Figures
Abstract
Meconopsis Viguier (Papaveraceae) is an iconic genus of alpine forbs that includes medicinal and ornamental species. This study extends previous phylogenetic analyses of Meconopsis, using ITS sequences representing all the major Meconopsis clades. Phenotypic traits are also analysed for all described species. Our results show that Meconopsis evolved as a ≥ octaploid clade, with considerable interior structure reflecting further changes in ploidy levels as well as phenotypic differentiation. We support the exclusion of a few species as Cathcartia or Papaver, making Meconopsis a Tibetan region clade. Based on average rates of nucleotide substitution in angiosperm herbs we estimate that the Meconopsis clade diverged from the Meconella clade of Papaver approximately 16.6 Ma. This is soon after the ‘hard’ collision of the Indian continent with Asia caused uplift of the Himalaya and Hengduan ranges, greatly extended the Tibetan plateau, and initiated monsoonal climates. Eight major clades within Meconopsis are well supported and these correspond closely to previously recognised subgenus groups. However, the relationship among the clades is poorly resolved, probably because they diverged rapidly ∼15-11 Ma. Two of these clades are ∼dodecaploid but appear to have originated independently. The eight clades have distinct distributions, variously associated with the Himalaya, the eastern Plateau and Hengduan ranges. Some Meconopsis species were not monophyletic, suggesting that a combination of multilocus molecular and phenotypic traits is required when defining and revising species.
Citation: Xie H, Ash JE, Linde CC, Cunningham S, Nicotra A (2014) Himalayan-Tibetan Plateau Uplift Drives Divergence of Polyploid Poppies: Meconopsis Viguier (Papaveraceae). PLoS ONE 9(6): e99177. https://doi.org/10.1371/journal.pone.0099177
Editor: Edward J. Louis, University of Leicester, United Kingdom
Received: January 19, 2014; Accepted: May 12, 2014; Published: June 16, 2014
Copyright: © 2014 Xie et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: A PhD scholarship and Vice-Chancellor's Travel Grant to HYX from the Australian National University. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Many studies suggest that mountain uplift can drive the evolution of new species including plants [1], [2], [3] fungi [4], invertebrates [5], fish [6], frogs [7], birds [8], [9] and mammals [10], [11], [12], [13]. Tertiary uplift of the Tibetan region [14] created the world's largest plateau and highest mountains (Map S1) so evidence of impacts of uplift upon phylogeny should be strong. Uplift increases the diversity of habitats to which species may adapt, extend their range and speciate. Local uplift generally creates cooler, cloudier and wetter habitat but extensive uplift modifies regional climates. Across the Himalo-Tibetan region, hyper-humid monsoonal forests and alpine herb-fields in the south graduate to desert and steppe at similar altitudes to the north [15]. Mountain ranges and plateau margins become deeply eroded creating steep topography with contrasting elevations and aspects that provide high local habitat diversity [16], [17]. Steep altitudinal gradients and complex topography can buffer the impacts of climate change on populations by reducing the need for long-distance migration [18], [19]. However, mountainous terrain also generates barriers that could foster allopatric speciation [20]. In combination, these impacts of mountain uplift on habitat diversity, distribution and stability should have facilitated divergence and speciation in many clades in the Tibetan region. Indications that this occurred are the very high floristic diversity across the southern Tibetan region and phylogenetic evidence from various plant and animal clades [13], [21], [22], [23], [24], [25].
Polyploidization can facilitate a host of rapid evolutionary changes that enable species to exploit the diverse habitats created by mountain uplift [26], [27]. Polyploidization is one of the most effective mechanisms of generating radically new genomes and of causing abrupt speciation [28], [29], [30], [31]. Hybridization introduces an additional genome that can fuel adaptation and divergence. Though autopolyploidy does not add novel genetic material it can cause immediate adaptive changes in phenotype, such as breakdown of self-incompatibility mechanisms [29], [32], [33], [34], as well as providing duplicate copies of genes that can diverge fuelling gradual adaptation [35], [36], [37], [38].
In this paper we focus on Meconopsis Viguier (Papaveraceae Jussieu) and examine whether its phylogeny is consistent with an uplift driven model of divergence and speciation. We selected Meconopsis, the ‘Tibetan’, ‘Himalayan’ or ‘Blue’ poppies, because the genus is species rich and has been the flagship for the regional flora since the mid-19th century [39], [40], [41], [42]. In 1814 Viguier proposed that the elongate pistil and absence of a protruding stigmatic disc on the capsule of the European Papaver cambrica L. warranted recognizing it as a new monotypic genus, Meconopsis. Since then more than 70 species with similar traits have been classified as Meconopsis [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53]. However, taxonomic revisions [39], [52], [53] indicated that these diagnostic traits are not limited to Meconopsis. Both the single American and European species have been placed in the Papaver clade [52], [54], [55] and one Asian species in Cathcartia [39]. The remaining ∼68 Meconopsis species are endemic to relatively humid alpine or sub-alpine parts of the Tibetan uplift region (Tibet plateau, Himalaya, Hengduan and adjacent ranges: Map 1). Most species are monocarpic perennial rosette forbs but a few can be polycarpic and one forms ramets. Successive morphological sub-generic classifications [52], [53], [56] have each defined about eight species groups, some of which are consistently identified but others are not, and their proposed hierarchical or evolutionary relationships differ because Prain [52] gave priority to hair shape, Taylor [53] to style shape, and Wu and Chuang [56], [57] to the inflorescence (scapose, racemose or paniculate). Recent collecting has resulted in many new species [43], [44], [45], [46], [47], [48], [49], [50], [51] but some are based on minor morphological differences so there is a need for better phylogenetic information.
DNA sequence analyses are providing a more robust understanding of the phylogeny of the Papaveraceae. The family is a monophyletic clade [58] that fossil evidence indicates diverged from other Ranunculales 121-106 Ma (million years ago) [59] or possibly slightly earlier [60]. An analysis of ITS and trnl sequences in selected Papaver-related species [55] placed Asian Meconopsis species as a monophyletic clade that was sister to the Meconella clade, the Arctic Poppies (not to be confused with Meconella Nuttall the American Fairy Poppies). These two clades combined (Papaver 1) are sister to most other Papaver species (Papaver 2). Carolan et al [55] and Yang et al [61] included too few Meconopsis species to adequately define relationships within the genus.
In this paper we developed an ITS phylogeny to investigate how the evolution of Meconopsis species might have been influenced by the uplift of eastern Tibetan Plateau and Hengduan ranges. We use the average rate of nucleotide substitution in herbaceous angiosperms [62] to place the ITS phylogeny onto a time scale. We also present a phylogeny based on phenotypic traits for all described Meconopsis species since this could broaden the analysis to test correlated evolutionary changes in various traits. We map chromosome numbers onto the trees to investigate the possible role of polyploidization in the evolution of Meconopsis. Lastly we explore the evolutionary implications of this in the discussion.
Materials and Methods
DNA extraction, PCR amplification, cloning and sequencing
We used 75 Meconopsis, Papaver and Argemone mexicana (outgroup) ITS sequences lodged on GenBank and sequenced 34 additional specimens representing 7 Meconopsis species that we collected in the eastern Tibetan area (Table S1). The only protected species, M. torquata, was donated from harvests made by the Tibetan Traditional Medicine Pharmaceutical Factory, Lhasa, Tibet Autonomous Region. Meconopsis speciosa was collected from a wild population (N 28.384014, E 98.993400) with permission from Bai Ma Xue Shan Natural Reserve Bureau. The other species are not protected and permits were not required at the sites where they occurred.
One to 2 g of dried leaf sample was lyophilized and homogenized on a FastPrep machine (Thermo Electron Corp., Milford, MA, USA) for DNA extraction. Genomic DNA was extracted using a DNeasy plant mini kit (QIAGEN Inc., Valencia, CA, USA) following the manufacturer's instructions. DNA extracts were quantified and visualized using agarose gel electrophoresis. ITS was amplified using primers ITS1+ITS4 [63] in an Eppendorf Mastercycler. If necessary, PCR bands of interest were excised from the gel and purified using Wizard SV Gel and PCR Clean-up system (Promega, Madison, WI, USA). Purified PCR products from each sample were cloned using the pGEM-T Easy Vector System (Promega). A total of 55 clones representing seven species were sequenced and analyzed. Prior to sequencing, all PCR products were purified using the Promega Wizard SV gel and PCR Clean-up System and ExoSAP-IT (GE Healthcare, Piscataway, New Jersey, USA) according to the manufacturer's instructions. Extension products were purified using an ethanol/EDTA/sodium acetate precipitation protocol according to the BigDye Terminator v3.1 sequencing kit instructions (Applied Biosystems, Foster City, CA, USA). Products were sequenced bi-directionally with ABI PRISM BigDye Terminator v3.1 sequencing kit (Applied Biosystems) on an ABI-3100 automated sequencer. Sequences were edited in Sequencher v 4.7 [64].
Phylogenetic reconstruction and divergence time estimation
We conducted BLAST [65] searches on the consensus sequences of the obtained Meconopsis and related taxa and included the top hits in the phylogenetic analysis. The multiple sequence alignment was conducted with MUSCLE in Geneious Pro v5.6.3 [66] using default settings, and manually optimised. Indels as well as a 12 bp and 17 bp hypervariable region were excluded from the final alignment, similar to Carolan et al. [55].
A maximum-likelihood (ML) tree was estimated using RAxML 7.0.4 [67]. We estimated the phylogeny and chronology of clade divergence using BEAUti v1.7.4, BEAST v1.7.4, Tracer v1.5 and TreeAnnotator v1.7.4 [68] to derive a maximum-credibility (MC) tree from the aligned ITS data. Trees were visualised using FigTree v1.3.1 [69]. For the MC tree we used a GTR nucleotide substitution model with 4 gamma categories, a Yule speciation process, and a lognormal relaxed clock that allows rates to vary independently among lineages. We set priors for the clock using the mean (4.13×10−9 site−1 year−1) and standard deviation (1.47×10−9 site−1 year−1) of ITS nucleotide substitution rates among 10 herbaceous angiosperm clades with calibration ages >0.5 Ma [62]. We used default settings for other priors and ran the Markov chain Monte Carlo algorithm for 2.5×107 generations, discarded the initial 10% as ‘pre-equilibrium’, and then sampled every 2500th generation to derive a consensus tree. The effective sample sizes exceeded 1000 for all summary statistics, which exceeds the threshold of 200 that is considered to be adequate [68]. Support for nodes in topologies was assessed using non-parametric bootstrapping (ML tree) and posterior probabilities (MC tree).
Phenotypic traits
We included all 68 well-described Meconopsis species and four Papaver species (Meconella clade) [55] as outgroup to root phenotypic cladistic trees. We also generated a tree that included species placed in the Cathcartia clade by ITS analyses: C. villosa, M. chelidonifolia, M. oliveriana and M. smithiana. Phenotypic traits were chosen that have been used to distinguish species and include measures of root, stem, leaf, inflorescence, flower and fruit (Table S2). We categorized count and size measures into 5–7 classes so that we could represent the variability of each trait within species. Successive size classes had a roughly geometric relationship. This produced 251 binary traits that we scored from live or herbarium specimens, published descriptions and keys [39], [40], [45], [46], [48], [49], [50], [51], [56], [70], [71], [72], [73], [74], [75]. Cladistic analysis was conducted using the heuristic method in Mesquite [76] with SPR pruning and regrafting routine to select the minimum tree length. After 2.85×106 rearrangements, 100 trees of equal length were found, and these were summarized as a ‘majority rules’ consensus tree.
Linear Discrimination Analysis (LDA in R [77]) is generally robust on binary data [78] and we used it to examine how well the clades defined by the ITS analysis could be predicted as groups from the binary phenotypic data. We used a Classification Tree (in R: [79], [80]) to define a minimal set of phenotypic traits that could discriminate all clades.
Results
Phylogenetic reconstruction and divergence time estimation
The ML tree (Fig. 1) reiterates the broad supra-generic phylogeny that Carolan et al. [55] found. The majority of Meconopsis species are within a monophyletic clade (74% bootstrap support), sister to the Arctic poppies (Papaver spp: Meconella clade), which together are sister to the Papaver 2 clade. The European M. cambrica (syn. P. cambrica) is placed in Papaver 2, and Californian Stylomecon heterophylla (syn. M. heterophylla) in Papaver 3, while Tibetan region M. chelidonifolia and M. smithiana (syn. Cathcartia smithiana) are placed in the Cathcartia clade, sister to Argemone (67%) in this analysis. Within the Meconopsis clade there are eight well supported clades on the Tibetan plateau and adjacent ranges (Fig. 2): Purpureae (99%), Horridulae (100%), Aculeatae (97%), Primulinae (89%), Puniceae (99%), Grandes (92%), Himalaicae (99%) and Bellae (M. bella). The Himalaicae comprises the largely Himalayan Eupolychaetia (99%) and Discogyne (100%) clades. Most of these clades correspond to and are named after existing taxonomic groups, however the traditional Aculeatae series [39], [53] corresponds to three distinct clades: Aculeatae (including M. aculeata), Purpureae (purple-flowered species) and Horridulae (including M. horridula). The traditional Simplicifolia series is split between Grandes and Puniceae (including M. punicea) clades. The Himalaicae plus Bellae clade (80%) is identified as sister clade (60%) to all other Meconopsis clades, however other clade nodes are less certain.
Bootstrap support (%) for nodes is shown. Key nodes are labeled with the inferred diploid (2n) number of chromosomes. Clades are highlighted in colour. Specimens of M. integrifolia that we consider could be classified as M. pseudointegrifolia are indicated (p-). The scale bar represents substitutions per nucleotide site.
Branch thickness indicates bootstrap support (%) for nodes (≥75%, ≥50–75%, ≥25%–50%, <25%). Clades are highlighted by colour. The estimated median ages (A) of nodes have approximately lognormal error distributions with 95% confidence intervals of ∼0.5 A to ∼2 A. 95% error bar for key nodes are indicated by grey lines. Specimens of M. integrifolia that we consider could be classified as M. pseudointegrifolia are indicated (p-). Scale represents median age estimate (Ma).
Key nodes (Fig. 1) are labeled with the inferred diploid (2n) number of chromosomes [81]. Ploidy is estimated with a base of x = 7. The crown node of the Papavereae is diploid with 2n = 2x = 14, whereas the crown of Meconopsis is 2n = 8x = 56, the Puniceae clade is 2n = 12x = 84, and the crown of the Grandes clade has 2n<12x = 76–80 chromosomes, which possibly indicates a chromosome loss from an x = 12 genome.
The consensus maximum credibility (MC) tree (Fig. 2) is broadly similar to the ML tree. The major difference is among the poorly supported nodes (<50% ML bootstrap or <50% MC posterior probability) in the Meconopsis clade. Key median crown ages are 18.8 Ma for Papaver 1 and Papaver 2, 16.6 Ma for Meconopsis and Meconella clades, 15.0 Ma for Meconopsis. All the major Meconopsis sub-clades had stem ages >12.5 Ma, and crown ages were: Purpureae 7.2 Ma, Bellae <13.2 Ma, Himalaicae 10.0 Ma, Eupolychaetia 8.2 Ma, Discogyne 5.6 Ma, Horridulae 3.7 Ma, Primulinae 9.2 Ma, Puniceae 4.7 Ma, and Grandes 6.6 Ma.
Phenotypic tree
The consensus phenotypic tree for all Meconopsis species (Fig. 3) had ≥62% for all nodes and ≥99% support for most nodes. The topology is significantly different from the ITS trees. The base of the tree was a polychotomy including the outgroup Meconella clade (Papaver) and two clades of Purpureae species. One Purpureae clade gave rise to all other clades, most of which correspond to named ITS clades. However, M. aculeata (Aculeatae) is sister to M. speciosa (Purpureae) within the Himalaicae, and M. bella (Bellae) is placed as sister to Horridulae. On this tree shared basal traits included a scapose inflorescence, 4–6 petals, lobed or dissected leaves, and the relatively small size of most organs. Derived traits included racemose and paniculate inflorescences, larger organ size and more floral parts. When the paniculate ITS Cathcartia clade was included in the analysis it was placed within the Eupolychaetia clade.
All other nodes had ≥62% and most had ≥99% support. Colours indicate clades identified on ITS trees except grey lines indicate nodes with <50% support. Scale represents branch lengths.
The LDA analysis required exclusion of 2 clades represented by a single species and also exclusion of phenotypic traits unique to a single ITS clade (potentially the best predictors) however from the remainder it correctly assigned all other species to ITS clades. The minimal Classification Tree (Fig. 4) required 9 nodes and 8 traits to define the 9 ITS clades, with M. florindae separated from other Primulinae. Other trees of similar length were found, using different trait combinations.
The tree was found using the Tree Classification routine in R, and is one of several trees of similar length.
Discussion
Our study extends previous phylogenetic studies that included Meconopsis species [55], [61] and confirms that Meconopsis is monophyletic with considerable interior structure reflecting changes in ploidy levels as well as gradual differentiation. Certain differences are clear between the molecular and phenotypic trees, indicating sources of confusion in previous work and suggesting parallel evolution of certain traits. Here we assess the differences between molecular and phenotypic trees, examine the role of polyploidy, and interpret our results into the context of uplift of the Tibetan region. Lastly we discuss problems associated with delineating species and modes of speciation.
Molecular and Phenotypic trees
The ITS phylogeny reveals several contrasts to the trees based on phenotypic data alone. Notably, the ITS phylogeny places the Cathcartia clade outside the Papaver - Meconopsis - Argemone clade whereas the phenotypic data place it within Eupolychaetia (Meconopsis), largely due to analogous large shoot and paniculate inflorescence traits. This example highlights the difference: the phenotypic tree suggests that various traits evolved from a small size and few parts to larger size and more parts or branches but this is not supported by the ITS trees. Nor is the trend supported if phenotypic traits are examined more widely in the Papaveraceae. Though the ITS trees support our choice of Meconella clade species as the outgroup to root the phenotypic tree, the Meconella clade is adapted to cold arctic and northern alpine habitats where small and simple plants prevail.
We infer that the shared ancestors of the Meconella and Meconopsis clades had racemose or paniculate inflorescences and that scapes were independently evolved as adaptations to colder and drier habitats. Scapes of high-arctic Meconella clade species (e.g. P. radicatum) are renowned for heliotropic growth [82] which is absent in Meconopsis. The hypothesis that scapes are a derived trait agrees with Chuang [57], however, none of our trees support her proposed phylogeny of the series within Meconopsis.
Our LDA and Classification Tree analyses (Fig. 4) indicate that the Meconopsis sub-clades defined by ITS sequences are also readily defined by phenotypic traits, giving support to their recognition as taxonomic units.
Abrupt polyploid speciation and the origin of Meconopsis
Changes in ploidy level clearly underlie the diversification of Meconopsis and some of its constituent clades. We infer that the ancestral chromosome number in the Papaver 1 clade was 2n = 2x = 14, since this is found in Papaver 2 and the Meconella clade of Papaver 1 [83], [84]. Cathcartia species also have 2n = 2x = 14, supporting the ITS phylogeny that places them outside Papaver, rather than within Meconopsis. Karyotype studies of Meconopsis species suggest that all are polyploid (2n = 8–12x = 56–84) [81], [85], [86], [87], but see Kumar et al. [88]. We suggest that the Meconopsis clade originated with the polyploid transition to 2n = 2x = 56, possibly via a tetraploid intermediate with 2n = 4x = 28 chromosomes that is now extinct. The ITS trees suggest that independent polyploid transitions subsequently gave rise to the Grandes (2n = 76–80<12x) and Puniceae (2n = 12x = 84) clades, however support for this topology is weak. We could not determine if polyploidy in Meconopsis was due to autoploidy or alloploidy, and we caution that cladistic methods are not designed to determine the phylogeny of hybrids.
Allopolyploids occur in many clades in the Papaveraceae [83], [89], [90] and autopolyploids are prevalent in the Meconella clade [91]. Homoploid hybrids have been induced amongst various related Meconopsis species with similar ploidy level [40], [86] and some natural hybrids are known [61], [92].
Tibetan plateau uplift and the evolution of Meconopsis
Placing the evolution of Meconopsis into a historical context requires accurate dating. There are reliable dates for the tectonic uplift of the Tibetan region [14], [93], [94], [95], [96], [97] and palaeoclimates [93], [94], [95], [98], [99], [100]. Uplift occurred in two stages [14]. About 52 Ma, the collision and accretion of the Indian Tibetan-Himalaya microcontinent to the Asian plate was followed by subduction of Greater Indian Basin oceanic crust (52 to 25 Ma) and Asian crustal shortening and thickening that caused uplift of the southern west-central Tibetan plateau. About 25-20 Ma Indian continental crust first reached the subduction zone, where upper layers were sheared off and accreted, contributing to the Himalaya, while deeper continental crust was subducted. This ongoing ‘hard’ collision caused uplift of the Himalaya, further uplift and major faulting of the plateau and, from 15-10 Ma, extension of the plateau to the north and east. The Hengduan ranges to the southeast were tilted up from 20-13 Ma, driven largely by deep crustal flow from the plateau [16], [96], [101], [102], [103], [104] (Map S1). The atmospheric pressure system that brings summer monsoonal precipitation and dense cloud cover to the Himalaya, Hengduan ranges and eastern Tibetan plateau was initiated by uplift about 20 Ma [100]. At that time global Tibetan-latitude climates were ∼4°C warmer than at present [105] so alpine climates probably only occurred on elevated parts of the plateau and ranges at 5000 to 6500 m altitude.
Due to a lack of fossils and definitive geological events, our phylogenetic divergence dates were based on the long-term average rate of nucleotide substitution in other angiosperm herbs [106]. This approach has been used in similar studies [61]. We found the median molecular clock dates to be consistent with the hard collision of India with the Tibetan region causing uplift that enabled the early evolution of both Papaver 1 and the Meconopsis clade.
We suggest that the Papaver 1 clade originated in mid- to north- central Asia, since Meconopsis is endemic to the Tibetan region and the Meconella clade has a center of taxonomic diversity to the north [55], [91], [107]. The uplift driven expansion of subalpine and alpine habitat in this region following the hard collision (starting 25-20 Ma) slightly precedes the median stem and crown ages (18.8 Ma and 16.6 Ma) for Papaver 1. The Meconopsis and Meconella clades include some of the most cold-tolerant angiosperms [108], [109] and we suggest that cold tolerance was an ancestral trait in Papaver 1 that enabled proliferation in uplifted regions. Although both clades tolerate extreme winter cold, Meconopsis is associated with summer monsoonal climates (that started ∼20 Ma), whereas Meconella is not. We suggest that the Meconella clade adapted to northern high-latitudes as global climates cooled from ∼15 Ma [110] and only gave rise to the extant clade during the late Cenozoic, with glacial episodes facilitating its wide dispersal [105], [111].
The ranges of most Meconopsis sub-clades overlap, and species richness peaks, around the junction of the Himalaya and Hengduan ranges (Map S1 and S2). We propose that Meconopsis originated here (∼16.6-15 Ma) soon after uplift commenced (∼20-13 Ma). The transition to tetraploidy (2n = 4x = 28) or octaploidy (2n = 8x = 56) possibly enabled Meconopsis to invade cold monsoonal environments. Most well-supported Meconopsis sub-clades have similar median stem ages ∼15-12.5 Ma, and poorly resolved ancestral relationships, indicating rapid early divergence. There are both habitat and biogeographic differences amongst the sub-clades. The Himalaicae (Eupolychaetia and Discogyne), Bellae and Primulinae clades spread west along the monsoonal Himalayan arc, with Discogyne adapting to the drier Trans-Himalaya. The ∼dodecaploid Grandes clade also occur in the Himalaya and Hengduan ranges but M. integrifolia and M. pseudointegrifolia have adapted to drier plateau habitats to the north. The Purpureae and dodecaploid Puniceae clades spread across the eastern plateau and Hengduan ranges. The Horridulae clade spread widely into relatively dry and cold habitats: west along the Himalaya and adjacent plateau, along the Hengduan ranges and northeast across the plateau. The Aculeatae clade is restricted to the dry western Himalaya, and morphologically appears to be an extension of the Horridulae clade.
Species variability and delineation
A few species were represented by several specimens in our ITS trees (Fig. 1 & 2), allowing evaluation of species monophyly. Several species in the Purpureae and Horridulae clades were not monophyletic but rather form groups (e.g. M. impedita, M. henrici and M. lancifolia; M. horridula and M. racemosa), indicating either that some specimens were incorrectly identified or that species boundaries require revision. Various Meconopsis species are only distinguished by minor morphological trait differences, apparent as short branch lengths on the phenotypic tree (Fig. 3). Analysis of DNA sequence loci on geographically distributed specimens of these and related species could clarify their phylogeny and taxonomic status. Many of these problematic taxa occur within small areas of the eastern Tibetan plateau and Hengduan ranges and may represent geographic differentiation of widespread species. DNA sequences of ‘M. integrifolia’ (including M. pseudointegrifolia) populations across this region show such local spatial differentiation [61]. From our knowledge of local morphology we assigned these ‘M. integrifolia’ populations to either M. integrifolia or M. pseudointegrifolia [39] and conclude from the phylogenetic trees [61] that these are northern and southern sister clades. A population genetic approach is needed to assess whether these populations represent a single variable species, a cline or separate entities.
We suggest that the complex mosaic of habitats on the eastern Tibetan plateau and northern Hengduan ranges facilitated local population divergence. Minor differences in traits are common amongst these populations of Meconopsis species and we expect similar patterns to emerge in other plant genera in this region. Ongoing uplift in the southern Hengduan ranges is raising peaks into the alpine zone facilitating ‘island’ speciation, e.g. M. delavayi. In contrast, alpine habitat along the 2400 km Himalayan arc forms a nearly continuous but convoluted (Fractal Dimension ∼1.3 [112], [113], [114]) contour belt averaging ∼4.4 km wide but ∼16000 km long (total area ∼1.4×105 km2) (Map S1 and S2). Himalayan Meconopsis species are mostly restricted to short portions of the arc, and speciation has possibly occurred by parapatric divergence. Meconopsis species diversity declines to the west, which receives lower precipitation, and we suggest that the Eupolychaetia and Primulinae clades originated in the Eastern Himalaya. These hypotheses could be tested by multilocus sequence loci and population-level DNA analyses.
Conclusions
We consider the ITS phylogeny provides a robust indication of the evolution of Meconopsis, whereas the phenotypic is confounded by false homologies and convergent evolution. We used the average and standard deviation of the nucleotide substitution rate in herbaceous angiosperms [61], [106] to date nodes, and we found the median dates follow shortly after geological and palaeoclimate events that created suitable habitat for Meconopsis. We support Carolan et al [55] in placing Meconopsis as sister clade to the Arctic Poppies (Meconella clade) and estimate that these diverged about 18.8-16.6 Ma, probably in uplifted areas of central Asia. We consider the transition to tetraploidy or octaploidy (2n = 4x or 8x = 28 or 56) was instrumental in the early evolution of Meconopsis. Subsequently two ∼dodecaploid (2n≈12x = 76–84) clades arose within Meconopsis, both of which are characterized by robust and abundant species. The relationship among clades within Meconopsis is poorly resolved and requires additional sequence information. In contrast, there is strong support for most of clades (Discogyne, Eupolychaetia, Bellae and Primulinae), and they correspond closely to existing taxonomic groups. We found that some species in the Aculeatae, Purpureae and Horridulae clades were not monophyletic and suggest that species boundaries require revision. We noted that these species are associated with the eastern Plateau and Hengduan ranges that provide an extensive pattern of habitat variation that could foster a mosaic of locally divergent populations. The charismatic Meconopsis group provides an exciting opportunity for further investigation of the role of geological events, ploidy and local adaption in the diversification of high elevation lineages – of which we know relatively little. We suggest that similar analyses of other alpine herbarium genera from this region can test the paleao-biogeographic hypothesis that we have proposed.
Supporting Information
Table S1.
Species, voucher specimens codes and GenBank accession numbers for ITS sequences used in the study.
https://doi.org/10.1371/journal.pone.0099177.s001
(DOCX)
Table S2.
Binary phenotypic traits of 68 Meconopsis species and 4 Papaver (Meconella clade) species used in the phenotypic analyses.
https://doi.org/10.1371/journal.pone.0099177.s002
(XLSX)
Map S1.
The Tibetan region and adjoining parts of Indian and Asian tectonic plates, showing the subduction zone and major uplifted topographic regions. The distribution of Meconopsis is indicated, including areas with ≥1, ≥5 and ≥10 species. Both Meconopsis species and sub-clade richness peak at the junction of the Himalaya and Hengduan ranges. Base map from NASA Terra Image.
https://doi.org/10.1371/journal.pone.0099177.s003
(TIF)
Map S2.
Map of the ranges of the nine Meconopsis clades defined by ITS sequences. Ranges are based on herbarium specimens and field observations. Note that species are patchily distributed within these ranges and that sampling is sparse over much of the region. The centre of clade and species diversity is indicated by a star, which could indicate where Meconopsis originated. Most clades have 2n = 56 chromosomes. The Eupolychaetia, Discogyne, Bellae and Primulinae clades are largely associated with the Himalayan and Hengduan ranges at the southern margin of the Tibetan uplift region. The Grandes clade (2n = 76–80) is widely represented on the plateau by M. integrifolia. The Puniceae clade (2n = 84) apparently evolved on the northeast plateau, and M. quintuplinervia has spread further East onto adjacent mountain ranges. The base map from NASA Terra Image.
https://doi.org/10.1371/journal.pone.0099177.s004
(TIF)
Acknowledgments
Tsenang Neema from Dechen was an invaluable assistant during our field surveys. Zhengdong Fang at the Alpine Botanic Garden, Zhongdian provided advice on sampling sites. Dr Rong Li at Kunming Institute of Botany, Chinese Academy of Sciences and Dr Hunkun Zhou at Northwest Institute of Plateau Biology, CAS provided access to herbarium collections. Tibetan Traditional Medicine Pharmaceutical Factory (Lhasa, Tibet Autonomous Region) donated fresh samples of M. torquata to us. Permission from Bai Ma Xue Shan Natural Reserve Bureau provided us the opportunities to work at there. Dr David Morris and Leon Smith provided instruction and assistance with DNA analysis, and Dr Suat Hui Yeoh, Bee Gunn and Chong Ren Ong gave advice on phylogenetic analyses. Dr Yvonne Pittelkow assisted us with the LDA and Classification Tree analyses. Brendan Lepcshi at Australian National Herbarium, Canberra provided advice on nomenclature. We thank the editor and a referee for their comments on the manuscript.
Author Contributions
Conceived and designed the experiments: HYX CCL. Performed the experiments: HYX. Analyzed the data: CCL JEA HYX. Contributed reagents/materials/analysis tools: CCL JEA HYX. Wrote the paper: HYX JEA CCL AN SC. Figure design and drawing: HYX JEA.
References
- 1. Hughes C, Eastwood R (2006) Island radiation on a continental scale: Exceptional rates of plant diversification after uplift of the Andes. Proceedings of the National Academy of Sciences of the United States of America 103: 10334–10339.
- 2. Winkworth RC, Wagstaff SJ, Glenny D, Lockhart PJ (2005) Evolution of the New Zealand mountain flora: Origins, diversification and dispersal. Organisms Diversity & Evolution 5: 237–247.
- 3. Meng LH, Yang R, Abbott RJ, Miehe G, Hu TH, et al. (2007) Mitochondrial and chloroplast phylogeography of Picea crassifolia Kom. (Pinaceae) in the Qinghai-Tibetan Plateau and adjacent highlands. Molecular Ecology 16: 4128–4137.
- 4. Albertin W, Marullo P (2012) Polyploidy in fungi: evolution after whole-genome duplication. Proceedings of the Royal Society B-Biological Sciences 279: 2497–2509.
- 5. Kawakita A, Sota T, Ito M, Ascher JS, Tanaka H, et al. (2004) Phylogeny, historical biogeography, and character evolution in bumble bees (Bombus: Apidae) based on simultaneous analysis of three nuclear gene sequences. Molecular Phylogenetics and Evolution 32: 422–423.
- 6. Peng ZG, Ho SYW, Zhang YG, He SP (2006) Uplift of the Tibetan plateau: Evidence from divergence times of glyptosternoid catfishes. Molecular Phylogenetics and Evolution 39: 568–572.
- 7. Roberts JL, Brown JL, Schulte R, Arizabal W, Summers K (2007) Rapid diversification of colouration among populations of a poison frog isolated on sky peninsulas in the central cordilleras of Peru. Journal of Biogeography 34: 417–426.
- 8. Weir JT, Schluter D (2008) Calibrating the avian molecular clock. Molecular Ecology 17: 2321–2328.
- 9. Qu YH, Ericson PGP, Lei FM, Li SH (2005) Postglacial colonization of the Tibetan plateau inferred from the matrilineal genetic structure of the endemic red-necked snow finch, Pyrgilauda ruficollis. Molecular Ecology 14: 1767–1781.
- 10. Flynn JJ, Nedbal MA, Dragoo JW, Honeycutt RL (2000) Whence the red panda? Molecular Phylogenetics and Evolution 17: 190–199.
- 11. Janis CM (1993) Tertiary mammal evolution in the context of changing climates, vegetation, and tectonic events. Annual Review of Ecology and Systematics 24: 467–500.
- 12. Leonard JA, Vila C, Wayne RK (2005) Legacy lost: genetic variability and population size of extirpated US grey wolves (Canis lupus). Molecular Ecology 14: 9–17.
- 13. Tseng ZJ, Wang X, Slater GJ, Takeuchi GT, Li Q, et al. (2014) Himalayan fossils of the oldest known pantherine establish ancient origin of big cats. Proceedings of the Royal Society B: Biological Sciences 281.
- 14. van Hinsbergen DJJ, Lippert PC, Dupont-Nivet G, McQuarrie N, Doubrovine PV, et al. (2012) Greater India Basin hypothesis and a two-stage Cenozoic collision between India and Asia. Proceedings of the National Academy of Sciences of the United States of America 109: 7659–7664.
- 15. Bookhagen B, Burbank DW (2006) Topography, relief, and TRMM derived rainfall variations along the Himalaya. Geophysical Research Letters 33.
- 16. Burbank DW, Blythe AE, Putkonen J, Pratt-Sitaula B, Gabet E, et al. (2003) Decoupling of erosion and precipitation in the Himalayas. Nature 426: 652–655.
- 17.
Salick J, Moseley RK (2012) Khawa Karpo: Tibetan traditional knowledge and biodiversity conservation. St Louis (MO), USA: Missouri Botanical Garden Press,. 273 p.
- 18. Corlett RT, Westcott DA (2013) Will plant movements keep up with climate change? Trends in Ecology & Evolution 28: 482–488.
- 19. Aitken SN, Yeaman S, Holliday JA, Wang TL, Curtis-McLane S (2008) Adaptation, migration or extirpation: climate change outcomes for tree populations. Evolutionary Applications 1: 95–111.
- 20. Givnish TJ (2010) Ecology of plant speciation. Taxon 59: 1326–1366.
- 21. Yang FS, Li YF, Ding X, Wang XQ (2008) Extensive population expansion of Pedicularis longiflora (Orobanchaceae) on the Qinghai-Tibetan Plateau and its correlation with the Quaternary climate change. Molecular Ecology 17: 5135–5145.
- 22. Wang L, Richard J, zheng W, Chen P, Wang Y, et al. (2009) History and evolution of alpine plants endemic to the Qinghai-Tibetan Plateau: Aconitum gymnandrum (Ranunculaceae). Molecular Ecology 18: 709–721.
- 23. Opgenoorth L, Vendramin GG, Mao KS (2010) Tree endurance on the Tibetan Plateau marks the world's highest known tree line of the Last Glacial Maximum. New Phytologist 185: 332–342.
- 24. Wang B-S, Mao J-F, Gao J, Zhao W, Wang X-R (2011) Colonization of the Tibetan Plateau by the homoploid hybrid pine Pinus densata. Molecular Ecology 20: 3796–3811.
- 25. Wu L-L, Cui X-K, Milne RI, Sun Y-S, Liu J-Q (2010) Multiple autopolyploidizations and range expansion of Allium przewalskianum Regel. (Alliaceae) in the Qinghai-Tibetan Plateau. Molecular Ecology 19: 1691–1704.
- 26. Abbott RJ, Brochmann C (2003) History and evolution of the arctic flora: in the footsteps of Eric Hulten. Molecular Ecology 12: 299–313.
- 27. Levin DA (1983) Polyploidy and novelty in flowering plants. American Naturalist 122: 1–25.
- 28. Soltis DE, Soltis PS (1993) Molecular-data and the dynamic nature of polyploidy. Critical Reviews in Plant Sciences 12: 243–273.
- 29. Otto SP, Whitton J (2000) Polyploid incidence and evolution. Annual Review of Genetics 34: 401–437.
- 30. Bretagnolle F, Felber F, Calame FG, Kupfer P (1998) Polyploidy in plants. Botanica Helvetica 108: 5–37.
- 31. Soltis DE, Soltis PS, Tate JA (2004) Advances in the study of polyploidy since plant speciation. New Phytologist 161: 173–191.
- 32. Miller JS, Venable DL (2000) Polyploidy and the evolution of gender dimorphism in plants. Science 289: 2335–2338.
- 33. Mable BK (2004) Polyploidy and self-compatibility: is there an association? New Phytologist 162: 803–811.
- 34. Husband BC, Ozimec B, Martin SL, Pollock L (2008) Mating consequences of polyploid evolution in flowering plants: Current trends and insights from synthetic polyploids. International Journal of Plant Sciences 169: 195–206.
- 35. Körner C (1999) Alpine plant life: functional plant ecology of high mountain ecosystems.
- 36. Lumaret R, Guillerm JL, Maillet J, Verlaque R (1997) Plant species diversity and polyploidy in islands of natural vegetation isolated in extensive cultivated lands. Biodiversity and Conservation 6: 591–613.
- 37. Thompson JN, Merg KF (2008) Evolution of polyploidy and the diversification of plant-pollinator interactions. Ecology 89: 2197–2206.
- 38. Segraves KA, Thompson JN, Soltis PS, Soltis DE (1999) Multiple origins of polyploidy and the geographic structure of Heuchera grossulariifolia. Molecular Ecology 8: 253–262.
- 39.
Zhang M, Grey-Wilson C (2008) Meconopsis Viguier, Hist. Nat. Pavots Argémones, 48. 1814. In: Committee FoCE, editor. Flora of China. Beijing and Saint Louis: Science Press & Missouri Botanical Garden Press. pp. 262–278.
- 40.
Cobb JLS (2012) Meconopsis World- A visual reference St Andrews University, East Coast of Scotland, U.K.
- 41.
Kessell M, Taylor P, CHristie I (2011) The Meconopsis Group. 1998–2011: The Meconopsis Group, Dunblane, Perthshire, UK.
- 42.
Mueggler E (2011) The Paper Road: Archive and Experience in the Botanical Exploration of West China and Tibet. Berkeley: Univeristy of California Press.
- 43. An LZ, Chen SY, Lian YS (2009) A New Species of Meconopsis (Papaveraceae) from Gansu, China. Novon 19: 286–288.
- 44. Cui ZL (2005) Yongshan (2005) A new variety of Meconopsis from Gansu. Guihaia 25: 106–109.
- 45. Egan PA (2011) Meconopsis autumnalis and M. manasluensis (Papaveraceae), two new species of Himalayan poppy endemic to central Napal with sympatric congeners. PHYTOTAXA 20: 47–56.
- 46. Grey-Wilson C (2006b) A new Meconopsis from Tibet. The Alpine Gardner 74: 22–225.
- 47. Lu M, Lian Y (2006) Meconopsis punicea Maxim. var. glabra M.Z.Lu et Y.Sh. Lian (Papaveraceae), a new variety from Gansu, China. Bulletin of Botanical Research 26: 8–9.
- 48. Ohba H, Yoshida T, Sun H (2009) Two New Species of Meconopsis (Papaveraceae) from Southern Biluo Xueshan, Yunnan, China. The Journal of Japanese Botany 84: 294–302.
- 49. Yoshida T, Sun H, Boufford DE (2010) New Species of Meconopsis (Papaveraceae) from Mianning, Southwestern Sichuan, China. Acta Botanica Yunnanica 32: 503–507.
- 50. Yoshida T, Sun H, Boufford DE (2011) New species of Meconopsis (Papaveraceae) from Balang Shan, Western Sichuan, China. Plant diversity and Resources 33.
- 51. Yoshida T, Sun H, Boufford DE (2012) New species of Meconopsis (Papaveraceae) from Laojun Shan and Yao Shan, Northern Yunnan, China. Plant diversity and Resources 34.
- 52. Prain D (1906) A Review of the genera Meconopsis and Cathcartia. Ann Bot os-20: 323–370.
- 53. Taylor G (1934) An account of the genus Meconopsis: New flora and silva ltd.
- 54. Kadereit JW, Preston CD, Valtuena FJ (2011) Is Welsh Poppy, Meconopsis Cambrica (L.) Vig. (Papaveraceae), truly a Meconopsis? . New Journal of Botany 1: 80–88.
- 55. Carolan JC, Hook ILI, Chase MW, Kadereit JW, Hodkinson TR (2006) Phylogenetics of Papaver and related genera based on DNA sequences from ITS nuclear ribosomal DNA and plastid trnL intron and trnL-F intergenic spacers. Annals of Botany 98: 141–155.
- 56. Wu CY, Chuang H (1980) A study on the taxonomic system of the genus Meconopsis. Acta Bot Yunnanica 2: 371–381.
- 57. Chuang H (1981) The Systematic Evolution and the Geographical Distribution of Meconopsis Vig. Acta Botanica Yunnanica 3: 139–146.
- 58. Wang W, Lu AM, Ren Y, Endress ME, Chen ZD (2009) Phylogeny and classification of Ranunculales: Evidence from four molecular loci and morphological data. Perspectives in Plant Ecology Evolution and Systematics 11: 81–110.
- 59. Anderson CL, Bremer K, Friis EM (2005) Dating phylogenetically basal eudicots using rbcL sequences and multiple fossil reference points. American Journal of Botany 92: 1737–1748.
- 60. Doyle JA (2012) Molecular and Fossil Evidence on the Origin of Angiosperms. Annual Review of Earth and Planetary Sciences 40: 301–326.
- 61. Yang FS, Qin AL, Wang XQ (2012) Great Genetic Differentiation among Populations of Meconopsis integrifolia and Its Implication for Plant Speciation in the Qinghai-Tibetan Plateau Plos One. 7: e37196.
- 62. Kay KM, Whittall JB, Hodges SA (2006) A survey of Nuclear Ribosomal internal transcribed spacer substitution rates across Angiosperms: an approximate molecular clock with life history effects. Bmc Evolutionary Biology 6(36).
- 63.
White TJ, Bruns T, Lee S, Taylor JW (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: PCR Protocols: A Guide to Methods and Applications edited by Innis, M A, D H Gelfand, J J Sninsky, and T J White 315–322.
- 64. Wang H, Qiong L, Sun K, Lu F, Wang Y, et al. (2010) Phylogeographic structure of Hippophae tibetana (Elaeagnaceae) highlights the highest microrefugia and the rapid uplift of the Qinghai-Tibetan Plateau. Molecular Ecology 19: 2964–2979.
- 65. Altschul SF, Gish W, Miller EW, Lipman DJ (1990) Basic local alignment search tool. Journal of Molecular Biology 215: 403–410.
- 66.
Drummond AJ, Ashton B, Buxton S, Cheung M, Cooper A, et al.. (2011) Geneious In: ed v, editor. Auckland, New Zealand.
- 67. Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the Raxml Web-Servers. Systematic Biology 75: 758–771.
- 68. Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian Phylogenetics with Beauti and the Beast 1.7. Molecular Biology and Evolution 29: 1969–1973.
- 69. Rambaut A (2009) FigTree 1.3.1. Institute of Evolutionary Biology, University of Edinburgh
- 70.
Liu S (1996) Meconopsis Vig. Flora of Qinghai. Xining, Qinghai: Qinghai People's Publishing House. pp. 387–392.
- 71.
Wu CY (1999) Papaveraceae Flora of China. Beijing: Science Press. pp. 7–48.
- 72.
Grey-Wilson C (2001) Poppies: a guide to the poppy family in the wild and in clutivation: Timber Press. 256 p.
- 73. Grey-Wilson C (1996) Meconopsis integrifolia, the yellow Poppywort and its allies. New Plantsman 3: 22–39.
- 74. Yoshida T, Sun H, Boufford DE (2007) Meconopsis wilsonii subsp. wilsonii (Papaveraceae) rediscovered. Acta Botanica Yunnanica 29.
- 75. Grey-Wilson C (2006a) The true identity of Meconopsis napaulensis. Curtis's Botanical Magazine 23: 176–209.
- 76.
Maddison WPaDRM (2009) Mesquite: a modular system for evolutionary analysis. Version 272.
- 77.
R. Core Team (2013) R: A language and environment for statistical computing. Vienna, Austria.: R Foundation for Statistical Computing.
- 78. Asparoukhov OK, Krzanowski WJ (2001) A comparison of discriminant prodecures for binary variables. Computational Statistics & Data Analysis 38: 139–160.
- 79.
Breiman L (1984) Classification and regression trees: Chapman & Hall.
- 80.
Ripley BD (1996) Pattern Recognition and Neural Networks. Cambridge, UK: Cambridge University Press.
- 81. Meng Y, Xie HY, Nie ZL, Gu Z, Yang YP (2006) A karyomorphological study on four species of Meconopsis Vig. (Papaveraceae) from the Hengduan Mountains, SW China. Caryologia 59: 1–6.
- 82. Corbett AL, Krannitz PG, Aarssen LW (1992) The influence of petals on reproductive success in the arctic poppy (Papaver radicatum). Canadian Journal of Botany 70: 200–204.
- 83. Kadereit JW (1991) A note on the genomic consequences of regular bivalent formation and continued fertility in triploids. Plant Systematics and Evolution 175: 93–99.
- 84. Kurup S, Ride JP, Jordan N, Fletcher G, Franklin-Tong VE, et al. (1998) Identification and cloning of related self-incompatibility S-genes in Papaver rhoeas and Papaver nudicaule. Sexual Plant Reproduction 11: 192–198.
- 85. Singhal VK, Kumar P (2008) Impact of cytomixis on meiosis, pollen viability and pollen size in wild populations of Himalayan poppy (Meconopsis aculeata Royle). Journal of Biosciences 33: 371–380.
- 86. Norton CR, Qu Y, Straley GB (1986) Self-Fertility and Cross-Compatability among 6 Species of Meconopsis (Papaveraceae). Journal of Horticultural Science 61: 257–263.
- 87.
Xie HY (1999) The ethnobotanical study of Meconopsis Vig. in NW Yunnan. Kunming, Yunnan. 127 p.
- 88. Kumar S, Jeelani S, Rani S, Gupta R, Kumari S (2013) Cytology of five species of subfamily Papaveroideae from the Western Himalayas. Protoplasma 250: 307–316.
- 89. Ojala A, Rousi A, Lewing E, Pyysalo H, Widen CJ (1990) INTERSPECIFIC HYBRIDIZATION IN PAPAVER .3. F1 HYBRIDS BETWEEN SPECIES OF SECT OXYTONA. Hereditas 112: 221–230.
- 90. Lavania UC, Srivastava S (1999) Quantitative delineation of karyotype variation in Papaver as a measure of phylogenetic differentiation and origin. Current Science 77: 429–435.
- 91. Solstad H, Elven R, Nordal I (2003) Isozyme variation among and within North Atlantic species of Papaver sect. Meconella (Papaveraceae) and taxonomic implications. Botanical Journal of the Linnean Society 143: 255–269.
- 92. Yuan C, Li P, Wang Y, Shi S (2004) The confirmation of putative natural hybrid species Meconopsis X cookei G. Taylor (Papaveraceae) Based on Nuclear Ribosomal DNA ITS Region sequence. ACTA Genetica Sinica 31: 901–907.
- 93. Rowley DB, Currie BS (2006) Palaeo-altimetry of the late Eocene to Miocene Lunpola basin, central Tibet. Nature 439: 677–681.
- 94. Saylor JE, Quade J, Dellman DL, DeCelles PG, Kapp PA, et al. (2009) The late miocene through present paleoelevation history of southwestern tibet. American Journal of Science 309: 1–42.
- 95. Ding ZL, Derbyshire E, Yang SL, Yu ZW, Xiong SF, et al. (2002) Stacked 2.6-Ma grain size record from the Chinese loess based on five sections and correlation with the deep-sea δ18O record PALEOCEANOGRAPHY.
- 96. Zhang PZ, Shen Z, Wang M, Gan WJ, Burgmann R, et al. (2004) Continuous deformation of the Tibetan Plateau from global positioning system data. Geology 32: 809–812.
- 97. Wang CS, Zhao XX, Liu ZF, Lippert PC, Graham SA, et al. (2008) Constraints on the early uplift history of the Tibetan Plateau. Proceedings of the National Academy of Sciences of the United States of America 105: 4987–4992.
- 98. Pollard D, DeConto RM (2009) Modelling West Antarctic ice sheet growth and collapse through the past five million years. Nature 458: 329–U389.
- 99. Gupta AK (2008) Evolution of the Indian Monsoon system and Himalayan-Tibetan Plateau uplift during the Neogene. Himalayan Journal of Sciences 5.
- 100. Sun JM, Ye J, Wu WY, Ni XJ, Bi SD, et al. (2010) Late Oligocene-Miocene mid-latitude aridification and wind patterns in the Asian interior. Geology 38: 515–518.
- 101. Schoenbohm LM, Burchfiel BC, Chen LZ (2006) Propagation of surface uplift, lower crustal flow, and Cenozoic tectonics of the southeast margin of the Tibetan Plateau. Geology 34: 813–816.
- 102. Yao HJ, Beghein C, van der Hilst RD (2008) Surface wave array tomography in SE Tibet from ambient seismic noise and two-station analysis - II. Crustal and upper-mantle structure. Geophysical Journal International 173: 205–219.
- 103. Reid AJ, Wilson CJL, Liu S (2005) Structural evidence for the Permo-Triassic tectonic evolution of the Yidun Arc, eastern Tibetan plateau. Journal of Structural Geology 27: 119–137.
- 104.
Ming QZ, Shi ZT. The Forming Factor Analysis for the First Bend of Yangtse River; 2006 Jul 31–Aug 04; Denver, CO. pp. 1587–1590.
- 105. NOAA (2009) World Data Center for Paleoclimatology. National Climatic Data Center, NOAA Satelite and Information service
- 106. Wallace LE (2003) Molecular Evidence for Allopolyploid Speciation and Recurrent Origins in Platanthera huronensis (Orchidaceae). International Journal of Plant Sciences 164: 907–916.
- 107. Kadereit JW, Schwarzbach AE, Jork KB (1997) The phylogeny of Papaver s l (Papaveraceae): Polyphyly or monophyly? Plant Systematics and Evolution 204: 75–98.
- 108. Billings WD, Mooney HA (1968) Ecology of arctic and alpine plants. Biological Reviews of the Cambridge Philosophical Society 43: 481–&.
- 109. Klimesova J, Dolezal J (2011) Are clonal plants more frequent in cold environments than elsewhere? Plant Ecology & Diversity 4: 373–378.
- 110.
IPCC (2007) Climate change 2007: the physical science basis. Cambridge, United Kingdom and New York, NY, USA.: Intergovernmental Panel on Climate Change. 235–336 p.
- 111. Petit JR, Jouzel J, Raynaud D, Barkov NI, Barnola JM, et al. (1999) Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399: 429–436.
- 112. Burrough PA (1981) Fractal dimensions of landscapes and other environmental data. Nature 294: 240–242.
- 113. Mandelbrot BB (1975) Stochastic-models for earths relief, shape and fractal dimension of coastlines, and number-area rule for islands. Proceedings of the National Academy of Sciences of the United States of America 72: 3825–3828.
- 114. Palmer MW (1992) The coexistence of species in fractal landscapes. American Naturalist 139: 375–397.