Radiation and diversification within the Ligularia–Cremanthodium–Parasenecio complex (Asteraceae) triggered by uplift of the Qinghai-Tibetan Plateau
Introduction
A central goal of the study of biological diversity is to understand why different regions with similar environments contain different numbers of species (Qian and Ricklefs, 2000, Qian et al., 2005). Determining the causes of high biodiversity in some regions is of primary importance in biology and a principal aim of biogeographic research (Willis and Niklas, 2004, Willis and Whittaker, 2002). Molecular phylogenetic reconstructions of evolutionary relationships between living organisms are increasingly used to infer these putative causes of diversification within an historic and geographic context (Avise, 2000). Recent studies show that high numbers of plant species within regions might be due in part to bursts of speciation that occurred during the last few million years triggered by major geophysical and/or climate change (Richardson et al., 2001a, Richardson et al., 2001b), and that a significant proportion of plant diversity originated during the late Tertiary, i.e., since approximately 10 million years ago (Willis and Whittaker, 2002). However, the number of studies conducted on species rich floras remains low with most centered on groups in the Southern Hemisphere (Pennington et al., 2004). Several areas recognised as biodiversity hotspots in the Northern Hemisphere (Myers et al., 2000, Wilson, 1992), have yet to be subjected to detailed investigation. Here, we report the first molecular phylogenetic investigation of the history and evolution of a component of the flora of the Qinghai-Tibetan (Q-T) Plateau.
The Q-T Plateau is the highest and largest plateau in the world, having a mean elevation of ∼4.5 km and an area of 2.5 × 106 km2 (Zheng, 1996). The eastern part of this region and the adjacent area of southeast China has been listed as one of the world’s 25 or 34 biodiversity hotspots, based on species richness and greatest danger of anthropogenic extinction (Myers et al., 2000, Wilson, 1992; http://www.biodiversityhotspots.org/xp/Hotspots). The Q-T Plateau contains more than 12,000 species of plants in more than 1500 genera, and it is estimated that about 50 genera and more than 20% of the total species are endemic to this region (Wang et al., 1993, Wu and Wu, 1996). Although levels of plant diversity and endemism in this region are much less than those of the Cape flora (Linder, 2003) and tropical rainforests (Richardson et al., 2001a), the flora is more speciose than might be expected based on comparisons made at similar latitudes in the Northern Hemisphere (Wu and Wu, 1996). For example, the Q-T flora contributes to the high plant diversity in eastern Asia (Wang et al., 1993, Wu, 1988, Wu and Wu, 1996), which is roughly twice as rich as that of eastern North America, a region of similar area and climate (Qian et al., 2005). The high species richness of the flora of the Q-T Plateau and adjacent areas has been attributed to two major factors (Axelrod et al., 1996). One hypothesis is that an unbroken gradient of vegetation from tropical rain forest to boreal coniferous forests was maintained in the region and adjacent areas throughout the Quaternary when massive extinctions occurred elsewhere in the Northern Hemisphere. This therefore acted as a major refugium for organisms in the region during the period of marked climatic oscillation. The other scenario assumes that accelerated speciation occurred following the collision of the Indian subcontinent with Asia commencing about 40 Ma. Some ancient taxa, i.e., Trochodendraceae, Cecidiphyllaceae, Eucommiaceae, and several primitive genera found in the area, are monotypic or contain few species (Wang et al., 1993, Wu, 1988, Wu and Wu, 1996), indicating that the existence of Quaternary refugia might not have played an important part in generating great species richness despite having maintained some ancient groups.
The uplift of the Q-T Plateau began approximately 40 million years ago (Ma) (Chung et al., 1998) following the collision of India with Asia. Recent evidence indicates that the southern margin of the plateau reached its present elevation approximately 15 Ma (Spicer et al., 2003), if not earlier (22 Ma) (Guo et al., 2002), with the total plateau being uplifted to its present altitude by 7–8 Ma (Harrison et al., 1992) or more recently during the late Pliocene and early Pleistocene (Shi et al., 1998). These uplifts since the early Miocene have created high mountains and deep valleys within the plateau (Li et al., 1995), which could have accelerated the production of new species in allopatry, and been partly responsible for the high local and regional species richness. To investigate this possibility, we have conducted a phylogenetic analysis of the Ligularia–Cremanthodium–Parasenecio complex (hereafter referred to as the L–C–P complex) and possible allies that comprise the subtribes Tussilagininae and Tephroseridinae of the tribe Senecioneae (Asteraceae). This group exhibits high species richness in the region and in adjacent eastern Asia (Liu, 2001, Liu, 2004).
Senecioneae, the largest tribe in the Asteraceae with ∼3200 species and ∼120 genera (Bremer, 1994), has been the subject of much debate with regard to its phylogenetic composition. Nordenstam (1977) recognized two subtribes: Blennospermatinae and Senecioninae, while Jeffrey and Chen (1984) divided the Senecioneae of eastern Asia into three subtribes: Senecioninae, Tussilagininae, and Tephroseridinae. Bremer (1994) incorporated the Tephroseridinae into Tussilagininae, and acknowledged Blennospermatinae and Senecioninae as additional subtribes. But this treatment was rejected by Chen (1999) who maintained the Tussilagininae and the Tephroseridinae as separate subtribes. The L–C–P complex of the Tussilagininae is composed of ∼120 species of Ligularia, ∼70 species of Cremanthodium, ∼60 species ofParasenecio plus six monotypic or small satellite genera, i.e., Farfugium, Syneilesis, Ligulariopsis, Sinacalia, Miricacalia, and Dendrocacalia (Chen, 1999, Jeffrey and Chen, 1984, Liu, 1989, Liu, 2001, Liu, 2004). Species of Ligularia occur in a great variety of habitats in the Q-T plateau region from forests to high alpine meadows, at elevations ranging from 1000 to 4000 m. Cremanthodium species occur in alpine meadow and scree areas at altitudes ranging from 2400 to 5600 m, while most species of Parasenecio are restricted to coniferous forests. More than 200 species in the complex are endemic to the Q-T Plateau (Liu, 2004) and comprise a typical group which exhibits great diversification in this region (Wu and Wu, 1996). Most endemics are restricted to small hills or valleys, and occur either allopatrically or occasionally sympatrically. These endemics are morphologically well defined and easily recognized in the field (Chen, 1999, Liu et al., 1994, Liu et al., 2002b). However, generic circumscriptions are extremely ambiguous, especially between members of Ligularia, Parasenecio, and Cremanthodium (Liu, 2001, Liu et al., 2001), due to a lack of diagnostic morphological traits (Liu, 2001, Liu, 2004). This may reflect possible bursts of recent speciation and random fixation of similar morphological features among unrelated lineages. Two small satellite genera, Ligulariopsis and Sinacalia, of the three large genera also occur mainly in the Q-T Plateau (Chen, 1999, Liu, 2001, Liu, 2004). Ligulariopsis, a monotypic genus, is distinguished from the three speciose genera in having a morphological combination of radiate capitula and none-vaginate leaf sheathing, while the latter genus comprising four species, differs by having a morphological combination of radiate capitula, none-vaginate leaf sheathing and tuberiform rhizomes (Chen, 1999, Jeffrey and Chen, 1984, Liu, 2001, Liu, 2004). The five genera comprising the core components of the L–C–P complex mainly distributed in the Q-T Plateau have similar morphology and their delimitation is unclear. Of the remaining four satellite genera, Farfugium and Syneilesis occur from central China to Japan, while Miricacalia, and Dendrocacalia are endemic to Japan. The relationships of the complex to other genera of the Tussilagininae of eastern Asia, i.e., Tussilago, Petasites, and Doronicum, and to genera of the Tephroseridinae, i.e., Sinosenecio, Tephroseris, Nemosenecio, are not well established. Both floral microcharacters and chromosomal data suggest that the L–C–P complex is more closely related to some species of three genera of the Tephroseridinae than to the remaining genera of the Tussilagininae (Liu, 2001, Liu, 2004).
Subtribal relationships in the Senecioneae remain poorly known despite the accumulation of molecular data for the group within recent years (e.g., Bain and Golden, 2000, Comes and Abbott, 2001, Fernandez et al., 2001, Pelser et al., 2002, Pelser et al., 2003). Blennospermatinae has been widely assumed to be the basal group of the Senecioneae (Bain and Golden, 2000, Bremer, 1994, Pelser et al., 2002); however, Swenson and Bremer (1999) found that Abrotanella, a genus of the Blennospermatinae, is only weakly (one step) associated with four sampled genera (Blennosperma, Syneilesis, Senecio, and Lopholaena) of the Senecioneae, casting doubt upon which genus is basal to the tribe. Doronicum has traditionally been placed in the Tussilagininae based on its cylindrical anther-collars and x = 30, suggesting a close relationship with the L–C–P complex (Bremer, 1994, Chen, 1999, Jeffrey and Chen, 1984). However, its “Helianthoid” pollen and small chromosomes indicate an aberrant position in this subtribe (Liu, 2001, Liu, 2004). Recently, Fernandez et al. (2001) placed it at the base of the sampled genera of the Senecioneae, sister to a clade containing Blennosperma, Lopholaena, Senecio, and Syneilesis (one genus of the L–C–P complex). These findings suggest that the traditionally circumscribed Asian Tussilagininae might not be monophyletic. However, except for these aberrant genera, other genera of Senecioneae that occur out of Asia have been shown to form two monophyletic clades: the Senecioninae group and the Tussilagininae group (Bain and Golden, 2000, Panero et al., 1999, Pelser et al., 2002, Pelser et al., 2003). Although not all non-Asiatic genera of Senecioneae have been examined, the available morphological traits indicate that most un-sampled genera fit well within the Senecioninae and Tussilagininae groups (Jeffrey, 1992). Most genera of Senecioneae whose phylogenetic position is unresolved occur in eastern Asia, within two subtribes: the Tussilagininae and the Tephroseridinae (Chen, 1999, Jeffrey, 1992, Liu, 2001). Therefore, our sampling strategy focused on the L–C–P complex, but extended to cover the most representative genera of the Tussilagininae, the Tephroseridinae and a few of the Senecioninae in eastern Asia.
Following a survey of newly sequenced chloroplast and nuclear DNA data of representative species of the L–C–P complex and related genera of the Senecioneae, we aimed to (1) evaluate the relationship of the L–C–P complex to the Tephroseridinae, and to refine its circumscription in eastern Asia; (2) examine the generic delimitation of the complex against the previous classification based on morphological characters; and (3) determine underlying causes of the radiation and diversification within the L–C–P complex, which might be correlated with past geological changes in the Q-T plateau.
Section snippets
Sampling strategy, plant materials, and datasets
Our sample of species within the L–C–P complex included 20 species representing eight of the nine sections in Ligularia and Cremanthodium, five species representing three of five sections in Parasenecio, and eight species representing the satellite genera: Sinacalia, Ligulariopsis, Farfugium, Syneilesis, Miricacalia, and Dendrocacalia from eastern Asia (Fig. 1). We further sampled four species representing the other three genera of Tussilagininae: Doronicum, Tussilago, and Petasite. Only the
Phylogenetic analyses of ndhF dataset
The ndhF sequence dataset analyzed comprised 52 species of the Senecioneae, representing all recognized subtribes, and 12 species representing eight additional tribes of Asteraceae. The aligned dataset contained 2131 sites of which 276 were variable but phylogenetically uninformative, and 191 that were variable and informative (gaps excluded). Five indels (one of 3-bp, another of 9-bp, and the remainder of 6-bp) were restricted to single species, and therefore yielded no phylogenetic
Discussion
A localized lack of phylogenetic signal and poorly resolved phylogenetic relationships has been interpreted as a signature of explosive speciation or rapid radiation in some floras (see Baldwin and Sanderson, 1998, Richardson et al., 2001a, Richardson et al., 2001b, Verboom et al., 2003). Our investigation into the evolution of a morphologically diverse group of species of Ligularia, Cremanthodium, Parasenecio and closely related taxa, most of which are endemic to the Q-T Plateau, revealed that
Conclusions
Several molecular phylogenetic studies conducted on species-rich plant groups that occur in biodiversity hotspots have now yielded similar findings with regard to species diversity being the product of recent bursts of speciation triggered most likely by geophysical and/or climatic changes within these regions since the middle Miocene (Richardson et al., 2001a). In the Neotropics, however, it has been argued that a mixture of both ancient and recent diversification should be involved to explain
Acknowledgments
We are grateful to Dr. Stephen Harris for his constructive comments on an earlier version of this paper. We thank Profs. Chen Zhiduan, Lu Anmin, Liu Shangwu, Ho Tingnong, Lu Xuefeng, Shuichi Norshiro, Richard Milne, and George Miehe for their help in collecting materials in the fields, sequencing and analysis, and fruitful discussions during the past 10 years. Support for this research was provided by Key Innovation Plan KSCX-SW-106, Special Fund of Outstanding Ph.D. Dissertation, FANEDD 200327
References (85)
- et al.
Ribosomal ITS sequences and plant phylogenetic inference
Mol. Phylogenet. Evol.
(2003) - et al.
Characterization of angiosperm nrDNA polymorphism, paralogy, and pseudogenes
Mol. Phylogenet. Evol.
(2003) - et al.
A phylogeny of Packera (Senecioneae; Asteraceae) based on internal transcribed spacer region sequence data and a broad sampling of outgroups
Mol. Phylogenet. Evol.
(2000) - et al.
A phylogenetic analysis of Doronicum (Asteraceae, Senecioneae) based on morphological, nuclear ribosomal (ITS), and chloroplast (trnL-F) evidence
Mol. Phylogenet. Evol.
(2001) - et al.
Phylogeny and biogeography of Chinese sisorid catfishes re-examined using mitochondrial cytochrome b and 16S rRNA gene sequences
Mol. Phylogenet. Evol.
(2005) - et al.
Karyological studies on the Sino-Himalayas endemic genus, Cremanthodium (Asteraceae: Senecioneae)
Bot. J. Linn. Soc.
(2001) - et al.
Molecular phylogeny and biogeography of the Qinghai-Tibet Plateau endemic Nannoglottis (Asteraceae)
Mol. Phylogenet. Evol.
(2002) Phylogeny and biogeography of Rhododendron subsection Pontica, a group with a Tertiary relict distribution
Mol. Phylogenet. Evol.
(2004)- et al.
The phylogeny of the Asteridae sensu lato based on chloroplast ndhF gene sequences
Mol. Phylogenet. Evol.
(2000) - et al.
Phylogeny reconstruction in the gap between too little and too much divergence: the closest relatives of Senecio jacobaea (Asteraceae) according to DNA sequences and AFLPs
Mol. Phylogenet. Evol.
(2003)