Molecular Phylogenetics and Evolution 68 (2013) 443–460 Contents lists available at SciVerse ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Evolutionary events in Lilium (including Nomocharis, Liliaceae) are temporally correlated with orogenies of the Q–T plateau and the Hengduan Mountains ⇑ Yun-Dong Gao a,b, AJ Harris c, Song-Dong Zhou a, Xing-Jin He a, a Key Laboratory of Bio-Resources and Eco-Environment of Ministry of Education, College of Life Science, Sichuan University, Chengdu 610065, China b Chengdu Institute of Biology, Chinese Academy of Sciences, Chengdu 610041, China c Department of Botany, Oklahoma State University, Oklahoma 74078-3013, USA article info abstract Article history: The Hengduan Mountains (H-D Mountains) in China flank the eastern edge of the Qinghai–Tibet Plateau Received 21 July 2012 (Q–T Plateau) and are a center of great temperate plant diversity. The geological history and complex Revised 24 April 2013 topography of these mountains may have prompted the in situ evolution of many diverse and narrowly Accepted 26 April 2013 endemic species. Despite the importance of the H-D Mountains to biodiversity, many uncertainties Available online 9 May 2013 remain regarding the timing and tempo of their uplift. One hypothesis is that the Q–T Plateau underwent a final, rapid phase of uplift 8–7 million years ago (Mya) and that the H-D Mountains orogeny was a sep- Keywords: arate event occurring 4–3 Mya. To evaluate this hypothesis, we performed phylogenetic, biogeographic, Hengduan Mountains divergence time dating, and diversification rate analyses of the horticulturally important genus Lilium, Lilium–Nomocharis complex Intercontinental dispersal including Nomocharis. The Lilium–Nomocharis complex is distributed throughout the temperate Northern Diversification rate Hemisphere but is most diverse within the H-D Mountains and Q–T Plateau. Our matK and ITS phyloge- Bayesian Binary Method (BBM) nies support previous studies showing that Nomocharis is nested within Lilium. However, we detected Reconstruct Ancestral State in Phylogenies incongruence between the two gene trees which may result from hybridization. Dating analyses per- (RASP) formed using the ITS dataset showed that the evolution of major lineages within Lilium–Nomocharis Biodiversity hotspot may be temporally coincident with Q–T Plateau uplift occurring 8–7 Mya and H-D Mountains uplift approximately 4–3 Mya. Our analyses of diversification times and rates among Lilium–Nomocharis clades are less conclusive. However, these do suggest high extinction rates among H-D Mountains lineages. Ó 2013 Elsevier Inc. All rights reserved. 1. Introduction mountains and deep valleys (Li et al., 1995), which are thought to have profoundly accelerated the diversification of plant species The Qinghai–Tibetan Plateau (Q–T Plateau) is bounded by asso- through local vicariance, secondary contact, and ecological specia- ciated mountain ranges, which reflect a complex history of uplift tion events (Liu et al., 2006). Thus, high species richness within the and quiescence in the region (Zhang et al., 2000; Pan et al., eastern Q–T plateau and adjacent areas may be strongly correlated 2012). The rich topography of the plateau and its broad-reaching with the region’s geomorphology. climatic influence have probably favored the evolution of the di- The Hengduan Mountain chain (H-D Mountains, Fig. 1; see also verse flora of eastern Asia; namely within the its mountain ranges http://hengduan.huh.harvard.edu/fieldnotes for further informa- but also within Japan and central China (Richardson et al., 2001a,b; tion) occupies the southeastern margin of Q–T Plateau and is Willis and Whittaker, 2002; Wharton et al., 2005; Liu et al., 2006). known as one of the world’s 25 or 34 biodiversity hotspots based Uplift of the plateau began approximately 40 million years ago on species richness and sensitivity to anthropogenic activities (Mya) in the Eocene (Chung et al., 1998). Recent evidence indicates (Myers et al., 2000; Wilson, 1992; http://www.biodiversityhot- that the southern margin reached its present elevation approxi- spots.org/xp/Hotspots; Richardson et al., 2001a,b). The region mately 15–22 Mya (Guo et al., 2002; Spicer et al., 2003) and that may be one of the most geologically complex in China (Pan, the whole plateau may have attained its present altitude by 1989) due to the extensive geomorphism that has occurred around 7–8 Mya (Harrison et al., 1992). Orogeny along the eastern margin the Sichuan Basin resulting from differences in crustal strength be- of the plateau since the early Miocene (22 Mya) has created high tween the basin and the surrounding, uplifted area (Royden et al., 2008). Evidence from molecular and floristic studies has implicated ⇑ Corresponding author. Fax: +86 028 85415006. the orogeny and complex topography of the H-D Mountains in E-mail address: [email protected] (X.-J. He). driving the high plant species richness observed in the region. 1055-7903/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ympev.2013.04.026 444 Y.-D. Gao et al. / Molecular Phylogenetics and Evolution 68 (2013) 443–460 Fig. 1. Major geographical features of East Asia and the location of Hengduan Mountains (H-D Mountains). For example, Li (1980) conducted a floristic survey the wetland speciation may provide insight into the timing of orogenic events plant genus Arisaema Mart. within the H-D Mountains and found (e.g., Antonelli et al., 2009; Picard et al., 2008; Stigall Rode and Lie- high diversity among both ‘‘primitive’’ and derived lineages. Thus, berman, 2005; Che et al., 2010). For example, a historical biogeo- Li (1980) hypothesized that multiple, divergent clades within graphic study of the family Rubiaceae was used to test the Arisaema underwent recent, independent diversifications within alternative hypotheses that the Andes Mountains rose continu- the region. Similarly, Wang et al. (2010) speculated that genetic ously or in stages of uplift and quiescence (Antonelli et al., 2009). differentiation among H-D Mountain populations of the aquatic Thus, it may be possible to test the competing hypotheses regard- herb Batrachium bungei (Steudel) L. is best explained by isolation ing the timing of the H-D Mountain orogeny using biological data. due to topographical complexity. Additional investigations using The genus Lilium, including Nomocharis, is an excellent model species-rich groups are needed to test the biological mechanisms system for testing the H-D Event hypothesis due to its geographic of these bursts of speciation. In particular, did uplift cause local distribution within the H-D Mountains and, more broadly, within vicariance, provide corridors for secondary contact, promote eco- Q–T Plateau region. Lilium is primarily distributed within the logical speciation, or all of these in concert? Northern Hemisphere and has three main ranges including the Further, despite frequent speculations about the role of the H-D Q–T Plateau in East Asia (including the Himalayas and the H-D Mountain orogeny in plant diversification, few studies have used Mountains), North America, and the Caucasus in Europe (Patterson the tools of evolutionary biology and historical biogeography to and Givnish, 2002; Liang, 1995). Of all these regions, the Q–T Pla- investigate the timing of these geological events. There are two teau harbors the greatest diversity of Lilium. Within Q–T Plateau conflicting hypotheses that seek to explain H-D Mountain uplift. region, there are 40 species of Lilium; 30 of these plus all species The most widely accepted of these suggests that the H-D Moun- of Nomocharis are endemic to the H-D Mountains (Lighty, 1968; tains rose in the latter stages of the eastern Q–T Plateau orogeny Baranova, 1969; Sealy, 1983; Liang, 1995; Liang and Tamura, and attained their present height around 7 Mya in the Miocene 2000). Further, a previous phylogenetic and biogeographic study (Harrison et al., 1992; Shi et al., 1998). More recently, Chen of Liliales revealed that the Himalayas were the geographic origin (1992, 1996) showed evidence from lithostratigraphy, biostratigra- of Lilium and that the genus was later dispersed into the rest of phy, magnetostratigraphy and geomorphology indicating that the Eurasia and North America (Patterson and Givnish, 2002). Despite uplift of the H-D Mountains occurred later in the Miocene, around the potential utility of Lilium and Nomocharis for better under- 3.4 Mya, following a period of relative geological stability, and he standing H-D Mountain orogeny, many questions about the origins called this the H-D Mountain Movement or Event. He proposed and intercontinental migrations of Lilium remain unanswered. that the H-D Movement should be recognized as an independent The inferred timing of evolution of Lilium also makes the genus phase in the Q–T Plateau orogeny (Chen, 1992, 1996). Previous a good model system. Previous studies show that Lilium probably authors have suggested that the timing and rate of biological evolved during the Miocene. Thus, eastern Asian clades may have Y.-D. Gao et al. / Molecular Phylogenetics and Evolution 68 (2013) 443–460 445 been affected by geological activity occurring within the H-D within or among sections (Comber, 1949; Lighty, 1960, and papers mountain region of the Q–T Plateau. Vinnersten and Bremer cited in). (2001) used rbcL sequences to estimate the divergence times of In this study, we used Lilium–Nomocharis as a model system for the major clades of Liliales and showed that the genus Lilium evaluating the timing and biological ramifications of the H-D evolved 10 Mya. Subsequent studies, in which dating was per- Mountain orogeny. However, we first addressed several remaining formed at the familial or ordinal levels, have supported the find- questions on evolutionary relationships and species boundaries ings of Vinnersten and Bremer (2001). Within Lilium, Ikinci_ within the complex by sampling taxa more completely than in pre- (2011) determined that members of the European section Lirioty- vious studies; particularly emphasizing H-D Mountain endemic pus diverged from the rest of the genus 9 Mya ago and that speci- species.