Phylogenetic and biogeographic complexity of Magnoliaceae in the Northern Hemisphere inferred from three nuclear data sets

https://doi.org/10.1016/j.ympev.2008.06.004Get rights and content

Abstract

This study employs three nuclear genes (PHYA, LFY, and GAI1) to reconstruct the phylogenetic and biogeographic history of Magnoliaceae. A total of 104 samples representing 86 taxa from all sections and most subsections were sequenced. Twelve major groups are well supported to be monophyletic within Magnoliaceae and these groups are largely consistent with the recent taxonomic revision at the sectional and subsectional levels. However, relationships at deeper nodes of the subfamily Magnolioideae remain not well resolved. A relaxed clock relying on uncorrelated rates suggests that the complicated divergent evolution of Magnolioideae began around the early Eocene (54.57 mya), concordant with paleoclimatic and fossil evidence. Intercontinental disjunctions of Magnoliaceae in the Northern Hemisphere appear to have originated during at least two geologic periods. Some occurred after the middle Miocene, represented by two well-recognized temperate lineages disjunct between eastern Asia and eastern North America. The others may have occurred no later than the Oligocene, with ancient separations between or within tropical and temperate lineages.

Introduction

The intercontinental disjunctions of plants in the Northern Hemisphere are considered to be the most complex biogeographic pattern observed at the global scale (Wen, 1999, Wen, 2001, Milne and Abbott, 2002, Donoghue and Smith, 2004, Milne, 2006). Such disjunctions are generally thought to be remnants of a more continuously distributed, mixed mesophytic forest during the Tertiary, known as the “boreotropical flora” (Wolfe, 1975). It subsequently became fragmented due to geologic and climatic changes. Some of the ancient relict mesic forests that once covered much of the temperate regions of the Northern Hemisphere can be found today in the southeast region of USA as well as in eastern to central China and central to southern Japan.

Although the “boreotropical flora” hypothesis has been well accepted, questions still remain (Donoghue and Smith, 2004). These concern their overall phylogenetic relationships, morphological correlations (convergence/stasis) of temperate pairs in different areas, and morphological divergence between temperate and tropical relatives (Wen, 1999, Donoghue and Smith, 2004, Ickert-Bond et al., 2007). The disjunct taxa with relatives in the tropics are biogeographically much more complicated (Wen, 1999). To better understand the complexity requires the evaluation of phylogenetic relationships and divergence times in a broader phylogenetic framework, including closely related elements spanning from the temperate regions to the subtropical and tropical zones.

One potentially important study group showing such a disjunct pattern is the family Magnoliaceae, which is distributed from the north temperate to the tropical regions. Based on the fossil record (Tao and Zhang, 1992, Frumin and Friis, 1996, Frumin and Friis, 1999), this family has been considered to be one of the earliest extant lineages of flowering plants (93.5–110 mya) and has played a crucial role toward our understanding of the origin and diversification of angiosperms. Both morphological and molecular evidence have supported this scenario (Takhtajan, 1969, Cronquist, 1981, Mathews and Donoghue, 1999, Parkinson et al., 1999, Qiu et al., 1999, Soltis et al., 1999, Graham and Olmstead, 2000).

Magnoliaceae is comprised of ca. 220–240 species characterized by an androecium of numerous spirally arranged stamens, a gynoecium with many simple carpels spirally arranged on an elongated axis, and separate tepals (Law, 1984, Law, 1996, Law, 2004, Liu et al., 1995, Figlar and Nooteboom, 2004). Roughly two-thirds of the species are currently distributed in temperate and tropical regions of eastern to southeastern Asia (Fig. 1). The other third occur in the New World from the temperate eastern North America through tropical South America as far as Brazil and Bolivia (Dandy, 1927, Dandy, 1971, Dandy, 1978b, Law, 1984, Nooteboom, 1993, Nooteboom, 1998, Thorne, 1993, Frodin and Govaerts, 1996). The distribution of Magnoliaceae in the north temperate as well as tropical regions in Asia and the Americas (Fig. 1), makes it an excellent model for understanding the evolution of intercontinental temperate disjunctions and their interactions with the tropical members.

Magnoliaceae are usually divided into two subfamilies: Magnolioideae and Liriodendroideae (Law, 1984, Nooteboom, 1985). Except for the very distinct Liriodendroideae (including only Liriodendron with two species), taxonomists have long debated over the classification of Magnolioideae due to a paucity of phylogenetically useful characters (Dandy, 1927, Dandy, 1978a, Dandy, 1978b, Law, 1984, Law, 1996, Qiu et al., 1995a, Qiu et al., 1995b, Jobes, 1998, Nooteboom, 1985, Nooteboom, 1998, Chen and Nooteboom, 1993, Azuma et al., 1999, Azuma et al., 2001, Kim et al., 2001, Sun and Zhou, 2004, Xu and Rudall, 2006). Most recently, a new classification for Magnolioideae was proposed by Figlar and Nooteboom (2004) based on the chloroplast phylogenetic results and morphological re-examinations (also see Figlar, 2006). This classification recognized only one genus Magnolia in the Magnolioideae subfamily.

Molecular phylogenetic studies on Magnoliaceae have been conducted using chloroplast data (e.g., Qiu et al., 1995a, Qiu et al., 1995b, Azuma et al., 1999, Azuma et al., 2001, Azuma et al., 2004, Shi et al., 2000, Kim et al., 2001, Wang et al., 2006). One of the first important findings from molecular evidence was the polyphyly of Magnolia section Rhytidospermum, with the North American Magnolia tripetala closely related to the Asian counterparts, rather than to M. macrophylla, or M. fraseri from the same continent (Qiu et al., 1995a, Qiu et al., 1995b). The studies also indicated that Manglietia and Michelia were nested within Magnolia. Kim et al. (2001) used ndhF sequences of 99 taxa representing all of the traditional Magnoliaceae lineages and produced a most parsimonious tree containing 11 clades. However, many clades were weakly supported and the relationships among these clades remained unclear. Other molecular studies produced similar results (e.g., Azuma et al., 2000, Azuma et al., 2001, Azuma et al., 2004, Ueda et al., 2000).

The limited divergence of chloroplast sequences and the relative morphological homogeneity in Magnoliaceae have made it difficult to resolve the phylogenetic relationships within the family (Qiu et al., 1995a, Azuma et al., 2001, Kim et al., 2001, Li and Conran, 2003). Nuclear data have not been employed for phylogenetic studies in Magnoliaceae so far. Since nuclear markers have been generally shown to provide stronger phylogenetic signals than many plastid sequences (Wolfe et al., 1987, Small et al., 2004), we herein use three nuclear genes to resolve the phylogenetic relationships of Magnoliaceae. Phytochrome evolution in land plants has been shown to result from a series of gene duplications (e.g., PHYA, PHYB, PHYC, and PHYE) that have led to independent and functionally distinct lines (Mathews et al., 1995, Mathews and Sharrock, 1996, Manabe and Nakazawa, 1997, Mathews et al., 2003). Nucleotide variation at phytochrome loci has been useful in various phylogenetic studies of basal angiosperms and of several other angiosperm families (Kolukisaoglu et al., 1995, Mathews and Donoghue, 1999). The low-copy nuclear LEAFY (LFY) gene has been used for phylogenetic analyses of plants with its introns showing a relatively high nucleotide substitution rate (Hoot and Taylor, 2001, Archambault and Bruneau, 2004). The Arabidopsis GA INSENSITIVE (GAI) locus and related genes of the DELLA subfamily encode growth regulators and have been implicated in quantitative variation for developmental traits (Peng et al., 1999, Silverstone et al., 1998, Thornsberry et al., 2001, Boss and Thomas, 2002). The GAI-like gene sequence (GAI1) derived from a grapevine dwarf mutant was examined in a phylogenetic context within the Vitaceae and has been shown to be a potentially important marker (Wen et al., 2007).

Biogeographic studies have been conducted on some disjunct lineages in Magnoliaceae (Parks et al., 1983, Parks et al., 1994, Parks and Wendel, 1990, Qiu and Parks, 1994, Qiu et al., 1995a, Qiu et al., 1995b, Azuma et al., 2001). Liriodendron was estimated to have diverged in the mid-Miocene based on allozyme and restriction fragment length polymorphism (RFLP) analyses of cpDNA and paleobotanical evidence (Parks and Wendel, 1990). Subsequently, the divergence time of the genus was specifically estimated to be 27.9 ± 4.4 mya based on plastid trnK, psbA_trnH, and atpB_rbcL sequences using strict molecular clocks calibrated with fossil evidence (Azuma et al., 2001). Other lineages in the family suggested additional disjunctions. Magnolia sect. Rhytidospermum, the North American M. tripetala was shown to be sister to the Asian counterparts and their divergence was estimated to be during the late Miocene to early Pliocene using the allozyme and RFLP data of cpDNA (Qiu et al., 1995a, Qiu et al., 1995b), and 20.9 ± 3.3 to 27.9 ± 4.4 mya using the cpDNA sequences by Azuma et al. (2001). Divergence times in Magnoliaceae previously were all dated based on strict molecular clocks and discrepancies were found between different data sets. To update our biogeographic understanding of the family, we performed age estimation using “relaxed clock” analyses and multiple fossil calibrations (Renner, 2005).

This study employed sequences of three nuclear markers to infer the phylogenetic relationships in Magnoliaceae, with comparison of previous chloroplast results. Divergence times for intercontinental disjunct clades were estimated using the nuclear sequences under a relaxed molecular clock. Geologic data were placed in the phylogenetic contexts to gain insights into the biogeographic origin and interactions of temperate and tropical elements disjunct between eastern Asia and eastern North America. The phylogenetic framework and the temporal scale that we present provide a foundation to examine the complexity and relationships of biogeographic diversification of angiosperms between temperate and tropical zones in the Northern Hemisphere.

Section snippets

Plant sampling and sequencing

A total of 104 samples representing 86 species and subspecies, and varieties were sequenced in this study (Appendix 1). This sampling scheme covers all sections and nearly all subsections (excepting subsections Dugandiodendron, Splendentes, and Maingola) of the most recent classification system of Magnoliaceae by Figlar and Nooteboom (2004). We followed Frodin and Govaerts (1996) for generic circumscription and scientific names, and Figlar and Nooteboom (2004) for the classification system.

Phylogenetic analyses based on nuclear data sets

The PHYA matrix comprised 1070 aligned positions without indels. The putative LFY sequence was 887 bp in aligned length including both a coding region (1–330) with a 3-bp indel and an intron ranging from 331 to 887. The alignment of the GAI1 sequences generated a data matrix of 1300 positions with a 9-bp indel. The strict consensus trees for each nuclear sequence significantly supported the separation of the two subfamilies (Liriodendroideae and Magnolioideae), but the phylogenetic relationships

Phylogenetic relationships within Magnoliaceae

All nuclear data clearly supported the separation of the two subfamilies, the speciose Magnolioideae and the monogeneric Liriodendroideae, as had been shown previously based on chloroplast data (Chase et al., 1993, Qiu et al., 1993, Azuma et al., 2001, Kim et al., 2001). Each of the three nuclear genes, PHYA, LFY, and GAI1 separately exhibited a relatively low level of sequence divergence, and thus the phylogenetic relationships within the subfamily Magnolioideae were not well resolved (their

Acknowledgments

This study was supported by grants from the National Basic Research Program of China (973 Program, 2007CB411601), the Natural Science Foundation of China (NSFC 30625004 and 40771073 to H. Sun), and the John D. and Catherine T. MacArthur Foundation (to J. Wen), and by the Laboratory of Analytical Biology of the Smithsonian Institution’s National Museum of Natural History. We appreciate the valuable comments from R. Figlar of the Magnolia Society International, South Carolina, USA.

References (139)

  • P.K. Boss et al.

    Association of dwarfism and floral induction with a grape ‘green revolution’ mutation

    Nature

    (2002)
  • S.R.S. Cevallos-Ferriz et al.

    Vegetative remains of the Magnoliaceae from the Princeton chert (Middle Eocene) of British Columbia

    Can. J. Bot.

    (1990)
  • M.W. Chase et al.

    Phylogenetics of seed plants, an analysis of nucleotide sequences from the plastid gene rbcL

    Ann. MO. Bot. Gard.

    (1993)
  • B.L. Chen et al.

    Notes on Magnoliaceae III, the Magnoliaceae of China

    Ann. MO. Bot. Gard.

    (1993)
  • W.Y. Chun

    Genus speciesque novae Magnoliacearum sinensium

    Acta Phytotaxon. Sin.

    (1963)
  • M.E. Collinson et al.

    Magnoliophyta (Angiospermae)

  • A. Cronquist

    An Integrated System of Classification of Flowering Plants

    (1981)
  • C.W. Cunningham

    Can three incongruence tests predict when data should be combined?

    Mol. Biol. Evol.

    (1997)
  • J.E. Dandy

    The genera of Magnoliaceae

    Kew Bul.

    (1927)
  • J.E. Dandy

    The classification of the Magnoliaceae

    Newsl. Am. Magnolia Soc.

    (1971)
  • J.E. Dandy

    Key to subgenera and sections of Magnolia

  • J.E. Dandy

    Revised survey of the genus Magnolia together with Manglietia and Michelia

  • J.H. Degnan et al.

    Discordance of species trees with their most likely gene trees

    PLoS Genet.

    (2006)
  • D.L. Dilcher et al.

    Archaeanthus: an early angiosperm from the Cenomanian of the western interior of North America

    Ann. MO. Bot. Gard.

    (1984)
  • M.J. Donoghue et al.

    Patterns in the assembly of temperate forests around the Northern Hemisphere

    Philos. Trans. R. Soc. Lond. B Biol. Sci.

    (2004)
  • P.I. Dorofeev et al.

    Magnoliaceae-Eucommiaceae

  • J.J. Doyle et al.

    A rapid DNA isolation procedure for small quantities of fresh leaf tissue

    Phytochem. Bul.

    (1987)
  • A.J. Drummond et al.

    Relaxed phylogenetics and dating with confidence

    PLoS Biol.

    (2006)
  • A.J. Drummond et al.

    BEAST: Bayesian evolutionary analysis by sampling trees

    BMC Evol. Biol

    (2007)
  • S.V. Edwards et al.

    Gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies

    Evolution

    (2000)
  • J. Felsenstein

    Phylogenies from molecular sequences, inference and reliability

    Ann. Rev. Genet.

    (1988)
  • R.B. Figlar

    Stone Magnolias

    Arnoldia

    (1993)
  • Figlar, R.B., 2006. Taxonomy topics - a new classification for Magnolia. In: Royal Horticultural Society (Ed.),...
  • R.B. Figlar et al.

    Notes on Magnoliaceae IV

    Blumea

    (2004)
  • E.M. Friis et al.

    Fossil history of magnoliid angiosperms

  • D.G. Frodin et al.

    World Checklist and Bibliography of Magnoliaceae

    (1996)
  • S.I. Frumin et al.

    Magnoliid reproductive organs from the Cenomanian-Turonian of north-western Kazakhstan: Magnoliaceae and Illiciaceae

    Plant Syst. Evol.

    (1999)
  • X. Gong et al.

    The cross compatibility of Magnoliaceae

    Acta Bot. Yunnan

    (2001)
  • A. Graham

    Late Cretaceous and Cenozoic History of North American Vegetation

    (1999)
  • S.W. Graham et al.

    Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms

    Am. J. Bot.

    (2000)
  • T.A. Hall

    BioEdit, a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT

    Nucl. Acids. Symp. Ser.

    (1999)
  • A.L. Hipp et al.

    Congruence versus phylogenetic accuracy, Revisiting the incongruence length difference test

    Syst. Biol.

    (2004)
  • S.Y. Ho et al.

    Accuracy of rate estimation using relaxed-clock models with a critical focus on the early metazoan radiation

    Mol. Biol. Evol.

    (2005)
  • S.B. Hoot et al.

    The utility of nuclear ITS, a LEAFY homolog intron, and chloroplast atpB–rbcL spacer region data in phylogenetic analyses and species delimitation in Isoetes

    Am. Fern J.

    (2001)
  • J.P. Huelsenbeck et al.

    MRBAYES, Bayesian inference of phylogenetic trees

    Bioinformatics

    (2001)
  • S.M. Ickert-Bond et al.

    Comparative infructescence morphology in Altingia (Altingiaceae) and discordance between morphological and molecular phylogenies

    Am. J. Bot.

    (2007)
  • Y.H. Ji et al.

    Phylogeny and classification of Paris (Melanthiaceae) inferred from DNA sequence data

    Ann. Bot.

    (2006)
  • D.V. Jobes

    Molecular systematics of the Magnoliaceae: a review of recent molecular studies

  • S. Kim et al.

    Phylogenetic relationships in family Magnoliaceae inferred from ndhF sequences

    Am. J. Bot.

    (2001)
  • S. Kim et al.

    DNA sequences from Miocene fossils, an ndhF sequence of Magnolia latahensis (Magnoliaceae) and an rbcL sequence of Persea pseudocarolinensis (Lauraceae)

    Am. J. Bot.

    (2004)
  • Cited by (140)

    • Sesquiterpene lactone and its unique proaporphine hybrids from Magnolia grandiflora L. and their anti-inflammatory activity

      2022, Phytochemistry
      Citation Excerpt :

      Paleontologists have suggested that the earliest plants in Magnoliaceae appeared 95 million years ago, and fossilized plant specimens of Magnolia acuminate L. appeared 20 million years ago. Thus, the Magnoliaceae family has been used to track the historical values of plant evolution and morphological changes (Nie et al., 2008). Magnoliaceae is generally a spreading evergreen or deciduous tree, or shrub characterized by large fragrant flowers that can be bowel-shaped or star-shaped in white, pink, or purple colors.

    • The evolutionary history of the Caribbean magnolias (Magnoliaceae): Testing species delimitations and biogeographical hypotheses using molecular data

      2022, Molecular Phylogenetics and Evolution
      Citation Excerpt :

      In Magnoliaceae, sister-relationships among well-supported and recognized clades (Table 2) other than section Talauma are inconsistently placed when comparing the different gene trees (Appendix D), hence, their relationships remain unresolved in the summary tree (Fig. 4). This result agrees with previous Sanger sequencing based studies (Azuma et al., 2011 and precursors; Kim and Suh, 2013 and precursor; Nie et al., 2008). The inconsistent topology of the clades in Magnoliaceae depending on the genetic region studied contrasts with the robustness of the separate clades.

    • Anthropogenic climate change increases vulnerability of Magnolia species more in Asia than in the Americas

      2022, Biological Conservation
      Citation Excerpt :

      This genus also has a wide altitudinal distribution range from northern latitudes to tropical ecosystems (Sánchez-Velásquez et al., 2016). In addition, Magnoliaceae is one of the earliest extant lineages of flowering plants (Nie et al., 2008a). Therefore, the Magnolia genus has been a model system in the study of intercontinental disjunctions and the origin and diversification of angiosperms, attracting research attention from biogeographers and molecular botanists (Hebda and Irving, 2004; Nie et al., 2008b; Veltjen et al., 2019; Wang et al., 2020; Ye et al., 2019).

    View all citing articles on Scopus
    View full text