© 2005 Society of Systematic Biologists
Phylogeny and Biogeography of Exacum (Gentianaceae): A Disjunctive Distribution in the Indian Ocean Basin Resulted from Long Distance Dispersal and Extensive Radiation
Edited by Roberta Mason-Gamer: Assiciate Editor
1 Laboratory of Evolutionary Botany, Institute of Botany, University of Neuchâtel Emile-Argand 11, CH-2007, Neuchâtel, Switzerland yong-ming.yuan{at}unine.ch
2 Royal Botanic Garden Edinburgh Edinburgh EH3 5LR, Scotland, United Kingdom
3 Department of Phanerogamic Botany, Swedish Museum of Natural History S-10405, Stockholm, Sweden
4 South China Botanical Garden, Chinese Academy of Sciences Guangzhou, P.R. China
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
Disjunctive distributions across paleotropical regions in the Indian Ocean Basin (IOB) often invoke dispersal/vicariance debates. Exacum (Gentianaceae, tribe Exaceae) species are spread around the IOB, in Africa, Madagascar, Socotra, the Arabian peninsula, Sri Lanka, India, the Himalayas, mainland Southeast Asia including southern China and Malaysia, and northern Australia. The distribution of this genus was suggested to be a typical example of vicariance resulting from the breakup of the Gondwanan supercontinent. The molecular phylogeny of Exacum is in principle congruent with morphological conclusions and shows a pattern that resembles a vicariance scenario with rapid divergence among lineages, but our molecular dating analysis demonstrates that the radiation is too recent to be associated with the Gondwanan continental breakup. We used our dating analysis to test the results of DIVA and found that the program predicted impossible vicariance events. Ancestral area reconstruction suggests that Exacum originated in Madagascar, and divergence dating suggests its origin was not before the Eocene. The Madagascan progenitor, the most recent common ancestor of Exacum, colonized Sri Lanka and southern India via long-distance dispersals. This colonizer underwent an extensive range expansion and spread to Socotra-Arabia, northern India, and mainland Southeast Asia in the northern IOB when it was warm and humid in these regions. This widespread common ancestor retreated subsequently from most parts of these regions and survived in isolation in Socotra-Arabia, southern India–Sri Lanka, and perhaps mainland Southeast Asia, possibly as a consequence of drastic climatic changes, particularly the spreading drought during the Neogene. Secondary diversification from these surviving centers and Madagascar resulted in the extant main lineages of the genus. The vicariance-like pattern shown by the phylogeny appears to have resulted from long-distance dispersals followed by extensive range expansion and subsequent fragmentation. The extant African species E. oldenlandioides is confirmed to be recently dispersed from Madagascar.
Keywords: Biogeography; DIVA; Exacum; Gentianaceae; ITS; phylogeny; trnL intron
Received January 8, 2003; Revised April 22, 2003; Accepted August 10, 2004
The historical biogeographic connection through the paleotropical regions around the Indian Ocean Basin (IOB) area stands out as a conspicuous pattern and a key topic in the biogeography of plants and animals of the world (Raven and Axelrod, 1974; Raxworthy et al., 2002). This is primarily due to the firmly established and well-documented geologic history of this region, the sequential breakup of the Gondwana landmass, and the differential shifting history of different tectonic plates (McLoughlin, 2001). Thus, biogeographic patterns in this region provide ideal models for testing the long-lasting debate between vicariance and dispersal (Rieppel, 2002). In addition to the Gondwana vicariance and long-distance dispersal explanations, noticeably the earlier land-bridge theory (Steenis, 1962), and the more recent "Lemurian stepping-stones" hypothesis (Schatz, 1996) have been proposed to interpret this biogeographic connection. The distribution of chameleons used to be considered as a typical vicariance example related to Gondwana fragmentation, but recent studies strongly suggest that their present distribution is the result of extensive oceanic dispersals (Raxworthy et al., 2002; Rieppel, 2002). A similar situation applies to certain freshwater fish: cichlids show a typical Gondwana vicariance distribution pattern, which is congruent with the phylogenetic relationships among the lineages. Recent divergence dating, however, revealed significant discrepancy between the times of lineage divergence and the tectonic events (Vences et al., 2001). Therefore, Cenozoic dispersal was considered as a more plausible explanation than Mesozoic vicariance. A few similar debates exist in plants; for example, the oceanic dispersal versus Gondwana vicariance for the biogeography of the genus Adansonia (Baum et al., 1998), the circum-global dispersal versus Gondwana vicariance for the history of the family Melastomataceae (Renner and Meyer, 2001; Renner et al., 2001), migration versus vicariance in Malpighiaceae (Davis et al., 2002a, 2002b), the post-Gondwana long-distance dispersal versus Gondwana vicariance for the historical biogeography of the genus Polyscias (Plunkett et al., 2001), and the widespread-extinction hypothesis versus vicariance or long-distance dispersal theories for the history of the genus Nepenthes (Meimberg et al., 2001). Recently, Sanmartin and Ronquist (2004) also found that most of the plant patterns of suspected Gondwanan origin fit a dispersal rather than a vicariance scenario. As urged by Rieppel (2002), studies on other groups of organisms with a geographical distribution similar to that of chameleons are needed to verify whether some regularity underlies their phylogenetic and biogeographical patterns. The plant genus Exacum (Gentianaceae, tribe Exaceae) appears as a good model for this purpose due to its disjunctive distribution pattern in the IOB.
|
The genus Exacum consists of 64 species and shows a typical paleotropical distribution (see Fig. 1) (Klackenberg, 1985, 2002; Thulin, 2001). Our previous study has addressed its phylogenetic position within the tribe Exaceae (Yuan et al., 2003). Klackenberg (1985)monographed the genus and divided it into two sections based on phylogenetic studies of morphological and anatomical characters: sect. Africana consisting of African, Madagascan, Socotran and southern Arabian species, and sect. Exacum consisting of species of other Asian regions (Sri Lanka, India, through mainland Southeast Asia). The majority of species (38 species) occur in Madagascar. Sri Lanka and the southern tip of the Indian subcontinent (mainly the Western Ghats) is the second most species-rich area (14 species limited to this area, 3 species shared with north India, the Himalayas, and other areas). Three species are found on the southern Arabian peninsula (Dhofar of Oman and nearby Mahrah of Yemen) and the Island of Socotra (Thulin, 2001). The African continent harbors only two species: E. oldenlandioides widespread throughout the entire tropical Africa and E. zombense endemic to the Shire Highlands in Malawi (Klackenberg, 1985). Two species, E. pedunculatum and E. petiolare, are widespread in India. Two species, E. hamiltonii and E. teres, are restricted to the Himalayas, and two species, E. pteranthum and E. sutaepense, are limited to the mountainous regions between Burma and Thailand (referred to as mainland Southeast Asia herein). Only one species, E. tetragonum, is widespread in Indomalesia (India, the Himalayas, mainland Southeast Asia including southern China, and Malaysia) and reaches the extreme north of Australia (Arnhem Land). Based on a phylogeny reconstructed from morphological and anatomical characters, a Gondwana vicariance hypothesis has been suggested to explain the distribution pattern of Exacum, especially for the divergence among the species of Madagascar, India, and Socotra-Arabia (Klackenberg, 1985, 2002). Obviously, this hypothesis requires the assumption of a very old age of the genus with species divergence as early as the Gondwana breakup, in order to link the disjunctions of species distributions to the fragmentation events of the Gondwana landmass. Paleomagnetic data and tectonic reconstruction suggested that the Gondwana breakup initiated ca. 180 million years ago (Mya) (Storey, 1995), and the Madagascar-Seychelles-India block began to separate from the Africa–South America block ca. 165 Mya, with movement ending by 121 Mya (Rabinowitz et al., 1983). Since then, Madagascar has remained in its position with respect to Africa (Coffin and Rabinowitz, 1988), whereas Australia and Antarctica separated from the Madagascar-Seychelles-India block ca. 132 Mya (Barron, 1987). India separated from Madagascar ca. 88 Mya (Storey et al., 1995) and split from the Seychelles ca. 65 Mya (Storey, 1995). Although the Seychelles block subsequently became fixed with respect to Africa, India rapidly drifted northward and collided with Laurasia during the Paleocene-Eocene. Some authors suggested the India-Asia collision occurred ca. 43 Mya (Lee and Lawver, 1995; McLoughlin, 2001), whereas the majority of authors suggest an earlier collision between 50 and 66 Mya (Besse et al., 1984; Patriat and Achache, 1984; Jaeger et al., 1989; Klootwijk et al., 1992; Beck et al., 1995; Valdiya, 2002). Sri Lanka had remained in full contact with India during all these processes until the last major sea level rise 6000 years ago, which separated Sri Lanka from India by the narrow and shallow Palk Strait (McLoughlin, 2001). After separation from Antarctica in the Cretaceous and Paleogene (96 to 35.5 Mya), the Australia–New Guinea block moved northward and collided with Southeast Asia in the late Oligocene to the beginning of the Miocene ca. 25 Mya (Lee and Lawver, 1995; Li and Powell, 2001). Socotra, a continental fragment, was located adjacent to the Dhofar region of southern Oman prior to the Gulf of Aden rifting (Jolivet and Faccenna, 2000). The Afro-Arabian plate collided with Eurasia in the eastern part around 25 Mya (Samuei et al., 1997). At about the same time the Gulf of Aden rift appeared (Richardson et al., 1995; Birse et al., 1997; Ghebreab, 1998; Fantozzi and Sgavetti, 1998), and subsequently the Arabian continent became separated from continental Africa and Socotra ca. 10 Mya (Ghebreab, 1998). Socotra remained in its position relative to the African continent ever since (Horn of Africa) (Richardson et al., 1995; Birse et al., 1997; Samuei et al., 1997).
Our recent investigations on the tribal relationships within Gentianaceae suggested a minimum age of 40 Mya for the tribe Exaceae (Yuan et al., 2003), which apparently argues against a Gondwanan origin of the genus Exacum. Alternative hypotheses to explain the historical biogeography of the genus relevant to post-Gondwana vicariance/dispersal events are therefore needed. To search for such rational explanations, we applied maximum likelihood (ML) and maximum parsimony (MP) methods to reconstruct the phylogeny of Exacum using chloroplast and nuclear ribosomal DNA sequences, and divergence dating and ancestral area reconstruction were conducted based on the molecular phylogeny.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
Ingroup Sampling and Outgroup Choice
To obtain an accurate divergence time estimation for the tribe Exaceae, our previously analyzed family data matrix (Yuan et al. 2003) was expanded by the addition of two trnL intron sequences of Lisianthius, L. longifolius (accession AF102450 [GenBank] ), and L. laxiflorus (accession AF102449 [GenBank] ), retrieved from GeneBank and was reanalyzed with additional calibration points (see below).
Our Exacum data set included 42 sequences of the nuclear ribosomal internal transcribed spacer (ITS) regions, including the spacer 1, the 5.8S gene, and the spacer 2 and 42 chloroplast trnL(UAA) intron sequences (trnL) (Appendix 1). Sampling of Exacum species was maximized to represent all areas of endemism (Africa, Madagascar, Socotra-Arabia, India/Sri Lanka, and mainland Southeast Asia) and morphological diversity. About half the taxa from each geographic area were sampled. The species cover all main morphological variations such as different habits (small herb, large herb, shrub, etc.), different floral types (size of corolla, merosity, etc.), and different types of testa cells (star-shaped, isodiametric, etc.). Thirty-seven accessions representing 30 species of Exacum were sampled, equivalent to about 47% of the genus. Two species (out of three, all endemic to Madagascar) of Ornichia, the sister genus of Exacum (Yuan et al., 2003), were also sampled as ingroups. An additional basal clade sister to the Exacum–Ornichia lineage (Yuan et al., 2003) was sampled as outgroup. This additional clade, composed of the genera Tachiadenus (2 species sampled out of 11, all endemic to Madagascar) and Gentianothamnus (monotypic, endemic to Madagascar), was used as an outgroup to accommodate the dispersal-vicariance (DIVA) analyses for more reliable estimates of the ancestral areas of Exacum (Ronquist, 1996).
The genus Sebaea, the most basal clade of the tribe, was not included as an additional outgroup, as our previous analyses of the whole family including African Sebaea confirmed that inclusion or exclusion of this clade did not interfere with the geographic assessment for Exacum.
Our previous studies have shown evidence that the saprophytic genus Cotylanthera resided inside a monophyletic Exacum. It consists of four species occurring in the Himalayas, mainland Southeast Asia, and the Malaysian area. We did not include this genus in our present study because, on the one hand its sequences were incomplete (the chloroplast sequence were not obtainable); on the other hand it has been shown to have undergone an accelerated molecular evolution (cf. Yuan et al., 2003: Fig. 3), which may be problematic in the present phylogenetic analysis. However, omitting sequences of this genus had no significant effect on the tree topology (data not shown) and thus should not affect analyses on Exacum. Furthermore, this group showed affinities only with the more terminal branches of the phylogenetic tree of Exacum (E. hamiltonii from the Himalayas and E. tetragonum widespread in Indomalesia) (cf. Yuan et al., 2003: Fig 2 and Fig 3), the biogeography patterns of Exacum across the IOB region is greatly depending on the basal branching patterns (see below).
|
|
Molecular Methods
Sequences of 18 species were previously obtained (Yuan et al., 2003). Sequences for an additional 24 accessions were newly acquired for this study. The procedures of DNA extraction, polymerase chain reaction (PCR) amplification of target fragments, purification of PCR products, and sequencing follow Yuan et al. (2003). All sequences have been deposited in GenBank (accession numbers for the trnL intron: AJ490202 [GenBank] –38, 40, 42–43, 49–50; for ITS: AJ489877 [GenBank] –914, 917–918, 923–924).
Sequence Alignment, Congruence Test, and Phylogenetic Analysis
The sequences were aligned with Clustal X (Thompson et al., 1997) followed by minor manual adjustments to improve indel events in a few tandem repeat regions. Nine base pairs near the beginning of the spacer ITS2 involved tandem repeats and could not be unambiguously aligned due to alternative alignment possibilities. They were excluded from subsequent phylogenetic analyses. The alignment does not involve severe ambiguities, therefore it was not necessary to use secondary structure for alignment. Sixteen unambiguously aligned indels (13 in ITS and 3 in trnL) were potentially informative. These indels were then scored as binary characters regardless of their length, and were added to the sequence data for MP analyses. The data matrix was deposited in TreeBASE and is also available from the corresponding author.
To assess the level of congruence between the trnL and ITS data sets, we analyzed each data set independently to see if they produced a similar topology. We also performed an incongruence length difference (ILD) test of Farris et al. (1995), implemented in PAUP* 4.0b10 (Swofford, 2000) as the partition-homogeneity test, after assuring ourselves that the properties of the data did not lend themselves to biasing this test (see results). One thousand replicates of repartitions of trnL intron versus ITS were conducted with heuristic searches (simple sequence addition and TBR branch-swapping). The two data sets were confirmed as congruent (P = 0.74) and were then combined for all further analyses.
MP analyses were conducted on separate trnL, ITS data sets, and on a combination of the two, applying accelerated transformation optimization (ACCTRAN) option. Heuristic searches of 1000 replicates of random sequence addition and TBR branch-swapping were performed, with MULTREES and STEEPEST DESCENT on. The relative clade support for MP analyses was also estimated by bootstrap analysis of 1000 replicates of heuristic searches with random sequence addition and TBR branch-swapping.
Phylogenetic reconstructions based on the combined data sets was also conducted using ML optimality criteria as implemented in PAUP*4.0b10 (Swofford, 2000). The TIM+I+
model (Posada and Crandall, 1998) and parameter settings were chosen by using the Akaike information criterion as suggested by Modeltest V3.06 (Posada and Crandall, 1998). Optimal gene trees were found via heuristic searches of 100 replicates of random sequence addition with TBR branch-swapping, MULTREES, and STEEPEST DESCENT on. The relative clade support for the ML analyses was estimated by bootstrap analysis of 100 replicates of heuristic searches using the same model and parameters (Felsenstein, 1985).
Molecular Clock Test, Divergence Calibration, and Divergence Time Calculation
To obtain approximate timings of branching events within Exacum, it is necessary to estimate the times of divergence of the main clades of the Exacum tree. By comparing the likelihood ratio statistic, –2(lnLclock–lnLnon–clock), to the 
2 distribution with n–2 degrees of freedom (n = number of taxa), a molecular clock was tested (Muse and Weir, 1992). Finding that a clocklike sequence evolution had to be rejected (P < 0.01), the ML tree obtained in the absence of a molecular clock was then subjected to a rate smoothing applying penalized likelihood (PL) approach using the software r8s v.1.50 (Sanderson, 2002a, b). By applying appropriate calibrations, divergence times can be estimated on the smoothened tree. PL is a semiparametric smoothing method that estimates relative branching time without assuming a molecular clock. Optimal smoothing factors were chosen based on a data-driven cross-validation procedure implemented in r8s (Sanderson, 2002b).
There is no fossil record for the tribe Exaceae to calibrate the molecular phylogeny of Exacum. Our previous studies (Yuan et al., 2003) estimated a minimum age of 40 Mya for the tribe Exaceae calibrated on the basis of pollen fossil of Gentianales (Muller, 1984). This minimum age of the tribe could be used directly as a calibration. However, this estimate was based on the nonparametric rate smoothing (NPRS) approach (Sanderson, 1997) with only one calibration. NPRS approaches have been proven to overfit the data, allowing too much rate variation and therefore losing predictive power (Sanderson, 2002a). As a further confirmation and cross-validation, we therefore reanalyzed our previous trnL intron data of the family Gentianaceae and its sister groups, with additional sequences and fossil calibrations that we recognized recently. Two trnL intron sequences of the genus Lisianthius, L. longifolius (accession AF102450 [GenBank] ) and L. laxiflorus (accession AF102449 [GenBank] ), were retrieved from GeneBank and were added to our previous data matrix to accommodate a fossil record of Lisianthius (Graham, 1984). This enlarged trnL intron matrix contained 68 taxa and 657 characters. This matrix was deposited in TreeBASE and is also available from the corresponding author. Phylogenetic reconstructions on the enlarged trnL intron matrix were performed using MP with 1000 replicates of heuristic searches with random sequence addition and TBR branch-swapping. Finding a molecular clock has to be rejected (P < 0.01) for the enlarged trnL intron data set, the resulting MP trees were then subjected to PL rate smoothing (Sanderson, 2002a) using the software r8s (Sanderson, 2002b). Four independent calibration points, minimum age of Gentianales (CP1 = 60 Mya) based on fossil pollen (Muller, 1984), minimum age of Lisianthius (CP2 = 40 Mya) based on fossil pollen (Graham, 1984), an inferred age of subtribe Swertiinae (CP3 = 15 Mya) (Hagen and Kadereit, 2001, 2002), and minimum age of Gentiana (CP4 = 5 Mya) based on fossil seeds (Mai and Walther, 1988), were used to infer the divergence dates between the outgroup Gentianothamnus-Tachiadenus clade and the ingroup Exacum-Ornichia clade. These inferred estimates of divergence dates from the global analyses were then used to calibrate the Exacum tree to obtain divergence estimates. Such a procedure has been successfully used by Davis et al. (2002a).
NPRS and PL may result in biased rates to the evenly weighted MP branch lengths because they are not corrected for multiple hits (Sanderson, 2002a). Ideally, the branch lengths should be corrected for multiple hits using an appropriate model of evolution (Yang, 1996). This is more critical for highly divergent sequences as the likelihood of multiple hits increases. However, in the enlarged trnL intron analysis, sequence divergence was moderate (maximal 18% for ingroup taxa) and our calibration dates for the Exacum tree inferred from PL smoothened MP trees was not affected significantly.
As an independent cross-check of the divergence-time estimates based on the smoothened ML tree, we also estimated the times of divergence of the trnL intron and ITS sequences employing various extreme rates measured for different plants, compiled by Richardson et al. (2001). Divergence time between a pair of species was calculated as half of divergence value divided by the rate. The highest sequence distance between a pair of species of two clades under consideration was used to ensure the divergence times were not underestimated.
Ancestral Area Reconstruction
In order to obtain the historical scenarios of the biogeography of Exacum, DIVA analyses were conducted to infer the ancestral areas for each internal node of the phylogeny, by using DIVA 1.1a (Ronquist, 1996, 1997), on the ML tree. DIVA reconstructs ancestral areas by minimizing dispersal and extinction events needed to explain the observed distribution pattern based on an inferred phylogeny. Vicariance is considered as the default mode of speciation. The two zero-length branches of the ML tree were arbitrarily resolved to obtain dichotomies; this had no effect on the area assignment. Based on the analyses of the distribution of Exacum and its relatives, we recognized eight areas of endemism: Africa, Madagascar, Sri Lanka and south India, Socotra and southern Arabian peninsula, central to northern India, the Himalayas, mainland Southeast Asia, and Malaysia plus northern Australia (Fig. 1). Terminal taxa were scored according to their distributions across the above eight areas to generate the distribution data matrix, and this data matrix was subsequently optimized onto the ML tree. Optimizations of both, unconstrained and area constrained to a maximum of two, were conducted following the reasoning of Ronquist (1996, 1997) and Donoghue et al. (2001).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
Sequence Characteristics
The trnL intron sequences ranged from 487 to 495 bp in length in Exacum, whereas the intron was as short as 381 to 382 bp in the two species of Ornichia mainly due to a long gap of 106 bp. The aligned trnL intron matrix had 508 sequence and three indel characters. The number of variable characters was 79 (15.5%), of which 36 (7.0%) were informative. The uncorrected pairwise sequence divergence was between 0% (E. tetragonum—K254 versus E. hamiltonii; E. macranthum—FK767 versus E. pallidum; E. marojeyense versus E. fruticosum; E. marojeyense versus E. humbertii; E. fruticosum versus E. humbertii) and 3.9% (E. sessile versus E. trinervium—Zsl002) within the ingroup and maximal 4.5% (E. sessile versus T. longiflorus) overall. The ITS sequences ranged from 611 to 638 bp in length. These length variations are due to single or short indels. The ITS data matrix had 659 sequence and 13 indel characters, of which 266 (39.6%) were variable and 179 (26.6%) informative. The taxa had 0% (E. affine—M8238a versus E. affine—Wcff5; E. tetragonum—K254 versus E. hamiltonii; E. marojeyense versus E. fruticosum) to 13.0% (E. sessile versus E. trinervium—Zsl002) within ingroup and maximal 16.4% (E. trinervium—Zsl002 versus T. longiflorus) uncorrected pairwise sequence divergence across ingroup and outgroup taxa.
The two sets of sequences were revealed as congruent by the ILD test (P = 0.74). Recent studies and simulations suggested that the ILD test could fail to detect congruence due to different noise levels of the data sets (Dolphin et al., 2000; Yoder et al., 2001) or incongruence (Dowton and Austin, 2002; Darlu and Lecointre, 2002) due to large difference of the sizes and evolutionary conditions of the data partitions. Our data sets had similar sizes (511 characters in trnL and 672 characters in ITS. ITS showed a ca. 3.5-fold higher divergence (0% to 16.4%) compared to trnL (0% to 4.5%), but both revealed similar tree topology when analyzed separately. Thus we consider that our two data sets were not suffering from these limitations, and, therefore, were combined. The combined data matrix had 1167 bp sequence characters plus 16 binary indel characters. It was deposited in TreeBASE and is also available on the web site http://www.unine.ch/bota/ebolab/gentianaceae/gentmain.html Of the 1167 bp combined sequence, 9 bp (0.8%) ambiguously aligned ITS sequences were excluded from phylogenetic analyses.
Phylogenetic Analysis
MP analyses on trnL intron sequences alone generated 1829 equally most parsimonious trees of 90 steps (CI = 0.89 excluding autapomorphic sites, RI = 0.94). The strict consensus (not shown) is poorly resolved, and only the south India–Sri Lanka clade (b in Fig. 2) is retained. MP analyses on ITS data alone generated 2893 trees of 558 steps (CI = 0.58 excluding autopomorphic sites, RI = 0.82). The strict consensus (not shown) is highly resolved and the topology of the consensus is almost identical to that generated from the combined data. The topology of the main clades remained the same as the combined analyses, with resolution within one Madagascan clade (a2 in Fig. 2) slightly lower.
MP analyses on the combined data generated 270 most parsimonious trees of 648 steps (CI = 0.60 excluding autapomorphic sites, RI = 0.83) (results not shown). The ML analysis on the combined sequence data generated a single tree (–ln = 4350.2744) (Fig. 2). The ML tree was congruent with the strict consensus of the combined MP trees and was identical to the most likely MP tree (one of the 270 MP trees that retained the highest likelihood score when measured with ML criteria).
As shown in Fig. 2, the genus Exacum was resolved as monophyletic, with moderate (68%, ML) to high (90%, MP) bootstrap support. The sister relationship between Exacum and Ornichia was highly supported (100%, both ML and MP). The most basal split within Exacum separated the highly supported Africa–Madagascar clade (93% for ML and 99% for MP; Fig. 2, clade a) from the highly supported Asia–Socotra clade (89% for ML and 85% for MP, Fig. 2). Although all the sampled Madagascan species together were monophyletic, they group into two clades: one strongly supported clade (94% for ML and 97% for MP), composed of the small-flowered species that are mainly found in open habitats (clade a1 in Fig. 2), and a weakly supported clade (51% for ML and 61% for MP), including the species with larger flowers that are found mainly in forests (clade a2 in Fig. 2). The African species, E. oldenlandioides, nested deeply inside the small-flowered Madagascan clade. Within the Asia–Socotra lineage, three highly supported clades were present (clades b, c, d in Fig. 2): a Sri Lanka–south India clade (100% for both ML and MP), composed of the showy species endemic to Sri Lanka and the very southern part of India (clade b); the Socotra–Arabia clade (99% for ML and 100% for MP), containing species endemic to Socotra and the southern Arabia peninsula (clade d); and the Indomalesia clade (87% for ML and 95% for MP), including the species endemic to the Himalayas, mainland Southeast Asia, the species spread in Sri Lanka and India, and the species widely spread over the entire Indomalesia area including northern Australia (clade c). The Socotra–Arabia clade showed a closer relationship to the Indomalesia clade than to any other clade (65% for ML and 83% for MP).
Divergence Calibration and Dating
In order to obtain calibrations for the Exacum tree, phylogenies were reconstructed for 68 representatives of the family Gentianaceae and its sister groups on trnL intron data. MP analyses with MIN collapse option generated six equally parsimonious trees of 463 steps (CI = 0.68 excluding autopomorphic sites, RI = 0.89). The strict consensus was well resolved and the main clades were highly supported (Fig. 3). As a molecular clock has to be rejected, these six MP trees with branch lengths optimized with ACCTRAN option were subjected to PL rate smoothing. Applying the four different calibration points to the smoothened MP trees, the divergence times between Gentianothamnus-Tachiadenus clade and Exacum-Ornichia clade (the node marked with a star in Fig. 3) are estimated as 20.3 to 25.9 Mya (CP1), 23.9 to 53.7 Mya (CP2), 18.2 to 40.8 Mya (CP3), 12.0 to 26.4 Mya (CP4), and 29.2 to 32.3 (CP1+CP2+CP3+CP4), respectively (Fig. 3).
Using these calibrations and the smoothened ML tree, the dates of the main divergence events within Exacum are estimated (Table 1). Dating based on maximum sequence divergences and extreme rates obtained comparable figures for corresponding nodes (Table 1). The divergence between Exacum and its sister group Ornichia was estimated to be 9.0 to 39.4 Mya, the divergence between the Africa–Madagascar clade ("a") and the Socotra–Asia clade ("b-c-d") to be about 8.2 to 35.6 Mya, the divergence between the Sri Lanka–South India clade ("b") and Socotra–Indomalesia clade ("c-d") to be about 7.4 to 31.9 Mya, and the divergence between the Socotra clade ("c") and the Indomalesia clade ("d") to be about 6.5 to 27.7 Mya. The divergence between the two Madagascan clades ("a1" and "a2") is estimated as 5.8 to 23.5 Mya, and the divergence between the African species E. oldenlandioides and its closest Madagascan ally is estimated to be only about 1.9 to 5.3 Mya. The divergence between the mainland Southeast Asian endemic E. sutaepense and its sister clade is estimated as 5.5 to 23.7 Mya. The divergence between the Socotra endemic species E. caeruleum and the species E. affine, that is common to both Socotra and the southern Arabian peninsula, is estimated to be about 2.7 to 11.6 Mya (Fig. 4). Noticeably, none of these estimates dated the divergence of Exacum beyond the Eocene. Meanwhile, it is conspicuous that, whatever calibration is used, the divergences among the main clades "a," "b," "c," and "d," clearpage corresponding to the main areas of endemism, seem have happened within a relatively short period of time, which suggests a relatively rapid and extensive radiation of the progenitors of the clades.
|
|
Dispersal-Vicariance Analysis
As anticipated (Ronquist, 1996, 1997), unconstrained reconstructions by DIVA (not shown) suggested the three basal nodes of the tree, the most recent common ancestor (MRCA) of the outgroup (Gentianothamnus-Tachiadenus) and the ingroup (Ornichia-Exacum), the MRCA of Exacum and Ornichia, and the MRCA of Exacum, to be widespread in the areas "B" through "H" or "B" through "G." The area "A" (Africa) is never postulated as a possible ancestral area for these basal nodes. The MRCA of the clades "b," "c," and "d" was suggested to be spread in the areas "C" through "G" or "C" through "H." Thus, a vicariance between Madagascar (area "B") and the Asian areas plus Socotra (areas "C" through "H") is suggested for the first splitting of Exacum.
The optimal reconstruction of the unconstrained analyses required 13 dispersals (mainly due to widespread terminal taxa such as E. tetragonum). When the maximum number of areas assigned to each node is restricted to two, the optimal reconstruction of the ancestral areas required only two more dispersals than the unconstrained optimization. The results of the two-area optimization are shown in Figure 4. Madagascar (area "B") is consistently assigned to the MRCA of Gentianothamnus-Tachiadenus and Ornichia-Exacum and the MRCA of Ornichia and Exacum. The constrained optimization reveals that Exacum originated in Madagascar and specifies an initial dispersal (as indicated by "+C" in Fig. 4) of the MRCA of Exacum from Madagascar (area "B") to Sri Lanka and south India (area "C") at early diversification stage of the genus. This single dispersal resulted in the vicariance pattern between Madagascar and Sri Lanka–south India. Similarly, DIVA specifies the MRCA of the clades "b" and "c-d" to have dispersed from Sri Lanka–south India (area "C") to Socotra–Arabia (area "D"), and the MRCA of the clade "c" further dispersed from the area "C" to the mainland Southeast Asia (area "G"). Each of these two dispersals is followed by a vicariance. The African species, E. oldenlandioides, is also resolved as a vicariant resulted from the dispersal of the MRCA of E. oldenlandioides and its sister clade from Madagascar to Africa.
DIVA gives less certain resolution about the ancestral areas of the Indomalesia lineage. Although the most recent ancestor of E. sutaepense in mainland Southeast Asia is confirmed as being dispersed from Sri Lanka–south India, multiple possibilities exist regarding the ancestral areas of the widespread species E. tetragonum and its closest relatives. Either Sri Lanka–south India, or the Himalayas, or both were involved as possible ancestral areas.
|
Discussions |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
Phylogeny of Exacum
Our present ML tree based on combined data was identical to the most likely MP tree (under ML criteria), and its topology is congruent with the Exacum clade structure of our previous study on the whole tribe Exaceae where fewer Exacum species were included (Yuan et al., 2003). The present phylogeny of Exacum based on molecular data is also highly congruent with the previous hypothesis inferred from morphological and anatomical characters (Klackenberg, 1985), with regard to the monophyly and the main infrageneric lineages resolved. However, significant difference is shown for the different position of the Socotran–Arabian clade. In our present molecular phylogeny, the Socotran–Arabian clade showed a closer relationship to the Indomalesia clade than to any other clade, whereas in Klackenberg's (1985) morphological phylogeny, they showed a closer relationship to the African–Madagascan clades rather than to any Asian clade: the Socotran–Arabian clade fell on a polytomy together with three other Madagascan–African clades. All African, Madagascan, and Socotran–Arabian clades together formed a monophyletic clade, sister to the Asian clade.
Klackenberg (1985) divided Exacum into two sections, sect. Exaum and sect. Africana, based on his morphological phylogeny. The relative length of the pedicel and the internode below was used as an important character to distinguish these two sections. Our present molecular phylogeny, however, suggests that his section Africana is polyphyletic.
Divergence Dating and Calibration
Obtaining an absolute time for a divergence/ speciation offers an elegant and explicit test on whether the divergence resulted from a given historical event. Different lines of evidence have indicated that the minimum ages of the order Gentianales (53.2 Mya: Magallon et al., 1999), the family Gentianaceae (ca. 50 Mya: Yuan et al., 2003), and the tribe Exaceae (ca. 40 Mya: Yuan et al., 2003) are all well after the Gondwana breakup. Without an explicit dating, Klackenberg (2002) speculated that the Gentianaceae might be over 90 Mya old, based on the vicariant distribution of the basal clades of the family and an assumed correlation between this pattern and the Gondwana history. Except for our previous study where only a few Exacum species were sampled (Yuan et al., 2003), the absence of a temporal scale in the previous phylogeny established for Exacumprohibits such a direct test (Klackenberg, 1985). A temporal calibration of the phylogenetic tree of the genus Exacum remains necessary in order to gain insight into biogeographic consequences of post-Gondwana vicariance/dispersal events, even if a Gondwana hypothesis can be rejected based on the dating of the order Gentianales (Magallon et al., 1999), the family, or the tribe (Yuan et al., 2003). Absence of fossil data for Exacum and even the tribe Exaceae prohibits a direct calibration of the Exacum tree. Fossil records for the entire family Gentianaceae are scarce. Muller (1984) estimated the minimum age of the order Gentianales as 60 Mya based on fossil pollen of the other families allied with the Gentianaceae. Here we used this date as the calibration point CP1; molecular divergence dating suggested a close figure for this order (53.2 Mya: Magallon et al., 1999). The earliest megafossil of suggested Gentianaceae origin was the fossil flowers with Pistillipollenites pollen from the Eocene (ca. 45 Mya) of North America. These fossil pollens were suggested to be associated with the relatively derived Macrocarpaea of the extant Gentianaceae (Crepet and Daghlian, 1981), but such a suggestion was unconfirmed (Stockey and Manchester, 1986). Struwe et al. (2002) doubted the Gentianaceae origin of these fossil flowers, and Crepet himself recently also upholds such a consideration (personal communication). We therefore did not use this fossil record as a calibration. Graham (1984) associated some fossil pollens isolated from lignite deposit developed from mangrove swamps of the middle to late Eocene (ca. 40 Mya) in Panama to the extant Lisianthius. We simply used this fossil record as our second calibration CP2. However, the taxonomic attribution of this fossil pollen needs to be further confirmed, because, as Graham (1984) admitted, the fossils differ from pollen of extant Lisianthius in shape, and, meanwhile, other associated fossil pollen from the same deposit were mostly mangroves such as Rhizophora and Pelliceria. The estimated age (15 Mya) of the subtribe Swertiinae was based on ITS sequence divergence of Gentianella calibrated with both fossil pollen and geological evidence (based on the ITS rate of 4.52 ± 2.12 x 10–9 substitutions per site per year obtained from an analysis on 44 Gentianella species: Hagen and Kadereit, 2001, 2002). Having similar habit and generation time, the rates of Gentianella may potentially applicable to Exacum. Thus we used both the estimated age of Swertiinae and the ITS rate to calibrate the Exacum divergence. Both gave consistent dating results (Table 1). The fossil seeds of Gentiana from the Upper Pliocene of Germany (ca. 5.0 Mya: Mai and Walther, 1988) seems to be a relatively recent representative. The lower bonds of the dating results we obtained are due to this "young" calibration point.
Even though our present dating on the branching events within Exacum was calibrated with inferred dates based on a global tree of Gentianaceae, a Gondwana vicariance scenario for the biogeography of Exacum can be refuted. The PL dating suggested that the divergence of Exacum was not beyond the Eocene. Meanwhile, our present study revealed a close correlation between PL dating and divergence rate dating on the estimates of the divergences within Exacum. Dating results on both trnL intron and ITS sequences with previously reported fast and slow rates (despite rejection of a molecular clock) are highly consistent with PL approaches in concluding that the divergence between Exacum and it sister group Ornichia was less than 40 Mya (Table 1). There is no evidence to assume that Exacum sequences evolve dramatically slower than other plants, and therefore it is less likely that the divergence of Exacum was underestimated. Fossil-calibrated NPRS and molecular clocks suggested that the divergence between the Gentianothamnus-Tachiadenus clade and the Exacum-Ornichia clade fell between 11.2 and 29.8 Mya (Yuan et al., 2003). Our present analyses applying PL approaches with additional calibrations confirmed our previous results and suggested this divergence to be 9.0 to 39.4 Mya, well after the breakup of the Gondwana continent. In fact, reconciling dates of fossil data with molecular data, Magallon et al. (1999) showed that few angiosperm lineages had evolved at that time, and the divergence of the order Gentianales (53.2 Mya) was well after Gondwana breakup. Our present PL estimate on the first divergence of Exacum, the divergence between the Madagascan and Indian–Sri Lankan lineages was 8.2 to 35.6 Mya, whereas geological data suggest that the most recent close contact between the Madagascar–Seychelles and Indian plates was before 65 Mya (Storey, 1995). If this time frame is correct, it is obviously impossible to interpret the history of biogeography of Exacum as a result of Gondwana breakup (Klackenberg, 1985, 2002).
Our Exacum divergence estimates further suggest that the main infrageneric lineages diverged rapidly within a relatively short period of time, regardless of which calibration point is used. The four lineages "a, b, c, d" corresponding to the main areas of endemism seemed to have emerged within ca. 8 million years (9.0–6.5 to 35.6–27.7 Mya depending on calibration point). If the Gondwana hypothesis has to be rejected, dispersals and extensive radiations have to be considered as the main causes to explain the present-day distribution pattern of Exacum.
Out-of-Madagascar Dispersals
Unconstrained DIVA analysis suggested the basal nodes, MRCA of Gentianothamnus-Tachiadenus and Ornichia-Exacum, the MRCA of Exacum and Ornichia, and the MRCA of Exacum as widespread, and the first divergence within Exacum as the result of a vicariance event between Madagascar and Asian areas. As mentioned above, such a vicariance without dispersal involvement is physically impossible, because at the time of this divergence (8.2 to 35.6 Mya), Madagascar (area "B") was already isolated from other areas (Storey, 1995). If we assume the coexistence of the MRCA of Gentianothamnus-Tachiadenus and Ornichia-Exacum, the MRCA of Exacum and Ornichia, and the MRCA of Exacum, in Madagascar, Socotra, and other Asian areas, we have to assume extinctions of Gentianothamnus, Ornichia, and Tachiadenus in Socotra and all other Asian areas, which is an unlikely scenario. Such a resolution is because DIVA considers wide distribution and vicariance as "default" optimization (Ronquist, 1996, 1997).
The constrained optimization of DIVA offers a more plausible estimation of the historical biogeography of Exacum, particularly with regard to the divergence of the basal lineages corresponding to the main areas of endemism around the IOB. As shown in Fig. 4, the constrained DIVA optimization suggested the MRCA of Exacum and its sister group, Ornichia, originated in Madagascar. The constrained DIVA optimization specified unambiguously four dispersal events within Exacum: an initial dispersal of the MRCA of Exacum from Madagascar (area "B") to Sri Lanka and south India (area "C") at early diversification stage of the genus (when there was only a single ancestral taxon in Exacum extant), the dispersal of the MRCA of the clades "b" and "c-d" from Sri Lanka–South India (area "C") to Socotra–Arabia (area "D"), the dispersal of the MRCA of the clade "c" from the area "C" to the mainland Southeast Asia (area "G"), and the dispersal of the MRCA of E. oldenlandioides and its sister clade from Madagascar to Africa. Each of these dispersal events was followed by a vicariance. Because the present dating suggested the first divergence of Exacum was well after the Gondwana breakup, these unambiguous vicariance events revealed by DIVA cannot be the results of tectonic history, but instead the results of single-dispersal events. By the time of Exacum divergence (< 35.6 Mya), the plates of mainland Africa, Madagascar, and Sri Lanka–south India had mostly attained their present-day positions (ca. 500 km between mainland Africa and Madagascar, and ca. 3600 km between Madagascar and Sri Lanka–south India), and therefore, these were necessarily long-distance dispersals.
Having originated in Madagascar, Exacum has experienced multiple out-of-Madagascar dispersals. The most important is the long-distance dispersal to Sri Lanka–south India, which resulted in the extensive radiation of the Socotra–Arabia and other Asian lineages in the northern IOB regions. More recent out-of-Madagascar dispersals include the dispersal to the African continent, as represented by the divergence of E. oldenlandioides, and to other islands around Madagascar, as indicated by the distributions of E. stenopterum, which occurs in both Madagascar and the Comoros, and E. quinquenervium, which occurs in both Madagascar and the volcanic island of Mauritius (Klackenberg, 1985).
These long-distance dispersals could be directly or via stepping-stones. The recent dispersal from Madagascar to Africa (4.7 Mya), which resulted in the divergence of E. oldenlandioides, could have happened directly or via the Comoros as stepping-stones (the oldest island, Mayotte, is 5.5 Mya old; Emerick and Duncan, 1982). As for the dispersal from Madagascar to the Indian subcontinent, in addition to the possibility of direct dispersal, an Eocene–Oligocene land bridge (Steenis, 1962) or the so-called "Lemurian stepping-stones" (Schatz, 1996) connecting India, Sri Lanka, the Seychelles, and Madagascar has been proposed as the main channel for plants and animals to disperse across the IOB. In fact, dispersals through this channel can be considered as stepwise long- or medium-distance dispersals because there is no geological evidence for a continuous land bridge known today (the Seychelles at present are ca. 1000 km from Madagascar and ca. 2600 km from Sri Lanka–south India) (Schatz, 1996). Further evidence is needed to confirm if such a dispersal channel has ever existed. Nevertheless, with regard to Exacum there is no reason to exclude the possibility of dispersals via the "Lemurian stepping-stones," although no extant Exacum species is found on the Seychelles (Klackenberg, 1985, 2002).
Yet the process and mechanism of these long-distance dispersals are not well understood. Without apparent epizoic or hydrozoic adaptations, the possible means of dispersal of Exacum are wind and birds. Birds have been proposed as the most plausible and common agents of long-distance dispersals for carrying seeds or fruits of continental plants to colonize oceanic islands. For example, it was suggested that the Hawaiian colonizer of the genus Viola was dispersed from the Arctic areas by migratory birds (Ballard and Sytsma, 2000). Birds might have played roles in transferring Exacum from Madagascar to the Sri Lanka–south India area. As soon as Exacum arrived in the Sri Lanka–south Indian area, it might have extensively expanded its range by means of anemochory dispersals. Although wind dispersal usually represents a slow and gradual process of range expansion, it could fulfill a relatively rapid radiation for Exacum: although the seeds of Exacum are small and light, and released from xerochastic capsules by external forces such as wind, they might be too large to be carried far by wind (like orchids). However, many extant species of Exacum have prominent wings along the ribs of the saclike calyx, and the calyx, particularly the wings, become conspicuously enlarged and hardened during fruit development (Klackenberg, 2002). Seeds or capsules entrapped in the calyx sac can quickly travel long distances along the ground in an open environment when caught in wind. The two most widespread species, E. tetragonum and E. oldenlandioides, indicate such a capability. The former is widespread in many open places across the entire Indomalesia region, from south India, the Himalayas, mainland Southeast Asia, to the Philippines, New Guinea, and northern Australia, whereas the latter is seen in almost the entire tropical Africa with little morphological variation, further supporting a rapid and extensive expansion in distribution.
An extensive radiation of Exacum in the northern IOB conforms to the paleoclimate variation in these regions (deMenocal, 1995; Cerling et al., 1997; Ramstein et al., 1997; Zachos et al., 2001; Billups and Schrag, 2002; Griffin, 2002). It is plausible to speculate that this genus originated in Madagascar when the global temperature was rising, and soon after its origin, the progenitor of Exacum was dispersed to the area of Sri Lanka–south India by long-distance dispersal. Following the colonization of this area, this ancestral lineage spread out in the northern IOB regions including the Socotra–Arabia area (by that time the Afro–Arabia plates had attained its connection with Asia but Socotra was still connected to Arabia) and diversified at different places as a response to the favorite warm and humid climate during the Eocene or the Miocene climatic optimum (17 to 15 Mya). At the same time, Madagascan lineages had undergone extensive diversification as well. Subsequently, following the spreading Neogene drought, particularly towards the end of the Miocene, due to a decrease of atmospheric CO2 (Cerling et al., 1997), a global cooling linked to the reestablishment of a major ice sheet over Antarctica (Zachos e al., 2001; Billups and Schrag, 2002) and the drying up in the northern IOB regions induced mainly by the rapid uplifting of the Himalayas (deMenocal, 1995; Ramstein et al., 1997; Griffin, 2002), global vegetation experienced a drastic change (e.g., widespread increase of C4 biomass in favor of the dry climate). Consequently, Exacum species were forced to retreat from most parts of the northern IOB regions and survived in isolation in Socotra, south India–Sri Lanka, and perhaps mainland Southeast Asia. The Madagascan lineage was probably less drastically affected. From these core areas, "secondary" diversifications and radiations occurred later on, which resulted in several relatively distantly related clusters of species—the main clades we observe today. A similar pattern has been suggested for the genus Nepenthes (Meimberg et al., 2001). It is thought that Nepenthes has originated in the north of the ancient Tethys sea or on the Indian subcontinent after it was separated from Madagascar. This progenitor is supposed to have undergone extensive spreading: on land migration to Southeast Asia, and dispersals through land bridge or stepping-stones to the Seychelles and Madagascar. This habitat expansion was followed by incisive extinctions of the widespread progenitors in the paleotropical regions (i.e., fragmentation). Subsequently, secondary diversification occurred on the Southeast Asian islands, whereas the species on the Seychelles and Madagascar remained relatively unchanged (Meimberg et al., 2001).
The collision of the Afro–Arabian plate with Asia opened another channel of migration for flora and fauna between Africa and Asia since the Miocene (Samuei et al., 1997). Furthermore, a possible land bridge between mainland Africa and Madagascar due to a temporal dry-out in large areas of the Mozambique Channel between 45 and 26 Mya has been suggested as a possible channel for mammals to colonize Madagascar (McCall, 1997). Our present reconstructions of the phylogeny and ancestral areas, however, do not offer any support for a dispersal of Exacum from Madagascar to the Socotra–Arabian area via mainland Africa and then to India–Sri Lanka, because our phylogeny suggested that the Sri Lanka–south India lineage diverged prior to the divergence between the Socotra–Arabian and Indomalesian clades. Moreover, such a consideration has to assume a complete extinction of the first African colonizers on the entire African continent. The widespread African species E. oldenlandioides is firmly shown to have originated from a recent dispersal from Madagascar. The only other extant African species, E. zombense, endemic to the Zomba mountain of Malawi, was unavailable for our present phylogenetic assessment. However, morphologically it is similar to the recently dispersed E. oldenlandioides (Klackenberg, 1985) and unlikely represents the relict of the first African colonizers.
The present article has given an interesting insight into the evolution and biogeographic history of the genus Exacum and sheds lights on the noticeable biogeographic patterns of the paleotropical regions around the IOB. Despite some uncertainty in dating caused by the scarcity of reliable fossils, we were able to establish, to the best of our knowledge, a geographical and temporal origin of Exacum and older lineages therein. The evolution history of Exacum apparently involved a combination of long-distance dispersal events and extensive radiations.
|
|
Acknowledgments |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
We thank the Botanical and Zoological Park of Tsimbazaza and Association Nationale pour la Gestion des Aires Protégées (Antananarivo, Madagascar), Louis Zeltner, Tony Miller, and Philippe Chassot for help in sample collections, and Jason Grant for critical reading of the manuscript. We are much indebted to anonymous reviewers for critical comments. This study was financially supported by Swiss National Science Foundation (Grant 3100-052885).
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionsAcknowledgmentsReferences |
|---|
-
Ballard H. E. Jr., Sytsma K. J. Evolution and biogeography of the woody Hawaiian violets (Viola, Violaceae): Arctic origins, herbaceous ancestry and bird dispersal. Evolution (2000) 54:1521–1532.[Web of Science][Medline]
Barron E. J. Cretaceous plate tectonic reconstruction. Paleogeogr. Paleoclimatol. Paleoecol. (1987) 59:3–29.[CrossRef]
Baum D. A., Small R. L., Wendel J. F. Biogeography and floral evolution of Baobabs (Adansonia, Bombacaceae) as inferred from multiple dat sets. Syst. Biol. (1998) 47:181–207.
Beck R. A., Burgen D. W., Sercombe W. J., Riley G. W., Barndt J. K., Berry J. R., Afzal J., Khan A. M., Jurgen H. J., Metje Cheema A., Shafique N. A., Lawrence R. D., Khan M. A. Stratigraphic evidence for an early collision between northwest India and Asia. Nature (1995) 373:55–58.[CrossRef]
Besse J. V., Courtillot Pozzi J. P., Westphal M., Zhou Y. X. Paleomagnetic estimates of crustal shortening in the Himalayan thrusts and Zangbo suture. Nature (1984) 311:621–626.[CrossRef]
Billups K., Schrag D. P. Paleotemperatures and ice volume of the past 27 Myr revisited with paired Mg/Ca and O-18/O-16 measurements on benthic foraminifera. Paleoceanography (2002) 17:1–11.
Birse A. C. R., Bott W. F., Morrison J., Samuel M. A. The Mesozoic and Early Tertiary tectonic evolution of the Socotra area, eastern Gulf of Aden, Yemen. Mar. Petr. Geol. (1997) 14:675–684.[CrossRef]
Cerling T. E., Harris J. M., MacFadden B. J., Leakey M. G., Quade J., Eisenmann V., Ehleringer J. R. Global vegetation change through the Miocene/Pliocene boundary. Nature (1997) 389:153–158.[CrossRef]
Coffin M. F., Rabinowitz P. D. Evolution of the conjugate East African-Madagascan margins and the western Somali Basin. Geo. Soc. Am. Spec. Pap. (1988) 226:1–78.
Crepet W. L., Daghlian C. P. Lower Eocene and Paleocene Gentianaceae: Floral and palynological evidence. Science (1981) 214:75–77.
Darlu P., Lecointre G. When does the incongruence length difference test fail. Mol. Biol. Evol. (2002) 19:432–437.
Davis C. C., Bell C. D., Fritsch P. W., Mathews S. Phylogeny of Acridocarpus-Brachlophon (Malpighiaceae): Implications for Tertiary tropical floras and Afroasian biogeography. Evolution (2002a) 56:2395–2405.[Web of Science][Medline]
Davis C. C., Bell C. D., Mathews S., Donoghue M. J. Laurasian migration explains Gondwana disjunctions: Evidence from Malpighiaceae. Proc. Natl. Acad. Sci. USA (2002b) 99:6833–6837.
deMenocal P. B. Pliopleistocene African climate. Science (1995) 270:53–59.
Dolphin K., Belshaw R., Orme C. D. L., Quicke D. L. J. Noise and incongruence: Interpreting results of the inconguence length difference test. Mol. Phylog. Evol. (2000) 17:401–406.[CrossRef][Web of Science][Medline]
Donoghue M. J., Bell C. D., Li J. Phylogenetic patterns in northern hemisphere plant geography. Int. J. Plant Sci. (2001) 162:S41–S52.[CrossRef][Web of Science]
Dowton M., Austin A. D. Increased congruenc does not necessarily indicate increased phylogenetic accuracy—The behavior of the incongruence length difference test in mixed-mode analyses. Syst. Biol. (2002) 51:19–31.
Emerick C. M., Duncan R. A. Age progressive volcanism in the Comoros Archipelago, western Indian Ocean and implications for Somali plate tectonics. Earth Planet. Lett. (1982) 60:415–428.
Fantozzi P. L., Sgavetti M. Tectonic and sedimentary evolution of the eastern Gulf of Aden continental margins: New structral and stratigraphic data from Somalia and Yemen. In: Sedimentation and tectonics in rift basins: Red Sea—Gulf of Aden—Purser B. H., Bosence D. W. J., eds. (1998) London: Chapman & Hall. 56–76.
Farris J. S., Källersjö M., Kluge A. G., Bult C. Testing significance of incongruence. Cladistics (1995) 10:315–319.[CrossRef][Web of Science]
Felsenstein J. Confidence limits on phylogenies: An approach using the bootstrap. Evolution (1985) 39:783–791.[CrossRef][Web of Science]
Ghebreab W. Tectonics of the Red Sea region reassessed. Earth Sci. Rev. (1998) 45:1–44.[CrossRef]
Graham A. Lisianthius pollen from the Eocene of Panama. Ann. Missouri Bot. Gard. (1984) 71:987–993.[CrossRef][Web of Science]
Griffin D. L. Aridity and humidity: Two aspects of the late Miocene climate of North Africa and the Mediterranean. Paleogeogr. Paleoclimatol. Paleoecol. (2002) 182:65–91.[CrossRef]
Hagen K. B., von Kadereit J. W. The phylogeny of Gentianella (Gentianaceae) and its colonization of the southern hemisphere as revealed by nuclear and chloroplast DNA sequence variation. Org. Divers. Evol. (2001) 1:61–79.[CrossRef]
Hagen K. B., von Kadereit J. W. Phylogeny and flower evolution of the Swertiinae (Gentianaceae-Gentianeae): Homoplasy and the principle of variabl proportions. Syst. Bot. (2002) 27:548–572.
Jaeger J., Courtillot V., Tapponnier P. Paleontological view of the ages of the Deccan traps, the Cretaceous Tertiary boundary, and the India-Asia collision. Geology (1989) 17:316–319.
Jolivet L., Faccenna C. Mediterranean extension and the Africa-Eurasia collision. Tectonics (2000) 19:1095–1106.[CrossRef][Web of Science]
Klackenberg J. The genus Exacum (Gentianaceae). Oper. Bot. (1985) 84:1–144.
Klackenberg J. Tribe Exaceae. In: Gentianaceae, systematics and natural history—Struwe L., Albert V. A., eds. (2002) Cambridge: Cambridge University Press. 66–108.
Klootwijk C. T., McFadden P. L., Gee J. S., Peirce J. W., Smith G. M. An early India-Asia contact—Paleomagnetic constraints from Ninetyeast Ridge. ODP LEG 121. Geology (1992) 20:395–398.
Lee T. -Y., Lawver L. A. Cenozoic plate reconstruction of the southeast Asia region. Tectonophysics (1995) 251:85–138.[CrossRef][Web of Science]
Li Z. X., Powell C. M. An outline of the palaeogeographic evolution of the Australasian region since the beginning of the Neoproterozoic. Earth Sci. Rev. (2001) 53:237–277.[CrossRef]
Magallon S., Crane P. R., Herendeen P. S. Phylogenetic pattern, diversity, and diversification of eudicots. Ann. Missouri Bot. Gard. (1999) 86:297–372.[CrossRef][Web of Science]
Mai D. H., Walther H. Die pliozaenen Floren von Thueringen, Deutsche Demokratische Republik. Quartaerpalaeontologie (1988) 7:55–297.
McCall R. A. Implications of recent geological investigations of the Mozambique Channel for the mammalian colonization of Madagascar. Proc. R. Soc. Lond. B (1997) 264:663–665.[Medline]
McLoughlin S. The breakup history of Gondwana and its impact on pre-Cenozoic floristic provincialism. Aust. J. Bot. (2001) 49:271–300.
Meimberg H., Wistuba A., Dittrich P., Heubl G. Molecular phylogeny of Nepenthaceae based on cladistic analysis of plastid trnK intron sequenc. data. Plan. Biol. (2001) 3:164–175.[CrossRef]
Muller J. Significance of fossil pollen for angiosperm history. Ann. Missouri Bot. Gard. (1984) 71:419–443.[CrossRef][Web of Science]
Muse S. V., Weir B. S. Testing for equality of evolutionary rates. Genetics (1992) 132:269–272.[Abstract]
Patriat P., Achache J. India Eurasia collision chronology has implications for crustal shortening and driving mechanism of plates. Nature (1984) 311:615–621.[CrossRef]
Plunkett G. M., Lowry P. P., Burke M. K. The phylogenetic status of Polyscias (Araliaceae) based on nuclear its sequence data. Ann. Missouri Bot. Gard. (2001) 88:213–230.[CrossRef][Web of Science]
Posada D., Crandall K. A. Modeltest: Testing the model of DNA substitution. Bioinformatics (1998) 14:817–818.
Rabinowitz P. D., Coffin M. F., Falvey D. The separation of Madagascar and Africa. Science (1983) 220:67–69.
Ramstein G., Fluteau F., Besse J., Joussaume S. Effect of orogeny, plate motion and land sea distribution on Eurasian climate change over the past 30 million. Nature (1997) 386:788–795.[CrossRef]
Raven P. H., Axelrod D. I. Angiosperm biogeography and past continental movements. Ann. Missouri Bot. Gard. (1974) 61:539–673.[CrossRef][Web of Science]
Raxworthy C. R., Forstner M. R. J., Nussbaum R. A. Chameleon radiation by oceanic dispersal. Nature (2002) 415:784–787.[Medline]
Renner S. S., Clausing G., Meyer K. Historical biogeography of Melastomaceae: The roles of Tertiary migration and long-distance dispersal. Am J. Bot. (2001) 88:1290–1300.
Renner S. S., Meyer K. Melastomeae come full circle: Biogeographic reconstruction and molecular clock dating. Evolution (2001) 55:1315–1324.[Web of Science][Medline]
Richardson S. M., Bott W. F., Smith B. A., Hollar W. D., Bermingham P. M. A new hydrocarbon play area offshore Socotra-Island, Republic of Yemen. J. Petr. Geol. (1995) 18:5–28.[CrossRef]
Richardson J. E., Pennington R. T., Pennington T. D., Hollingsworth P. M. Rapid diversification of a species rich genus of Neotropical rain forest trees. Science (2001) 293:2242–2245.
Rieppel O. A case of dispersing chameleons. Nature (2002) 415:744–745.[Web of Science][Medline]
Ronquist F. (1996) DIVA version 1.1. Computer program and manual available by anonymous FTP from Uppsala University, ftp.uu.seftp.systbot.uu.se.
Ronquist F. Dispersal-vicariance analysis: A new approach to the quantification of historica. biogeography. Syst. Biol. (1997) 46:195–203.
Samuei M. A., Harbury N., Bott R., Thabet A. M. Field observations from the Socotran platform: Their interpretation and correlation to Southern Oman. Mar. Petr. Geol. (1997) 14:661–673.[CrossRef]
Sanderson M. J. A nonparametric approach to estimating divergence times in the absence of rate constancy. Mol. Biol. Evol. (1997) 14:1218–1231.[Web of Science]
Sanderson M. J. Estimating absolute rates of molecular evolution and divergence time: A penalized likelihood approach. Mol. Biol. Evol. (2002a) 19:101–109.
Sanderson M. J. (2002b) r8s, version 1.50 user's manual. Available via http://ginger.ucdavis.edu/r8s/.
Sanmartin I., Ronquist F. Southern hemisphere biogeography inferred by event-based models: Plant versus anima. patterns. Syst. Biol. (2004) 53:216–243.
Schatz G. E. Malagasy/Indo-Australo-Malesian phytogeographic connections. In: Biogeography of Madagascar—Lourenco R., ed. (1996) ORSTOM editions. Paris: Steenis, C. G. G. J. van. 73–83. The land-bridge theory in botany. Blumea. 1962, 11, 235-372.
Stockey R. A., Manchester S. R. A fossil flower with in situ Pistillipollenites from the Eocene of British Columbia. Can J. Bot. (1986) 66:313–318.[CrossRef]
Storey B. C. The role of mantle plumes in continental breakup: Case histories from Gondwanaland. Nature (1995) 377:301–308.[CrossRef]
Storey M., Mahoney J. J., Saunders A. D., Duncan R. A., Kelley S. P., Coffin M. F. Timing of hot spot-related volcanism and the breakup of Madagascar and India. Science (1995) 267:852–855.
Struwe L., Kadereit J. W., Klackenberg J., Nilsson S., Thiv M., von K. B., Hagen Albert V. A. Systematics, character evolution, and biogeography of Gentianaceae, including a new tribal and subtribal classification. In: Gentianaceae—Systematics and natural history—Struwe L., Albert V. A., eds. (2002) Cambridge: Cambridge University Press. 21–309.
Swofford D. L. PAUP* phylogenetic analysis using parsimony (*and other methods) (2000) Sunderland, Massachusetts: Sinauer Associates. Version 4.
Thompson J. D., Gibson T. J., Plewniak F., Jeanmougin F., Higgins D. G. The Clustal X windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. (1997) 24:4876–4882.
Thulin M. Exacum (Gentianaceae) on the Arabian peninsula and Socotra. Nord. J. Bot. (2001) 21:243–247.[CrossRef]
Valdiya K. S. Emergence and evolution of Himalaya: Reconstructing history in the light of recent studies. Prog. Phys. Geogr. (2002) 26:360–399.[CrossRef]
Vences M., Freyhof J., Sonnenberg R., Kosuch J., Veith M. Reconciling fossils and molecules: Cenozoic divergence of cichlid fishes and the biogeography of Madagascar. J. Biogeogr. (2001) 28:1091–1099.[CrossRef]
Yang Z. Among-site rate variation and its impact on phylogenetic analysis. Trends Ecol. Evol. (1996) 11:367–372.[CrossRef]
Yoder A. D., Irwin J. A., Payseur B. A. Failure of the ILD to determine data combinability for slow lori. phylogeny. Syst. Biol. (2001) 50:408–424.
Yuan Y. -M., Wohlhauser S., Möller M., Chassot P., Mansion G., Grant J., Küpfer P., Klackenberg J. Monophyly and relationships of the tribe Exaceae (Gentianaceae) inferred from nuclear ribosomal and chloroplast DNA sequences. Mol. Phylog. Evol. (2003) 28:500–517.[CrossRef][Web of Science][Medline]
Zachos J., Pagani M., Sloan L., Thomas E., Billups K. Trends, rhythms, and aberrations in global mate 65 Ma to present. Science (2001) 292:686–693.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. W. Clayton, P. S. Soltis, and D. E. Soltis Recent Long-Distance Dispersal Overshadows Ancient Biogeographical Patterns in a Pantropical Angiosperm Family (Simaroubaceae, Sapindales) Syst Biol, August 14, 2009; (2009) syp041v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Moller, M. Pfosser, C.-G. Jang, V. Mayer, A. Clark, M. L. Hollingsworth, M. H. J. Barfuss, Y.-Z. Wang, M. Kiehn, and A. Weber A preliminary phylogeny of the 'didymocarpoid Gesneriaceae' based on three molecular data sets: Incongruence with available tribal classifications Am. J. Botany, May 1, 2009; 96(5): 989 - 1010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. K. M. Kulju, S. E. C. Sierra, S. G. A. Draisma, R. Samuel, and P. C. v. Welzen Molecular phylogeny of Macaranga, Mallotus, and related genera (Euphorbiaceae s.s.): insights from plastid and nuclear DNA sequence data Am. J. Botany, October 1, 2007; 94(10): 1726 - 1743. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Syring, K. Farrell, R. Businsky, R. Cronn, and A. Liston Widespread Genealogical Nonmonophyly in Species of Pinus Subgenus Strobus Syst Biol, April 1, 2007; 56(2): 163 - 181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Spalik and S. R. Downie The evolutionary history of Sium sensu lato (Apiaceae): dispersal, vicariance, and domestication as inferred from ITS rDNA phylogeny Am. J. Botany, May 1, 2006; 93(5): 747 - 761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Schwarz, S. Fuller, S. M. Tierney, and S. J. B. Cooper Molecular Phylogenetics of the Exoneurine Allodapine Bees Reveal an Ancient and Puzzling Dispersal from Africa to Australia Syst Biol, February 1, 2006; 55(1): 31 - 45. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Hordijk and O. Gascuel Improving the efficiency of SPR moves in phylogenetic tree search methods based on maximum likelihood Bioinformatics, December 15, 2005; 21(24): 4338 - 4347. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||