Linnaeoideae is a small subfamily of erect or creeping shrubs to small trees in Caprifoliaceae that exhibits a wide disjunct distribution in Eurasia, North America and Mexico. Most taxa of the subfamily occur in eastern Asia and Mexico but the monospecific genus Linnaea has a circumboreal to north temperate distribution. In this study, we conducted phylogenetic and biogeographic analyses for Linnaeoideae and its close relatives based on sequences of the nuclear ribosomal ITS and nine plastid (rbcL, trnS-G, matK, trnL-F, ndhA, trnD-psbM, petB-D, trnL-rpl32 and trnH-psbA) markers. Our results support that Linnaeoideae is monophyletic, consisting of four eastern Asian lineages (Abelia, Diabelia, Dipelta and Kolkwitzia), the Mexican Vesalea, and Linnaea. The Mexican Vesalea was formerly placed in Abelia, but it did not form a clade with the eastern Asian Abelia; instead Vesalea and Linnaea are sisters. The divergence time between the eastern Asian lineages and the Mexican Vesalea plus the Linnaea clade was dated to be 50.86 Ma, with a 95% highest posterior density of 42.8 Ma (middle Eocene) to 60.19 Ma (early Paleocene) using the Bayesian relaxed clock estimation. Reconstructed ancestral areas indicated that the common ancestor of Linnaea plus Vesalea may have been widespread in eastern Asia and Mexico or originated in eastern Asia during the Eocene and likely migrated across continents in the Northern Hemisphere via the North Atlantic Land Bridges or the Bering Land Bridge. The Qinling Mountains of eastern Asia are the modern-day center of diversity of Kolkwitzia-Dipelta-Diabelia clade. The Diabeliaclade became highly diversified in Japan and eastern China. Populations of Diabelia serrata in Japan and eastern China were found to be genetically identical in this study, suggesting a recent disjunction across the East China Sea, following the last glacial event.
Citation: Wang H-F, Landrein S, Dong W-P, Nie Z-L, Kondo K, Funamoto T, et al. (2015) Molecular Phylogeny and Biogeographic Diversification of Linnaeoideae (Caprifoliaceae s. l.) Disjunctly Distributed in Eurasia, North America and Mexico. PLoS ONE 10(3): e0116485. https://doi.org/10.1371/journal.pone.0116485
Academic Editor: Helge Thorsten Lumbsch, Field Museum of Natural History, UNITED STATES
Received: August 1, 2014; Accepted: December 8, 2014; Published: March 10, 2015
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication
Data Availability: Data can be found in the Supporting Information, i.e. S1 Dataset. Data matrix. All sequence data have been deposited in GenBank under accession numbers KP297477-KP297801.
Funding: This study was partly supported by grants from the Ministry of Science and Technology of China (2012AA021602, 2011FY120200, 2012BAC01B05), National Natural Science Foundation of China (31270239, 41201049, 31129001), IICT and the Innovation International Collaborative Team Grant from the Chinese Academy of Sciences. 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.
Intercontinental disjunct distributions of plants have fascinated botanists for centuries. The East Asian—North American plant disjunctions have attracted much attention in the last two decades [1–9]. The disjunct patterns have also been utilized to understand the histories of plant spreading between continents as well as allopatric speciation [5,6,10]. Fossil, molecular and geological data all suggest that the disjunctions between East Asia and North America originated many times in multiple areas throughout the Tertiary [4,5,9,11,12]. However, these studies have primarily focused on distributions restricted to the Northern Hemisphere, north of Mexico . Fewer studies explored the evolution of the intercontinental disjunct pattern involving lineages distributed in Central America, such as in Mexico [14–16]. Linnaeoideae is such an example with taxa distributed in Eurasia and North America including Mexico  (Fig. 1).
Linnaeoideae (Dumortier) P.F. Stevens is a subfamily of Caprifoliaceae s.l. (including Caprifolioideae, Diervilloideae, Dipsacoideae, Linnaeoideae, Morinoideae, Valerianoideae and Zabelia) [17,95]. This re-circumscribed family of Caprifoliaceae s.l. has been shown to be monophyletic in several recent studies [18–22]. The nomenclature of the clade is also subject to debate; we herein use the subfamily rank Linnaeoideae, although some authors used the tribe rank Linnaeeae Dumortier [21,25,28,38,73].
Improvements from successive molecular studies have provided important insights into the phylogeny of Linnaeoideae. (1) Based on rbcL, ITS sequences and morphological data from 45 samples (not including Diabelia, Vesalea and Zabelia), Donoghue et al.  found that Linnaeoideae was monophyletic and more closely related to Dipsacoideae than to Caprifolioideae. (2) Based on ndhF and trnL-F from 26 samples (not including Diabelia, Vesalea and Zabelia), Zhang et al.  supported the classification of Caprifoliaceae s.l. as in Donoghue et al. . (3) Based on ndhF, trnL-F, trnL, matK and rbcL-atpB regions from 30 samples (not including Diabelia, Vesalea and Zabelia), Bell et al. supported that Linnaeoideae was monophyletic. (4) Based on nuclear and plastid sequence data (ITS, trnK, matK, atpB-rbcL and trnL-F) from 51 samples (including all six genera), Jacobs et al.  were able to resolve a greater number of monophyletic subgroups within Linnaeoideae, now including Abelia, Dipelta, Kolkwitzia, Linnaea as well as Vesalea. They also questioned the position of Zabelia (Rehder) Makino and a sister relationship with Morinoideae or Valerianoideae was proposed but with low support . (5) Using five plastid regions (rbcL, ndhF, trnL intron, trnL-F spacer and matK) from 31 samples (including all six genera), Landrein et al.  identified a new monophyletic group designated at the generic level as Diabelia.
Christenhusz  argued for the broadest sense of Linnaea L. to include all members of Linnaeoideae except Zabelia. Major morphological differences within Linnaeoideae as well as a need to keep the number of nomenclatural changes to a minimum would direct to keep the monophyletic clades distinct. The redefined Linnaeoideae here includes six genera: Abelia R.Br., Diabelia Landrein, Dipelta Maxim., Kolkwitzia Graebn., Linnaea Dumortier and Vesalea M. Martens & Galeotti (Table 1). The subfamily is characterized by achenes topped with persistent sepals, three to four locular ovaries with only one or two fertile locules, a distinct epicalyx, and four stamens.
Linnaeoideae exhibits a wide intercontinental disjunct distribution in Eurasia, North America and Mexico (Fig. 1, and Figs. 6–8 of Tang and Li ). Abelia, Dipelta and Kolkwitzia (Fig. 8 of Tang and Li ) are endemic to China except for A. chinensis, which reaches the Ryukyu Islands, Japan. Diabelia is restricted to Japan with one locality found on the eastern coast of China [31,32]. Vesalea is endemic to Mexico. Linnaea has a circumboreal distribution with relic populations found in Japan (e.g. Iwate, Nagano, Yamanashi), Northeast China, Central Asia, the Caucasus, the Alps, the Appalachian and the Rocky Mountains (Fig. 6 of Tang and Li ).
The phylogeny and biogeography of Linnaeoideae remain poorly understood, and the unique disjunct distribution in East Asia and Mexico deserve far more attention. We thus conducted a phylogenetic study on Linnaeoideae using one nuclear marker (ITS) and nine chloroplast markers (rbcL, trnS-G, matK, trnL-F, ndhA, trnD-psbM, petB-D, trnL-rpl32 and trnH-psbA) with a nearly complete taxon sampling, including all genera, and most species, except Dipelta wenxianensis. Our objectives are to: (1) reconstruct the phylogeny of Linnaeoideae; (2) estimate the divergence times of the major lineages of Linnaeoideae using a fossil-calibrated molecular clock method; and (3) explore the evolution of the unique biogeographic distributions in Linnaeoideae, emphasizing on the intercontinental disjunctions in Eurasia, Mexico, and North America.
Materials and Methods
Linnaeoideae are not included in any Eurasian, North American or Mexican official list of threatened plants. No special permits were required for this study. The field studies did not involve endangered or protected species and the specific location of our study was provided in Table 2. Herbarium voucher specimens were deposited in the Institute of Botany, Beijing (PE) and Kew (K). The sequences determined in this study were listed in Table 2 (GenBank accession numbers: KP297477-KP297801). All sequence data have been deposited in GenBank under accession numbers KP297477-KP297801.
The chloroplast fragments were chosen amongst the core DNA barcodes for land plants and also from the most variable plastid regions previously used in Caprifoliaceae . Abelia is the most taxa-rich genus in Linnaeoideae. Rehder  recognized 13 species; Hu  accepted five species; Yang and Landrein  accepted three species and a species complex which includes all species with two sepals. In order to test relationships among taxa of the two-sepal group, we distinguished five names in this publication (A. macrotera, A. myrtilloides, A. engleriana, A. uniflora and A. parvifolia: specimens were selected and identified by S. L. Zhou). A total of 32 accessions of Linnaeoideae representing seven species of Abelia, five species of Vesalea [35,36,74], three species of Dipelta, three subspecies of Linnaea and the only species of Kolkwitzia were collected from China, Finland, Japan, and Mexico; 12 accessions representing ten outgroup genera were also added (Table 2). The voucher information and GenBank accession numbers are given in Table 2.
Based on previous analyses [19,24,40,70], we included Heptacodium, Leycesteria, Lonicera, Symphoricarpos and Triosteum in Caprifolioideae, Morina and Acanthocalyx in Morinoideae, Weigela in Diervilloideae plus unplaced Zabelia as outgroups for this study.
DNA extraction, amplification and sequencing
Total DNA was extracted from silica gel-dried leaf tissue using the modified Cetyltrimethyl Ammonium Bromide (mCTAB) method . Approximately 20 mg of dried plant tissue was used per extraction. DNA fragments were amplified and sequenced using the primers suggested by Olmstead and Palmer  for rbcL, Sun et al.  for matK, Taberlet et al.  for trnL-F, Shaw et al.  for trnS-G, and Sun et al.  for ITS. Primers for ndhA, trnD-psbM, petB-D, trnL-rpl32 and trnH-psbA are from Dong et al. . Each polymerase chain reaction amplification was carried out in a 25 μL volume with the following reagents: Taq polymerase buffer, 10–50 ng total genomic DNA, 2.0 μM MgCl2, 0.4 μM each of both forward and reverse primers, 0.25 μM each dNTP, and 2 units of Taq DNA polymerase (Takara Biotechnology Co., Dalian, China). The thermal cycling conditions were 3 min at 94°C, followed by 35 cycles of 30 s at 94°C, 40 s at 52°C and 1.5 min at 72°C, with a final extension of 10 min at 72°C. The obtained PCR products were purified with PEG8000 and sequenced using ABI Prism BigDye Terminator Cycle Sequencing Kits v. 3.1 on an ABI 3730xl DNA Analyzer (Life Technologies, 5791 Van Allen Way, Carlsbad, California 92008, USA) following the manufacturer’s instructions.
The sequences were edited and assembled using Sequencher v. 4.7 (Gene Codes Corporation, Ann Arbor, Michigan, USA). The resulting sequences were combined with those downloaded from GenBank, aligned using Clustal W implemented in Mega version 6.0 software  and manually adjusted using Se-Al 2.0 . Prior to concatenating the dataset of each marker, incongruence length difference (ILD) tests were performed on all ten datasets. The datasets were finally concatenated using SequenceMatrix .
Phylogenetic analyses were performed using PAUP* v4b10  for maximum parsimony (MP), RAxML  for maximum likelihood (ML) analyses, and MrBayes 3.2.2  for Bayesian inference (BI). The MP analyses used heuristic searches with 1,000 random addition sequence replicates, tree bisection reconnection (TBR) branch swapping, and MULTREES on. All character states were treated as unordered and equally weighted with gaps treated as missing data. To evaluate the relative robustness of clades in the MP trees, the bootstrap analysis  was performed with 1 000 replicates using the same options as above except that a maximum of 100 trees were saved per replicate.
MrModeltest 3.7  was run for each of the data sets to determine a model of sequence evolution. The models chosen under the Akaike information criterion (AIC) were used in the ML and BI analyses  (see the last row of Table 3). For the ML analyses, ten independent runs were conducted using automatic termination following 20 000 generations without a significant (lnL increase of 0.01) topology change. To estimate the support for each node, 1 000 bootstrap replicates were performed with automatic termination at 10 000 generations, All final runs were performed on the CIPRS Science Gateway (http://www.phylo.org/portal2/) .
A partitioned Bayesian analysis of the plastid dataset was also implemented by applying the previously determined models to each data partition . For BI 40 million generations were run with four chains, each starting with a random tree. Trees were sampled every 1 000 generations. Posterior probabilities (PP) were calculated from the majority consensus of all the sampled trees. When the standard deviation of the split frequencies (SDSF) permanently fell below 0.01, the trees sampled during the burn-in phase were discarded. All final runs were performed on the CIPRS Science Gateway (http://www.phylo.org/portal2/) .
Estimation of divergence times
Seven Abelia-like fruit fossils were reported from Late Oligocene to Middle Eocene . The most reliable character to distinguish Abelia fossils is the shape of fruits. In the extant Abelia, the typical fruit is an oblong achene crowned with 2~5 persistent sepals. Crane  thought only A. trialata, A. quadrialata, A. quinquealata and one Abelia sp. were probably correctly determined, while three additional Abelia-like fossils were wrongly identified. Fruits of A. quadrialata and A. trialata have been found to have hypogynous rather than epigynous fruits . Moreover, even if Abelia-like fossils are correctly determined, it is difficult to identify which genera they belong to. Due to this uncertainty, we did not use the Abelia-like fossils. Manchester and Donoghue  also discounted the Abelia-like fossils for their study.
Our tree was calibrated with three points. First, seeds of Weigela are known from the Miocene and Pliocene in Poland , the Miocene in Mammoth Mountain, Eastern Russia, the Oligocene and Miocene of Western Siberia [102,103], and the Miocene in Denmark . Therefore, we set lognormal prior of the divergence between Weigela and its sister Diervilla at 23 Ma with mean = 0, SD = 1.0, offset = 23 Ma.
Second, Manchester and Donoghue  described the fossil genus Diplodipelta from the late Eocene Florissant flora of Colorado (36–35 Ma), and from the Ruby, and Mormon Creek floras of Montana. The infructescence is made of two achene-like fruits of similar size enclosed by three bracts; two are wing-like, fused to the peduncle at base and the third one is hypothesized to be folded transversely and enveloping the two achenes. In order to interpret the Diplodipelta fossils we present two theoretical morphologies that could correspond to extant inflorescences in Linnaeoideae . (1) The paired achenes could be of similar size (maturing simultaneously) and the infructescence similar to the extant genus Diabelia (Fig. 2A-C). (2) The paired achenes could be of different size (maturing consecutively) and the infructescence similar to the extant genus Kolkwitzia (Fig. 2D-F).
A-C, hypothetical reconstruction of infructescence with achenes of similar size. A, infructescence; B, one pair of fused prophyll-1 removed; C, cross section. D-F, hypothetical reconstruction of infructescence with achenes of different sizes. D, infructescence; E, one pair of fused prophyll-1 removed; F, cross section. G-I, hypothetical reconstruction of Dipelta europaea fossil infructescence. G, infructescence; H, one bract removed; I. cross section. b: bract; b1: bracteole-1; b2: bracteole-2.
In both cases the epicalyx is theoretically formed of six prophylls of different origins (bracts, bracteole-1 and bracteole-2; see Landrein el al.  for terminology).
As shown by Manchester and Donoghue , it seems that paired achenes of similar sizes are the most plausible morphology. Nevertheless, this configuration does not allow the bracteole-1 (b1) to be larger than the bracts (b) like the situation in extant Dipelta species (Fig. 2G-I), and a fusion of the bracteole-1 in two pairs has to be hypothesized (this fusion can also be observed in extant species of the genus Heptacodium but not in Linnaeoideae). If Diplodipelta is related to the genus Dipelta, it is hard to explain why the bracts have reduced in size whereas the bracteole-1 became wing-like. The fact that the bracts of Diplodipelta are fused at the base to the peduncle could form an explanation. We thus differ from Manchester and Donoghue’s  reconstruction which only shows three bracts; the median wing in Fig. 8 p710 showing both sides is interpreted as a single folded bract which is slightly peltate at base and cordate at apex. Although we have not examined critical specimens of the fossil (e.g., Manchester and Donoghue’s Figs. 8, 10, 11 showing the connection between the front and back sides of the same specimen), our reinterpreted bract configuration of four bracteoles-1 fused in two pairs, slightly peltate on one side and cordate on the other side, seems to resemble extant Linnaeoideae fruit morphology (the bracteole-1 of D. floribunda is often either peltate or cordate).
When taking into account new phylogenetic results and inflorescence ontogenetic data, Diplodipelta infructescences could not be dissociated from Linnaeoideae but the fossil genus could also be sister to Diabelia as well as Dipelta. The stratigraphic record of Diplodipelta, together with the occurrence of genuine Dipelta fruits in the late Eocene of England and Mississippi , indicates that the divergence of these genera occurred during or prior to late Eocene . We therefore consider the split of Diplodipelta, Dipelta and Diabelia fossils at about 36–35 Ma and set the stem of Dipelta with lognormal mean = 0, SD = 1.0, offset = 36 Ma.
Third, Caprifoliaceae is a family within eudicots, the oldest fossils of eudicots were recorded at about 125 Ma with their distinctive tricolpate pollen [75,105–107]. Bell and Donoghue  suggested that the Dipsacales originated by the mid-Cretaceous, well before previous age estimates for eudicots. They estimated the Dipsacales node to be 102–110 Ma. In this study the Dipsacales node (the root of our tree) was constrained to 103 Ma, with a normal prior, mean = 103 Ma, SD = 5, despite the lack of fossil evidence.
The estimation of divergence times was obtained using a Yule process speciation prior and an uncorrelated lognormal (UCLN) model of rate change with a relaxed clock . The analyses were run for 30 million generations with parameters sampled every 1 000 generations. Trace files were loaded into Tracer v.1.5  to look for an effective sampling size (ESS), and to examine the posterior distributions of all parameters and their associated statistics including 95% highest posterior density (HPD) intervals. Initially to optimize efficiency in BEAST, we undertook several trial runs of 10–20 million generations and analyzed the results using Tracer v.1.5 . These results were then used to determine the number of generations necessary to achieve the desired ESS of at least 200 and to optimize the operator settings for our abovementioned final analysis. The program Tree Annotator v. 1.8.0  was used to summarize the set of post burn-in trees and their parameters (burn-in set to 4 000), to produce a maximum clade credibility (MCC) chronogram showing mean divergence time estimates with 95% HPD intervals. FigTree v.1.3.1  was used for visualization of the resulting divergence times.
As with many other genera endemic to China , the species diversity of Linnaeoideae is high in western China along the Qinghai-Tibetan plateau and low in eastern China. Five areas of endemism were defined to cover the distribution ranges of Linnaeoideae and its relatives as follows (Fig. 1) based on the distribution of taxa in the subfamily: (A), eastern and northern Asia: Japan, eastern China (Anhui, Fujian, Guangdong, Henan, Hunan, Hubei, Jiangsu, Jiangxi, Shanxi, Taiwan, and Zhejiang provinces), Korea, the Far East of Russia, and the adjacent regions; (B), central and western China (Chongqing, Gansu, Guangxi, Guizhou, Shaanxi, Sichuan, and Yunnan provinces); (C), Europe; (D), North America; and (E), Mexico.
The distribution of each species was assigned to at least one of these regions. The ancestral distributions were inferred using a likelihood approach under the dispersal-extinction-cladogenesis (DEC) model implemented in Lagrange . Python scripts were generated using the online Lagrange configurator (http://www.reelab.net/lagrange/configurator). The MCC tree from BEAST analysis was used as the input tree. The probability of dispersal between areas was modeled as equal, and all values in the dispersal constraint matrix were set to 1.
Many recent studies have incorporated fossils into biogeographic reconstruction [e.g., 63–67]. The ancestral distribution was thus optimized with Lagrange using the phylogeny of the extant species inferred from the combined data set with and without reliable Dipelta and Diplodipelta fossil taxa.
- (i). The fossil of Dipelta europaea was found in southern England in the late Eocene to early Oligocene . We incorporated the age 32.8 Ma as the time of occurrence of D. europaea in Europe into the BEAST tree.
- (ii). The fossil Dipelta sp. was reported from the Eocene of Mississippi . The age 33.5 Ma indicating the occurrence of Dipelta sp. in North America was incorporated into the BEAST tree.
- (iii). Diplodipelta fossil was described by Manchester and Donoghue  from the late Eocene Florissant flora of Colorado. The age 37 Ma was incorporated to indicate the position of Diplodipelta as sister to the Diabelia—Dipelta clade in the BEAST tree.
Given that the fossil of Diplodipelta may be sister to Diabelia or Dipelta, we made two alternative estimations: (1) sister relationship between Diplodipelta and Dipelta in the BEAST tree, and (2) sister relationship between Diplodipelta and Diabelia in the BEAST tree.
Sequence variability within Linnaeoideae
Length of aligned matrices, number of constant characters and potentially parsimony-informative characters, as well as consistency and retention indices of the nine chloroplast regions and the nuclear ribosomal ITS are summarized in Table 3. Of the nine chloroplast regions, trnL-rpl32 is the most variable fragment with a π (nucleotide diversity) value of 0.020 35, while psbM-trnH is the least variable fragment with a π value of 0.005 4. The concatenated length of the nine chloroplast regions reached 7641 bp with 230 parsimony-informative characters. ITS is also very variable in Linnaeoideae with π = 0.019 2 and 40 parsimony-informative characters.
The concatenated plastid markers resolved the tree topologies well at generic level while polytomies existed on all single marker trees. The tree topologies based on different markers were similar. One exception is the tree based on trnL-F which showed sistership between Zabelia and Morina + Cryptothladia, while other markers suggested a more basal position of Zabelia. ILD tests showed that ITS is incongruent with the nine plastid markers at a significant level (p = 0.01). Thus, we only concatenated all nine plastid markers to build better resolved phylogenetic trees using MP, ML and BI methods (Fig. 3, S1 Fig., and S1 Dataset).
MP (first) and ML (middle) bootstrap branch support and Bayesian posterior probabilities (last) are indicated above a cut-off value of 50 and 0.5, respectively.-indicates bootstrap value < 50%.
The monophyly of Linnaeoideae is strongly supported (PB = 100, LB = 100, PP = 1; Fig. 3). Within the subfamily the monophyly of each of the narrowly circumscribed genera is also strongly supported. Zabelia is shown to be a sister group to Morinoideae (PB = 100, LB = 100, PP = 1; Fig. 3) in agreement with previous study from Jacobs et al. . The resolution within Mexican species of Vesalea and Chinese Abelia species is low.
Divergence times of major lineages
The inferred divergence times of Linnaeoideae and its lower ranks are shown in Fig. 4. The crown group of Linnaeoideae was estimated at 50.86 (95% HPD 43.39–63.23) Ma from the Paleocene to early Eocene. Almost all genera of Linnaeoideae had diverged in the Eocene, but the divergences of extant species are inferred to have occurred mostly in the Miocene and Pliocene. According to our estimates, Linnaea split from Vesalea at around 41.03 (95% HPD 24.07–55.19) Ma in the middle Eocene, and Kolkwitzia split from Dipelta at 40.18 (95% HPD 36.85–44.68) Ma. The crown group of Abelia was dated at 23.76 (95% HPD 12.99–35.66) Ma, and the divergence of Abelia chinensis with other Abelia species was dated at 19.47 (95% HPD 10.63–29.94) Ma. The crown groups of Vesalea, Dipelta and Linnaea were estimated at 11.04 (95% HPD 4.23–19.95) Ma, 9.92 (95% HPD 1.83–20.85) Ma and 7.42 (95% HPD 2.23–14.07) Ma, respectively (Fig. 4).
Biogeography of Linnaeoideae
Given Diplodipelta might be the common ancestor of Dipelta or Diabelia, we placed Diplodipelta as sister to the Dipelta—Diabelia clade. In this scenario, our Lagrange analyses reconstructed the ancestral area of the Linnaeoideae in western China as well as central and western China plus Mexico (BE|B with 0.25 relative probability, Fig. 5). Without incorporating fossils, our Lagrange analysis reconstructed the ancestral area of Linnaeoideae in western China as well as central and western China plus Mexico (BE|B with 0.52 relative probability, S2 Fig.).
However, we could not exclude Diplodipelta from being at the lower nodes of the BEAST tree. Therefore, Diplodipelta was placed alternatively as sister to Dipelta or Diabelia. When placed with Dipelta, our Lagrange analysis reconstructed the ancestral area of the Linnaeoideae in western China as well as central and western China plus Mexico (BE|B with 0.30 relative probability, S3 Fig.); when placed instead with Diabelia, the Lagrange analysis reconstructed the ancestral area of the Linnaeoideae in western China as well as central and western China plus Mexico (BE|B with 0.32 relative probability, S4 Fig.).
The tree was based on a 50% majority-rule BI consensus tree. For the Lagrange results, a slash indicates the split of areas into two daughter lineages, i.e., left/right, where “up” and “down” are the ranges inherited by each descendant branch. The values in brackets represent relative probabilities.
Phylogenetic relationships in Linnaeoideae
Rehder  divided taxa now recognized in Linnaeoideae into two major groups, sect. Abelia and sect. Zabelia. Abelia sect. Zabelia was later raised to a generic level in 1948 by Makino . The generic name Zabelia was however, rarely used in later publications and recently Yang and Landrein  used it for the Flora of China. Recent molecular studies have shown that Zabelia is closer to Morinoideae or Valerianoideae than to Linnaeoideae . This study also shows Zabelia is sister to the rest of Linnaeoideae (PB = 100, LB = 100, PP = 1; Fig. 3).
When Zabelia is excluded from the analyses, Abelia s.l. is still polyphyletic [25,38]. The removal of the species belonging to Vesalea M. Martens and Galeotti (= Abelia ser. Vesalea Zabel) and Diabelia Landrein (= Abelia ser. Serrata Graebner) left Abelia as a monophyletic genus.
Six monophyletic groups: Abelia, Diabelia, Dipelta, Kolkwitzia, Linnaea, and Vesalea are well supported in this study and the recently established genera Diabelia and Vesalea are also supported [38,69] (Fig. 3). Diabelia and Vesalea were formerly treated as part of the genus Abelia [23,73]. The strong morphological similarities among Abelia, Diabelia and Vesalea are probably due to convergent evolution, a common phenomenon among the mesic forest elements of the eastern Asian—North American disjunct plants [5, 6, 13]. Some distinct characters can nevertheless be observed concerning inflorescence architecture, corolla shape, and nectaries (Table 4). Abelia, Diabelia and Vesalea share three synapomorphies: accrescent calyx, reduced episepals (except in V. floribunda var. foliacea) and ovaries flattened dorsi-ventrally. Kolkwitzia and Dipelta also have ovaries flattened dorsi-ventrally but they have a well developed epicalyx and reduced calyx but this is clearly linked to dispersal mechanisms (Table 4).
The clade formed by the circumboreal Linnaea and the Mexican Vesalea seems surprising at first because Linnaea has very distinct morphological features like the creeping habit, paired flowers and specialized epicalyx bracts. Nevertheless Vesalea and Linnaea share several morphological synapomorphies (Table 4): (1) raceme-like inflorescences with few flowers, forming on short shoots and appearing in the spring; (2) nectary which is not forming a bulge at the base of the corolla tube but a zone of dense glandular hairs in between the abaxial filaments, and (3) similar creeping habit between Linnaea and Vesalea floribunda except in dry conditions.
Divergent characters could be explained by an adaptation to their different environments. The two genera vicariously occupy the Rocky Mountains and the Sierra Madre Oriental, with extant populations being only separated by a few hundred kilometers.
All six genera have distinct morphological characters that have been described in detail by some taxonomists [71,72]. However, there are species problems within Abelia and Vesalea. The 5-sepaled species of Abelia (A. chinensis and A. forrestii, with the former occupying eastern and southern China, and the latter restricted to a small area in northwestern Yunnan and southwestern Sichuan) are well resolved, but the 2-sepaled species (A. macrotera, A. myrtilloides, A. engleriana, A. uniflora and A. parvifolia) are poorly delimited. Yang and Landrein  treated the latter group as the A. uniflora species complex. This study suggests that most of the 2-sepaled taxa of the genus are very closely related. Similarly to the 2-sepaled Abelia species, the species in Vesalea, which are 5-sepaled, are poorly resolved. Five species, V. floribunda, V. coriacea, V. grandifolia, V. mexicana and V. occidentalis, have been considered to occur in Mexico . Many species within Vesalea are also difficult to separate morphologically.
Linnaeoideae is a group of shrubby or small tree species producing achenes; these fruits are adapted for wind dispersal in the genera Abelia, Diabelia, Dipelta and Vesalea. Achenes of Diabelia serrata and Abelia uniflora only possess two sepals instead of five. In Dipelta the wings originate from the epicalyx and two large bracts are present allowing for wind dispersal. In Kolkwitzia the spiny achenes are surrounded by corky episepals  and the fruits are called Hedge-Hog in Chinese, suggesting possible animal dispersal by clinging to animal fur. Finally in Linnaea the calyx is deciduous in fruit but two of the episepals are covered by large and numerous sticky glandular hairs. Fruit dispersal adaptations, number of sepals and episepals do not seem good indicators of systematic relationships and this could be due to convergence.
Biogeography of Linnaeoideae
Dipelta and Diplodipelta distribution: The crown group for Diabelia, Dipelta and Kolkwitzia was dated at 40.18 (95% HPD 36.85–44.68) Ma in the middle Eocene (Fig. 4). The Lagrange analyses inferred that Dipelta originated in central and western China (B) in the Eocene (Fig. 5). As discussed previously, the Linnaeoideae fruits and Dipelta in particular are adapted to wind dispersal; though it is not known how far they can travel, long distance dispersal events cannot be ascertained. In Kolkwitzia the achenes are possibly carried away in animal fur but long distance dispersal has not been tested. Starting in the Miocene, there was a distinct climatic cooling period across the high-latitude areas of the Northern Hemisphere, which may have resulted in a reduction of the distribution of forests . Dipelta once reached southern England as well as the Mississippi region in the late Eocene, as evidenced from the fossil species D. europaea and D. sp., respectively [58,80]. The genus is now restricted to central and western China (B). Evidence for an early North American origin can be inferred from the fossil genus Diplodipelta with its Dipelta-like infructescences that existed in the late Eocene of western North America . Diplodipelta may represent a sister group of Diabelia and Dipelta. Dipelta is inferred to have been more broadly distributed in the Miocene (although no Miocene fossil occurrences are known), with occurrences in Europe and North America, although the genus is restricted to western China today. The Tertiary disjunct distribution of Dipelta between Europe and North America may be explained by extinctions in large parts of its former ranges. Extinction events could have extirpated the old stem relatives that diverged prior to the extant crown radiation, leaving a phylogeny that includes only extant taxa with long stems and species-rich crowns . A remarkably long “temporal gap” occurs between the Dipelta stem and the beginning of the extant radiation in the early Miocene (Fig. 3).
The lack of fossils from paleobotanically rich deposits of Asia might mean that the ancestral area was not in Asia, but in Europe and/or North America. The lack of DNA from extirpated populations of Europe and North America may have given a false impression that the area of modern diversity is the area of origin. Diplodipelta would also possibly have occurred in the lower node of the BEAST tree, therefore, in this study, we discuss all possible scenarios on the likely phylogenetic position of Diplodipelta.
Reconstruction of ancestral areas with Lagrange including fossils (i.e., Diplodipelta in three different positions of the BEAST tree) and without the fossils showed the same ancestral area for Linnaeoideae (c.f. Fig. 5, S2 Fig., S3 Fig., and S4 Fig.), which suggested an ancestral distribution and early diversification of Linnaeoideae in central and western China as well as central and western China plus Mexico, and subsequent dispersal into eastern Asia, Europe as well as into North America and Mexico.
The incorporation of fossils had little impact on the ancestral area of Linnaeoideae in this study (c.f. Fig. 5, S2 Fig., S3 Fig., and S4 Fig.). This may be due to the fact that these fossils are deeply nested within a clade which is now only found in central and western China, while the Vesalea plus Linnaea clade did not incorporate any fossil. This resulted in the same ancestral area for the four scenarios (Fig. 5, S2 Fig., S3 Fig., and S4 Fig.). Nevertheless, all Lagrange analyses had a comparatively low probability (less than 0.60) and did not clearly show the origin place for the subfamily (BE|B). A broader phylogenetic framework is also needed for Linnaeoideae and its close relatives.
The North Atlantic Land Bridges (NALB)  and the Bering Land Bridge (BLB)  have been hypothesized to have played important roles for the spreading of many intercontinental disjunct taxa of the Northern Hemisphere in the Tertiary [5,9,11–13]. NALB existed from the late Cretaceous to early Tertiary, which is an important migration channel for thermophilic plants in the Northern Hemisphere [11,79]. Similarly, BLB provided a stepping-stone migration route for high-latitude distributed (69–75°N) temperate plants from the Eocene to the present except for several temperature decreasing periods [5,12]. The NALB existed from the late Cretaceous to early Tertiary [12,79], and our dating and biogeographic results as well as the fossil records are consistent with a hypothesis of the migration of the Diplodipelta-Dipelta-Kolkwitzia-Diabelia clade from Eurasia to North America via NALB.
The circumboreal distribution of Linnaea
Linnaea borealis is divided into three subspecies, subsp. borealis in Europe, Asia and Alaska; subsp. longiflora (Torr.) Piper & Beattie along the Pacific coast of western North America from Alaska to California, and subsp. americana (J. Forbes) Hultén in the rest of Canada and USA as well as Greenland [33,34]. The wide disjunct distribution of the monotypic genus Linnaea most likely represents an example of migration from Eurasia to North America via Beringia. Linnaea borealis is the only Linnaeoideae showing a continuous intercontinental extant distribution with populations along the Bering Strait islands, Chukotka and the Alaska Peninsula. Linnaea borealis is clearly the most cold-resistant species in Linnaeoideae and could have survived the conditions in the Bering Land Bridge area in the late Tertiary. Smith  concluded that the Caprifolieae clade originated within Asia and migrated around the Northern Hemisphere during the Cenozoic, including several migrations through the BLB. The Beringian route was also reported as a possible hypothesis for the disjunction between the East Asian Weigela middendorffiana and the North American Diervilla .
Vesalea and Linnaea
Our BEAST and biogeographic analyses suggest that the Linnaea-Vesalea clade originated in central and western China (B) and Mexico (E) at 41.03 (95% HPD 24.07–55.19) Ma in the middle Eocene (Fig. 4). The Lagrange analysis supports dispersal from central and western China (B) to Mexico (E) as the explanation of the intercontinental disjunction between Linnaea-Vesalea and the rest of Linnaeoideae.
Vesalea and Abelia are both thermophilic genera, but Linnaea is well adapted to cold conditions. Two alternative hypotheses regarding the migration of the clade are consistent with our results. Unfortunately the lack of well preserved fossil for Vesalea as well as Abelia does not allow us to strongly favor one or the other.
First, since the genera of Linnaeoideae originated in the Eocene, migration through BLB seems likely. Tiffney and Manchester  argued that BLB may be too cold for the thermophilic plants in the late Tertiary. The Pleistocene glaciations disrupted gene flow and drove thermophilic species southward widening their genetic divergences. A migration of the Linnaea-Vesalea common ancestor through the BLB and subsequent radiation of Vesalea in Mexico cannot be excluded, as it allows both conditions to be met (cold resistant and thermophilic).
Second, the NALB might be a more likely route for the migration of the Mexican Vesalea or its common ancestor. Vesalea species were presumably more commonly distributed than its present range prior to the Pleistocene glaciations, and its perennial growth habit may have allowed it to survive the subsequent millennia locally within this former range in various high-elevation or otherwise cool and moist habitats in the highlands, like a few other Northern Hemisphere disjunct plants, such as Aralia L. , Liquidambar L. , Platanus L. , and Toxicodendron Mill. . Many thermophilic disjunct plants of the Northern Hemisphere have been attributed to fragmentation of a once continuous belt of mixed mesophytic broadleaf-evergreen vegetation, i.e., the boreotropical flora [12,13,93] in the Northern Hemisphere. Remnants of the boreotropical floristic elements occur today in East Asia and eastern North America. The lineages that once grew in other areas became extinct by the late Eocene period due to a combination of climatic and geologic changes [5,13,41,93].
Diabelia diversification in the Sino-Japanese Floristic Region
Diabelia is a widespread genus in Japan and is only recorded in one locality of Zhejiang province of East China [31,32]. The Sino-Japanese Floristic Region (SJRF) is a major region of plant diversity mostly composed of temperate deciduous forest in eastern China, Korea and Japan . Kolkwitzia, Dipelta and Diabelia form a well-supported clade (81/99/1) (Fig. 3); their stem group was dated at 45.79 (95% HPD 39.04–53.52) Ma, and their crown group at 40.53 Ma in the Middle Eocene. The genera Kolkwitzia and Dipelta are endemic Chinese floristic elements and most diverse in the Qinling Mountains. Diabelia is a component of the Sino-Japanese floristic region and most diverse in Eastern China and Japan. About 63.8% of the genera of the Qinling range also occur in Japan . Kolkwitzia, Dipelta and Diabelia could therefore represent a typical element of this flora which has also shown a slow decline and extinction of the genus Diabelia in East China. Diabelia serrata (collected in Zhejiang) and the same species collected in Japan could not be differentiated with the molecular markers we employed (100/100/1). It is therefore likely that Diabelia in China represents a relatively recent refugium following the last glacial event.
During glaciations in the Quaternary, the East China Sea level was lowered and a continuous belt of forests connected the now disjunct populations of East China, South Japan and Korea . The East China Sea Land Bridge [90, 91] may have allowed dispersal and gene exchange between woodland species of East China, Korea and Japan. Examples such as Cercidiphyllum japonicum  and Kalopanax septemlobus  have been documented. However the East China Sea Land Bridge may also have acted as a ‘filter’ during the last glacial event for certain species and produced genetic differentiation among populations in South Japan, South Korea and East China. This is the case of many rare species such as Platycrater arguta  and Kirengeshoma palmata . Despite those results we were not able to reproduce this scenario with the species Diabelia spathulata also growing in Zhejiang. Further phylogeographic studies using population genetic data may allow discovering whether a possible admixture, isolation or ‘filter’ event occurred in this species.
The following new combinations are made in light of the phylogenetic results;
Vesalea occidentalis (Villarreal) H.F. Wang & Landrein, comb. nov.
Basionym,—Abelia occidentalis Villarreal, Brittonia 49 (1), 84. 1997
Holotype,—Mexico. DURANGO, Mpio. Suchil, Reserva de la Michilia, Cienega Los Caballos, Villarreal-Quintanilla, J.A. (with Carranza, N.A.) 8180 (MEXU).
Vesalea grandifolia (Villarreal) H.F. Wang & Landrein, comb. nov.
Basionym,—Abelia grandifolia Villarreal, Brittonia 52(2), 174. 2000
Holotype,—Mexico. QUERÉTARO, Mpio. de Jalpan, Cerro Grande, 13 June 1991, Servin, B. 1101 (CAS).
Vesalea mexicana (Villarreal) H.F. Wang & Landrein, comb. nov.
Basionym,—Abelia mexicana Villarreal, Brittonia 52 (2), 172. 2000
Holotype,—Mexico. OAXACA, Mpio. San Sebastian Tecomaxtlahuaca, Calzada, J.I. 21100 (MEXU).
Vesalea coriacea Hemsl. var. subcoriacea (Villarreal) H.F. Wang & Landrein, comb. nov.
Basionym,—Abelia coriacea Hemsl. var. subcoriacea Villarreal, Acta Bot. Mex. 102, 115, 2013
Holotype,—Mexico. COAHUILA, Mpio. Sierra Mojada, Sierra Mojada, Near Esmeralda, above San Salvador Mine, Stewart, R.M. 1081 (MEXU).
S1 Dataset. Data matrix.
The aligned sequence data as presented in nexus format.
S1 Fig. Bayesian tree of Linnaeoideae and outgroups based on the internal transcribed spacer (ITS) sequence data.
S2 Fig. Ancestral area reconstruction of Linnaeoideae using Lagrange without fossil information.
S3 Fig. Ancestral area reconstruction of Linnaeoideae using Lagrange including three fossils, in which Diplodipelta was placed as sister to Dipelta.
We thank Helge Thorsten Lumbsch, Steven Manchester and an anonymous reviewer for their constructive suggestions; Zhi-Xin Zhu, Ning Zhang, Wei-Tao Jin and Chao Xu for their help during laboratory work; Rico Arce, Maria de Lourdes; Soccoro Gonzales; Lizeth Ruacho; Arturo de Nova; Pedro Castillo Lara; Allen Coombes; Maricela Rodriguez Acosta; Lucio Caamano Onofre; Rodolfo Solano; Miguel Cervantes-Reyes; J. A. Villarreal-Quintanilla and Holly Forbes for helping collecting Mexican materials.
Conceived and designed the experiments: SLZ JW. Performed the experiments: HFW WPD. Analyzed the data: HFW ZLN JW. Contributed reagents/materials/analysis tools: SLZ SL KK TF JW. Wrote the paper: HFW SL SLZ JW.
- 1. Hong DY (1993) Eastern Asian—North American disjunctions and their biological significance. Cathaya 5: 1–39.
- 2. Fritsch P (1996) Isozyme analysis of intercontinental disjuncts within Styrax (Styracaceae) implications for the Madrean-Tethyan hypothesis. Am J Bot 83: 342–355.
- 3. Wen J, Jansen RK, Zimmer EA (1996) Phylogenetic relationships and DNA sequence divergence of eastern Asian and eastern North American disjunct plants. In: Nei M, Takahata N (ed). Current Topics on Molecular Evolution. University Park and Hayama: Pennsylvania State University Press and the Graduate University for Advanced Studies, Hayama University, pp. 37–44.
- 4. Wen J (1998) Evolution of the eastern Asian and eastern North American disjunct pattern: insights from phylogenetic studies. Korean J Pl Taxon 28: 63–81.
- 5. Wen J (1999) Evolution of eastern Asian and eastern North American disjunct distributions in flowering plants. Ann Rev Eco Syst 30: 421–455.
- 6. Wen J (2001) Evolution of eastern Asian—eastern North American biogeographic disjunctions: a few additional issues. Int J Pl Sci 162: S117–S122.
- 7. Milne RI, Abbott RJ (2002) The origin and evolution of tertiary relict floras. Adv Bot Res 38: 282–314.
- 8. Donoghue MJ, Smith SA (2004) Patterns in the assembly of temperate forests around the Northern Hemisphere. Philos Trans Royal Soc B Biol Sci 359: 1633–1644.
- 9. Xiang QY, Zhang H, Ricklefs RE, Qian H, Chen Z, et al. (2004) Regional differences in rates of plant speciation and molecular evolution: a comparison between eastern Asia and eastern North America. Evolution 58: 2175–2184. pmid:15568220
- 10. Boufford DE, Spongberg SA (1983) Eastern Asian—eastern North American phytogeographical relationships—a history from the time of Linnaeus to the twentieth century. Ann Missouri Bot Gard 70: 423–439.
- 11. Tiffney BH (1985a) The Eocene North Atlantic Land Bridge: its importance in Tertiary and modern phytogeography of the Northern Hemisphere. J Arnold Arbor 66: 243–273.
- 12. Tiffney BH (1985b) Perspectives on the origin of the floristic similarity between Eastern Asia and eastern North America. J Arnold Arbor 66: 73–94.
- 13. Wen J, Ickert-Bond S, Nie Z, Li R (2010) Timing and modes of evolution of eastern Asian—North American biogeographic disjunctions in seed plants. In: Long M, Gu H, Zhou Z (eds). Darwin’s heritage today—Proceedings of the Darwin 200 Beijing International Conference. Beijing: Higher Education Press, pp. 252–269.
- 14. Wu ZY (1991) The areal-types of Chinese genera of seed plants. Acta Botanica Yunnanica Supp 4:1–139. (In Chinese with English abstract)
- 15. Feng Y, Oh SH, Manos PS, Zomlefer WB (2005) Phylogeny and historical biogeography of the genus Platanus as inferred from nuclear and chloroplast DNA. Syst Bot 30: 786.
- 16. Huang WP, Sun H, Deng T, Razafimandimbison SG, Nie Z, et al. (2013) Molecular phylogenetics and biogeography of the eastern Asian—eastern North American disjunct Mitchella and its close relative Damnacanthus (Rubiaceae, Mitchelleae). Bot J Linn Soc 171: 395–412.
- 17. APG III (2009) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Bot J Linn Soc 161: 105–121.
- 18. Bell CD, Edwards EJ, Kim AT, Donoghue MJ (2001) Dipsacales phylogeny based on chloroplast DNA sequences. Harv Pap Bot 6: 481–499.
- 19. Donoghue MJ, Eriksson T, Reeves PA, Olmstead RG (2001) Phylogeny and phylogenetic taxonomy of Dipsacales with special reference to Sinadoxa and Tetradoxa (Adoxaceae). Harv Pap Bot 6: 459–479.
- 20. Zhang WH, Chen ZD, Li JH, Chen HB, Tang YC (2003) Phylogeny of the Dipsacales sl based on plastid trnL-F and ndhF sequences. Mol Phylogenet Evol 26:176–189. pmid:12565029
- 21. Pyck N, Smets E (2004) On the systematic position of Triplostegia (Dipsacales): a combined molecular and morphological approach. Belgian J Bot 137(2): 125–139.
- 22. Winkworth RC, Bell CD, Donoghue MJ (2008) Mitochondrial sequence data and Dipsacales phylogeny: mixed models, partitioned Bayesian analyses and model selection. Mol Phylogenet Evol 46: 830–843. pmid:18255318
- 23. Jacobs B, Pyck N, Smets E (2010) Phylogeny of the Linnaea clade: Are Abelia and Zabelia closely related? Mol Phylogenet Evol 57: 741–752. pmid:20727975
- 24. Zhang ZY, Zhou ZK, Gu ZJ (2002) Karyomorphology of Heptacodium (Caprifoliaceae s.str.) and its phylogenetic implications. Taxon 51(3): 499–505.
- 25. Landrein S, Prenner G (2013) Unequal twins? Inflorescence evolution in the twinflower tribe Linnaeeae (Caprifoliaceae s.l.). Int J Pl Sci 174: 200–233.
- 26. Rehder A (1911) Caprifoliaceae. In: Sargent CS, editor. Plantae Wilsonianae. Vol. 1. Cambridge: Cambridge University Press. Pp. 118–129.
- 27. Ikuse M, Kurosawa S (1954) Notes on sect. Zabelia Rehder of the genus Abelia. J Jap Bot 29(4): 11.
- 28. Yang Q, Landrein S (2011) Caprifoliaceae In: Wu Z, Hong D, Raven PH (eds.) Flora of China, vol 19. Beijing and St. Louis: Science Press and Missouri Botanical Garden Press. pp: 617–618.
- 29. Christenhusz M (2013) Twins are not alone: a recircumscription of Linnaea (Caprifoliaceae). Phytotaxa 125: 25–32.
- 30. Tang YC, Li LQ (1994) The phytogeography of Caprifoliaceae s. str. with its implications for understanding eastern Asiatic flora. Acta Phytotax Sin 32: 197–218.
- 31. Zhou SL, Funamota T, Wen J (2004) Species relationships in Abelia sect. Abelia (Caprifoliaceae) in eastern Asia and verification of distribution of A. serrata in China: evidence from AFLP analysis. Acta Bot Yunnan 26: 405–412.
- 32. Zhou SL, Funamota T, Huang PH, Wen J (2006) Discovery of Abelia spathulata (Caprifoliaceae) in Eastern China. Acta Phytotax Sin 4: 467–470.
- 33. Hultén E (1968) Flora of Alaska and neighboring territories. Stanford: Stanford Uni Press.
- 34. Hultén E (1986) Atlas of North European vascular plants. North of the Tropic of Cancer, vol 2. Königstein: Koeltz Scientific Books.
- 35. Villarreal-Q JA (1997) A new species of Abelia (Caprifoliaceae) from western Mexico. Brittonia 52: 172–176.
- 36. Villarreal-Q JA, Rosa-I MDL. (2000) Two new species of Abelia (Caprifoliaceae) from Mexico. Brittonia 52: 172–176.
- 37. Hu JQ (1988) Caprifoliaceae (Linnaeeae excluding Heptacodium and Symphoricarpos). In: Hsu P-S, editor. Flora Reipublicae Popularis Sinicae, vol. 72. Beijing: Science Press. pp. 112–131.
- 38. Landrein S, Gerhard P, Chase M, Clarkson JJ (2012) Abelia and relatives: phylogenetics of Linnaeeae (Dipsacales—Caprifoliaceae s.l.) and a new interpretation of their inflorescence morphology. Bot J Linn Soc 169: 692–713.
- 39. Dong WP, Liu J, Yu J, Wang L, Zhou SL (2012) Highly variable chloroplast markers for evaluating plant phylogeny at low taxonomic levels and for DNA barcoding. PLoS ONE 7(4): e35071. pmid:22511980
- 40. Howarth DG, Donoghue MJ (2009) Duplications and expression of DIVARICATA-like genes in Dipsacales. Mol Biol Evol 26: 1245–1258. pmid:19289599
- 41. Li J, Wang S, Yu J, Wang L, Zhou S (2013) A modified CTAB protocol for plant DNA extraction. Chin Bull Bot 48:72–78. (In Chinese with English abstract)
- 42. Olmstead RG, Palmer JD (1994) Chloroplast DNA systematics: a review of methods and data analysis. Am J Bot 81: 1205–1224.
- 43. Sun H, McLewin W, Fay M (2001) Molecular phylogeny of Helleborus (Ranunculaceae) with an emphasis on the East Asia—Mediterranean disjunction. Taxon 50:1001–1018.
- 44. Taberlet P, Pautou LG, Bouvet J (1991) Universal primers for amplification of three non-coding regions of chloroplast DNA. Pl Mol Biol 17:1105–1109.
- 45. Shaw J, Beck JT, Farmer SB, Liu W, et al. (2005) The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Am J Bot 92: 142–166. pmid:21652394
- 46. Sun Y, Skinner DZ, Liang GH, Hulbert SH (1994) Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacer of nuclear ribosomal DNA. Theor Appl Genet 89: 26–32. pmid:24177765
- 47. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S (2013) MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0. Mol Biol Evol 30: 2725–2729. pmid:24132122
- 48. Rambaut A (1996) Se-Al: sequence alignment editor, Version 2.0 a11. Available: http://treebioedacuk/software/seal/.
- 49. Vaidya G, Lohman DJ, Meier R (2011) Sequence matrix concatenation software for the fast assembly of multi-gene datasets with character set and codon information. Cladistics 27: 171–180.
- 50. Swofford DL (2003) PAUP* Phylogenetic Analysis Using Parsimony (* and Other Methods), version 4. Sunderland: Sinauer Associates.
- 51. Stamatakis A (2006) RAxML-VI-HPC maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690. pmid:16928733
- 52. Ronquist F, Huelsenbeck JP (2003) MrBayes version 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19(12): 1572–1574. pmid:12912839
- 53. Miller MA, Pfeiffer W, Schwartz T (2010) Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In: Proceedings of the Gateway Computing Environments Workshop (GCE). New Orleans. pp 1–8.
- 54. Crane RP (1986) Abelia-like fruits from the Palaeogene of Scotland and North America. Tert Res 9: 21–30.
- 55. Crave PR (1998) The phylogenetic position and fossil history of the Magnoliaceae. In: Hunt D, editor. Magnolias and their allies. Milborne Port: David Hunt. pp: 21–36.
- 56. Manchester SR, Hably L (1997) Revision of ‘Abelia’ fruits from the Paleogene of Hungary, Czech Republic and England. Rev Palaeob Palyn 96: 321–240.
- 57. Manchester SR, Donoghue MJ (1995) Winged fruits of Linnaeeae (Caprifoliaceae) in the Tertiary of Western North America, Diplodipelta gen nov. Int J Pl Sci 156: 709–722.
- 58. Manchester SR, Chen ZD, Lu AM, Uemura K (2009) Eastern Asian endemic seed plant genera and their paleogeographic history throughout the Northern Hemisphere. J Syst Evol 47: 1–41.
- 59. Drummond AJ, Ho SYW, Phillips MJ, Rambaut A (2006) Relaxed phylogenetics and dating with confidence. PLoS Biol 4: 699–710.
- 60. Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol Biol 7: 214. pmid:17996036
- 61. Rambaut A (2008) FigTree v1.1.1. Tree Fig Drawing Tool <http://wwwtreebioedacuk/software/figtree/>
- 62. Fang JY (2002) Vegetation-climate relationship and its application in the division of vegetation zones. Acta Bot Sinica 44(9): 1105–1122.
- 63. Tiffney BH, Manchester SR (2001) The use of geological and paleontological evidence in evaluating plant phylogeographic hypotheses in the Northern Hemisphere Tertiary. Int J Pl Sci 162: S3–S17.
- 64. Kearney M, Clark JM (2003) Problems due to missing data in phylogenetic analyses including fossils: a critical review. J Vertebr Paleontol 23: 263–274.
- 65. Xie L, Yi TS, Li R, Li DZ, Wen J (2010) Evolution and biogeographic diversification of the witch-hazel genus (Hamamelis L., Hamamelidaceae) in the Northern Hemisphere. Mol Phylogenet Evol 56(2): 675–689. pmid:20171295
- 66. Zhang JB, Li RQ, Xiang XG, Manchester SR, Lin L, et al. (2013) Integrated fossil and molecular data reveal the biogeographic diversification of the eastern Asian—eastern North American disjunct hickory genus (Carya Nutt.). PLoS ONE 8(7): e70449. pmid:23875028
- 67. Lu L, Wang W, Chen ZD, Wen J (2013) Phylogeny of the non-monophyletic Cayratia Juss (Vitaceae) and implications for character evolution and biogeography. Mol Phylogenet Evol 68: 502–515. pmid:23669013
- 68. Ree RH, Smith SA (2008) Maximum likelihood inference of geographic range evolution by dispersal, local extinction and cladogenesis. Syst Biol 57: 4–14. pmid:18253896
- 69. Landrein S (2010) Diabelia: a new genus of tribe Linnaeeae subtribe Linnaeinae (Caprifoliaceae). Phytotaxa 3: 34–38.
- 70. Soltis DE, Smith SA, Cellinese N, Wurdack KJ, Tank DC, et al. (2011) Angiosperm phylogeny: 17 genes 640 taxa. Am J Bot 98: 704–730. pmid:21613169
- 71. Fukuoka N (1968) Phylogeny of the tribe Linnaeeae. Acta Phytotax Geobot. 23: 82–94.
- 72. Hara H (1983) A revision of Caprifoliaceae of Japan with reference to all allied plants in other district and the Adoxaceae. Ginkgoana 5: 102–133.
- 73. Kim T (1998) Phylogenetic studies of tribe Linnaeeae (Caprifoliaceae). PhD Thesis, Chonbuk National University.
- 74. Villarreal-Q JA, Villaseñr-Ríos JL, Estrada-Castillón E (2013) Sistematic del genero Abelia (Caprifoliaceae) en Mexico. Acta Bot Mex 102: 99–128.
- 75. Bremer K, Friis EM, Bremer B (2004) Molecular phylogenetic dating of asterid flowering plants shows early Cretaceous diversification. Syst Biol 53: 496–505. pmid:15503676
- 76. Bell CD, Donoghue MJ (2005) Dating the diversification of Dipsacales: comparing models, genes, and evolutionary implications. Am J Bot 92: 284–296. pmid:21652406
- 77. McKenna MC (1983) Cenozoic paleogeography of North Atlantic Land Bridges. In: Bott MHP, Saxov S, Talwani M, Thiede J (eds). Structure and Development of the Greenland-Scotland Ridge: New Methods and Concepts. New York: Plenum Press. pp. 351–395.
- 78. Hopkins DM (1967) The Bering land bridge. Stanford: Stanford University Press.
- 79. Morley RJ (2003) Interplate dispersal paths for megathermal angiosperms. Persp Pl Ecol Evol Syst 6: 5–20.
- 80. Reid E, Chandler MEJ (1926) Catalogue of Cainozoic Plants in the Department of Geology. London: printed by order of the Trustees, British Museum (Natural History), Dept of Geology.
- 81. Smith SA (2009) Taking into account phylogenetic and divergence-time uncertainty in a parametric biogeographical analysis of the Northern Hemisphere plant clade Caprifolieae. J Biogeogr 36: 2324–2337.
- 82. Kim YD, Kim SH (1999) Phylogeny of Weigela and Diervilla (Caprifoliaceae) based on nuclear rDNA ITS sequences: biogeographic and taxonomic implications. J Plant Res 112: 331–341.
- 83. Wen J (2011) Systematics and biogeography of Aralia L. (Araliaceae): revision of Aralia sects. Aralia, Humiles, Nanae and Sciadodendron. Contr U S Nat Herb 57: 1–172.
- 84. Ickert-Bond SM, Wen J (2006) Phylogeny and biogeography of Altingiaceae: evidence from combined analysis of five non-coding chloroplast regions. Mol Phylogenet Evol 39: 512–528. pmid:16439163
- 85. Nie ZL, Sun H, Meng Y, Wen J (2009) Phylogenetic analysis of Toxicodendron (Anacardiaceae) and its biogeographic implications on the evolution of north temperate and tropical intercontinental disjunctions. J Syst Evol 47: 416–430.
- 86. Wu ZY, Wu SG (1996) A proposal for a new floristic kingdom (realm)—the E. Asiatic kingdom: its delimitation and characteristics. In: Zhang AL, Wu SG, editors. Proceedings of the First International Symposium on Floristic Characteristics and Diversity of East Asian Plants. Beijing: Springer-Verlag. pp. 3–42.
- 87. Harrison SP, Yu G, Takahara H, Prentice IC (2001) Palaeovegetation (Communications arising): diversity of temperate plants in east Asia. Nature 413: 129–130. pmid:11557970
- 88. Qi XS, Yuan N, Qiu YX (2012) Development of 12 microsatellite markers for Platycrater arguta (Hydrangeaceae) endemic to East Asia. Am J Bot 99: e304–e306. pmid:22837408
- 89. Sakaguchi S, Qiu YX, Liu YH, Qi XS, Kim SH, et al. (2012) Climate oscillation during the Quaternary associated with landscape heterogeneity promoted allopatric lineage divergence of a temperate tree Kalopanax septemlobus (Araliaceae) in East Asia. Mol Ecol 21: 3823–3838. pmid:22646502
- 90. Qi XS, Na Y, Comes HP, Sakaguchi S, Qiu YX (2014) A strong ‘filter’ effect of the East China Sea Land Bridge for East Asia’s temperate plant species: inferences from molecular phylogeography and ecological niche modelling of Platycrater arguta (Hydrangeaceae). BMC Evol Biol 14: 41. pmid:24593236
- 91. Qiu YX, Qi XS, Jin XF, Tao XY, Fu CX, et al. (2009) Population genetic structure, phylogeography, and demographic history of Platycrater arguta (Hydrangeaceae) endemic to East China and South Japan inferred from chloroplast DNA sequence variation. Taxon 58: 1226–1241.
- 92. Boufford DE, Ying TS (1998) Phytogeography of the Qinling Mountains and a Comparison with the Flora and Vegetation of Japan. Tokyo: University Museum, University of Tokyo.
- 93. Wolfe JA (1975) Some aspects of plant geography of the Northern Hemisphere during the late Cretaceous and Tertiary. Ann Mo Bot Gard 62: 264–279.
- 94. Antonelli A, Sanmartín I (2011) Mass extinction, gradual cooling, or rapid radiation? Reconstructing the spatiotemporal evolution of the ancient angiosperm genus Hedyosmum (Chloranthaceae) using empirical and simulated approaches. Syst Biol 60: 596–615. pmid:21856636
- 95. Stevens PF (2013) Angiosperm Phylogeny Website. Version 12, July 2012 http://www.mobot.org/MOBOT/research/APweb. Accessed in July 2014.
- 96. Fukuoka N (1972) Taxonomic study of the Caprifoliaceae. Mem Fac Sci Kyoto Univ Ser Biol 6: 15–58.
- 97. Donoghue MJ (1983) A preliminary analysis of phylogenetic relationships in Viburnum (Caprifoliaceae s.l.). Syst Bot 8: 45–58.
- 98. Takhtajan A (1987) Systema Magnoliophytorum. Leninopoli. MCML XXXVII (in Russian).
- 99. Felsenstein J (1985) Confidence-limits on phylogenies: an approach using the bootstrap. Evol 39: 783–791.
- 100. Nylander JAA (2004) MrModeltest, version 2. Uppsala: Evolutionary Biology Centre, Uppsala University.
- 101. Lańcucka-Środoniowa M (1967) Two new genera: Hemiptelea Planch. and Weigela Thumb. in the younger Tertiary of Poland. Acta Palaeobotanica 8: 1–17.
- 102. Dorofeev PI (1957) Contributions to the knowledge of western Siberian Tertiary flora (fossil flora of S. Yekaterina near Mt. Tara). A. N. Krishtofovich memorial volume. Moscow-Leningrad: Akademia Nauk SSSR. 277–312. (In Russian).
- 103. Dorofeev PI (1963) Tertiary floras of western Siberia. Moscow-Leningrad: Akademia Nauk SSSR. (In Russian)
- 104. Friis EM (1985) Angiosperm fruits and seeds from the Middle Miocene of Jutland (Denmark). Biologiske Skrifter 24 (3):1–165.
- 105. Doyle AJ, Donoghue MJ (1993) Phylogenies and angiosperm diversification. Paleobiology 19: 141–167.
- 106. Magallon S, Crane PR, Herendeen PS (1999) Phylogenetic pattern, diversity, and diversification of eudicots. Ann Missouri Bot Garden 86: 297–372.
- 107. Sanderson MJ, Doyle AJ (2001) Sources of error and confidence intervals in estimating the age of angiosperms from rbcL and 18S rDNA data. Am J Bot 88: 1499–1516. pmid:21669683
- 108. Posada D, Buckley TR (2004) Model selection and model averaging in phylogenetics: advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Syst Biol 53: 793–808. pmid:15545256
- 109. Brown JM, Lemmon AR (2007) The importance of data partitioning and the utility of Bayes factors in Bayesian phylogenetics. Syst Biol 56: 643–655. pmid:17661232