China. cies, but beneficial to the small seeds of the Fagopyrum-like Bot. J. Linn. Soc. 130, 53–64. species. Darwin, C., 1859. The Origin of Species by Means of Natural Selection. John Murray, It seems likely that insular woodiness, driven by habitat isola- London. Doyle, J.J., Doyle, J.L., 1987. A rapid DNA isolation procedure for small quantities of tion and environmental change through the uplift of the QTP, has fresh leaf tissue. Phytochem. Bull. 19, 11–15. also emerged in other woody endemics with herbaceous relatives Drummond, A.J., Ho, S.Y., Rawlence, N., Rambaut, A., 2007. A Rough Guide to BEAST in the region. In addition, our results suggest that some classical 1.4. University of Auckland, Auckland, New Zealand. western–eastern Eurasian disjunctions of plant groups inferred Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1995. Testing significance of incongruence. Cladistics 10, 315–319. from morphological traits require further detailed examination Fay, M.F., Swensen, S.M., Chase, M.W., 1997. Taxonomic affinities of Medusagyne based on independent molecular evidence; for example, Psammos- oppositifolia (Medusagynaceae). Kew Bull. 52, 111–120. ilene W.C. Wu and C.Y. Wu (Caryophyllaceae) and Cotinus Mill. Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791. (Anacardiaceae) (Sun, 2002). These genera may have local origins Friesen, N., Fritsch, R.M., Pollner, S., Blattner, F.R., 2000. Molecular and in eastern Eurasia, rather than being closely related to correspond- morphological evidence for an origin of the aberrant genus Milula within ing groups that occur disjunctly in western Eurasia, although some Himalayan species of Allium (Alliaceae). Mol. Phylogenet. Evol. 17, 209–218. Givnish, T.J., 1998. Adaptive radiation of plants on oceanic islands: classical species are thought to have originated from long-dispersal from patterns, molecular data, new insights. In: Grant, P. (Ed.), Evolution on Islands. that region (Tu et al., 2010). Oxford Univ. Press, Oxford, UK, pp. 281–304. 520 X. Tian et al. / Molecular Phylogenetics and Evolution 61 (2011) 515–520
Givnish, T.J., Kendra, M.C., Mast, A.R., Paterson, T.B., Theim, T.J., Hipp, A.L., Henss, Saiki, R.K., Celfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B., J.M., Smith, J.F., Kenneth, R., Wood, K.R., Sytsma, K.J., 2009. Origin, adaptive Erlich, H.H.A., 1988. Primer-directed enzymatic amplification of DNA with a radiation and diversification of the Hawaiian lobeliads (Asterales: thermostable DNA polymerase. Science 239, 487–491. Campanulaceae). Proc. Roy. Soc. B 276, 407–416. Sanchez, A., Schuster, T.M., Kron, K.A., 2009. A large-scale phylogeny of Guindon, S., Cascuel, O., 2003. A simple, fast, and accurate algorithm to estimate Polygonaceae based on molecular data. Int. J. Plant Sci. 1708, 1044– large phylogenies by maximum likelihood. Syst. Biol. 52, 696–704. 1055. Harrison, T.M., Copeland, P., Kidd, W.S.F., Yin, A., 1992. Raising Tibet. Science 255, Sanchez, A., Schuster, T.M., Burke, J.M., Kron, K.A., 2011. Taxonomy of Polygonoideae 1663–1670. (Polygonaceae): a new tribal classification. Taxon 60, 151–160. Hong, S.P., 1995. Pollen morphology of Parapteropyrum and some putatively related Shi, Y.F., Li, J.J., Li, B.Y., 1998. Uplift and Environmental Changes of Qinghai–Tibetan genera (Polygonaceae–Atraphaxideae). Grana 34, 153–159. Plateau in the Late Cenozoic. Guangdong Science and Technology Press, Hong, S.P., 1998. Systematic significance of tepal surface morphology in tribes Guangzhou. Persicarieae and Polygoneae (Polygonaceae). Bot. J. Linn. Soc. 127, 91–116. Spicer, R.A., Harris, N.B.W., Widdowson, M., Herman, A.B., Guo, S., Valdes, P.J., Wolfe, Huelsenbeck, J.P., Rannala, B., 1997. Phylogenetic methods come of age: testing J.A., Kelley, S.P., 2003. Constant elevation of southern Tibet over the past 15 hypotheses in an evolutionary context. Science 276, 227–232. million years. Nature 421, 622–624. Huelsenbeck, J.P., Ronquist, F., 2001. MrBayes: Bayesian inference of phylogeny. Sun, H., 2002. Tethys retreat and Himalayas–Hengduanshan Mountains uplift and Bioinformatics 17, 754–755. their significance on the origin and development of the Sino-Himalayan Hughes, C., Eastwood, R., 2006. Island radiation on a continental scale: exceptional elements and alpine flora. Acta Bot. Yunnan. 24, 273–288. rates of plant diversification after uplift of the Andes. Proc. Natl. Acad. Sci. USA Swofford, D.L., 2002. paup⁄: Phylogenetic Analyses Using Parsimony (⁄ and Other 103, 10334–10339. Methods), Version 4. Sinauer & Associates, Sunderland, MA. Kadereit, J.W., 1994. Molecules and morphology, phylogenetics and genetics. Bot. Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary Acta 107, 369–373. genetics analysis (MEGA) v755 software version 4.0. Mol. Biol. Evol. 24, 1596– Lamb-Frye, A.S., Kron, K.A., 2003. RbcL phylogeny and character evolution in 1599. Polygonaceae. Syst. Bot. 28 (2), 326–332. Tavakkoli, S., Osaloo, S.K., Maassoumi, A.A., 2008. Morphological cladistic analysis of Li, A.R., 1981. Parapteropyrum A. J. Li—Unum genus novum Polygonacearum sinicum. Calligonum and Pteropyrum (Polygonaceae) in Iran. Iran. J. Biotechnol. 14, 117– Acta Phytotax. Sin. 19, 330–331. 125. Li, A.R., 1998. Flora Republicae Popularis Sinicae. Science Press, Beijing. Tavakkoli, S., Osaloo, S.K., Maassoumi, A.A., 2010. The phylogeny of Calligonum and Li, J.J., Shi, Y.F., Li, B.Y., 1995. Uplift of the Qinghai-Xizang (Tibet) Plateau and Global Pteropyrum (Polygonaceae) based on nuclear ribosomal DNA ITS and chloroplast Change. Lanzhou University Press, Lanzhou. trnL-F sequences. Iran. J. Biotechnol. 8, 7–15. Liedo, M.D., Crespo, M.B., Cameron, K.M., Fay, M.F., Chase, M.W., 1998. Systematics Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. of Plumbaginaceae based upon cladistic analysis of rbcL sequence data. Syst. Clustal-X windows interface: flexible strategies for multiple sequence Bot. 23, 21–29. alignment aided by quality analysis tools. Nucl. Acids Res. 25, 4876–4882. Liu, J.Q., Chen, Z.D., Lu, A.M., 2000. The phylogenetic relationships of the Qinghai– Tian, X.M., Liu, R.R., Tian, B., Liu, J.Q., 2009. Karyological studies of Parapteropyrum Tibet endemic Sinadoxa, revealed by the ITS data. Acta Phytotax. Sin. 42, 656–658. and Atraphaxis (Polygonaceae). Caryologia 62 (4), 261–266. Liu, J.Q., Wang, Y.J., Wang, A.L., Hideaki, O., Abbott, R.J., 2006. Radiation and Tu, T.Y., Volis, S., Dilon, M.O., Sun, H., Wen, J., 2010. Dispersals of Hyoscyameae and diversification within the Ligularia–Cremanthodium–Parasenecio complex Mandragoreae (Solanaceae) from the New World to Eurasia in the early (Asteraceae) triggered by uplift of the Qinghai–Tibetan Plateau. Mol. Miocene and their biogeographic diversification within Eurasia. Mol. Phylogenet. Evol. 38, 31–49. Phylogenet. Evol. 57, 1226–1237. Maddison, W.P., Maddison, D.R. 2004. Mesquite: A Modular System for Evolutionary Wallace, A.R., 1878. Tropical Nature and Other Essays. Macmillan, London. Analysis, Version 1.01.