Deep phylogenetic incongruence in the angiosperm clade Rosidae
Graphical abstract
Introduction
Genome-scale data can provide the power to resolve some of the most perplexing parts of the tree of life (e.g., Dunn et al., 2008, Lee et al., 2011, Simon et al., 2012, Smith et al., 2011, Yoder et al., 2013). Furthermore, estimates from numerous independent loci also can reveal phylogenetic incongruence caused by different evolutionary processes, such as gene duplication and loss, recombination, hybridization, lateral gene transfer, or incomplete lineage sorting (e.g., Cui et al., 2013, Degnan and Rosenberg, 2009, Doyle, 1992, Goodman et al., 1979, Hudson, 1983, Maddison, 1997, Oliver, 2013). Molecular phylogenetic analyses have resolved much of the backbone angiosperm phylogeny (e.g., Ruhfel et al., 2014, Soltis et al., 2009, Soltis et al., 2011) and clarified long-standing questions regarding relationships within major clades such as monocots (Monocotyledoneae; Chase et al., 2000, Givnish et al., 2006, Givnish et al., 2010, Graham et al., 2006, Jerrold et al., 2004; Saarela et al., 2008; Saarela and Graham, 2010), asterids (Asteridae; Albach et al., 2001, Bremer et al., 2001, Bremer et al., 2004, Hilu et al., 2003, Moore et al., 2011, Olmstead et al., 2000), and rosids (Rosidae; Hilu et al., 2003, Jansen et al., 2007, Moore et al., 2010, Qiu et al., 2010, Soltis et al., 2007, Soltis et al., 2011, Wang et al., 2009). Yet much of this work is based either largely or exclusively on chloroplast sequence data, which represent a single, linked, and usually maternally inherited locus. New sequencing technologies make it feasible to obtain data sets of numerous independent nuclear loci, which can be used to evaluate results from analyses of chloroplast gene sequence data and reveal phylogenetic conflict among loci (e.g., Burleigh et al., 2011, Duarte et al., 2010, Lee et al., 2011, Xi et al., 2014, Zeng et al., 2014).
Introgressive hybridization has played an important role in plant evolution, and incomplete lineage sorting also likely occurred during some rapid radiations. Consequently, there are numerous examples of discordance between chloroplast and nuclear gene trees in plants (e.g., Acosta and Premoli, 2010, Okuyama et al., 2005, Rieseberg and Soltis, 1991; Rieseberg and Wendel, 1993; Rieseberg et al., 1995, Rieseberg et al., 1996, Soltis and Kuzoff, 1995, Soltis and Soltis, 2009, Tsitrone et al., 2003, Wendel et al., 1995, Xi et al., 2014). Although phylogenetic analyses of angiosperm backbone relationships based on nuclear, mitochondrial, and chloroplast loci have largely agreed, one major point of conflict is the placement of COM (Celastrales–Oxalidales–Malpighiales; Endress and Matthews, 2006, Zhu et al., 2007) within the large Rosidae clade.
Rosidae comprise approximately one quarter of all angiosperm species, which are morphologically diverse, exhibit extraordinary heterogeneity in habit, habitat, and life form, and include most temperate and tropical forest trees (Wang et al., 2009). Some members possess novel biochemical pathways (e.g., production of glucosinolate, and cyanogenic glycosides for defense), and many are important crops (e.g., Fabaceae and Rosaceae). Symbioses with nitrogen-fixing bacteria are largely confined to this clade as well. Resolving relationships within Rosidae has been difficult (e.g., Hilu et al., 2003, Jansen et al., 2007, Lee et al., 2011, Moore et al., 2010, Moore et al., 2011, Qiu et al., 2010, Ruhfel et al., 2014, Soltis et al., 2005, Soltis et al., 2007, Soltis et al., 2011, Wang et al., 2009, Zhu et al., 2007) due to a series of rapid radiations (Wang et al., 2009). However, multi-gene studies have recovered two major, well-supported clades—the Fabidae (i.e., eurosids I, fabids) and Malvidae (i.e., eurosids II, malvids) (Hilu et al., 2003, Judd and Olmstead, 2004, Moore et al., 2010, Moore et al., 2011, Soltis et al., 1999, Soltis et al., 2000, Soltis et al., 2005, Soltis et al., 2007, Soltis et al., 2011, Wang et al., 2009, Xi et al., 2014).
COM contains approximately one third of all Rosidae, 870 genera and ∼19,000 species (APG, 2009). Molecular analyses, largely dominated by chloroplast genes, have usually placed COM with Fabidae (Table 1; e.g., Burleigh et al., 2009, Hilu et al., 2003, Jansen et al., 2007, Moore et al., 2010, Moore et al., 2011, Soltis et al., 2005, Soltis et al., 2007, Soltis et al., 2011, Wang et al., 2009). Analyses of the mitochondrial gene matR first suggested the placement of COM with Malvidae (Zhu et al., 2007), and subsequent studies based on nuclear or mitochondrial genes supported this placement, although typically with limited taxon sampling (Table 1; Burleigh et al., 2011, Duarte et al., 2010, Finet et al., 2010, Lee et al., 2011, Morton, 2011, Qiu et al., 2010, Shulaev et al., 2010, Xi et al., 2014, Zhang et al., 2012). Several floral characters also appear to link COM with Malvidae. For example, in COM and Malvidae species, the inner integument of the ovule is thicker than the outer integument at the time of fertilization, a feature that is extremely rare in Fabidae and other eudicots. Additionally, contorted petals and a tendency towards polystemony and polycarpy also suggest a placement of COM members with Malvidae rather than with Fabidae (Endress and Matthews, 2006, Endress et al., 2013).
Although analyses of chloroplast gene sequence data generally appear to conflict with analyses of mitochondrial and nuclear gene sequence data, these studies often differ greatly in taxon sampling and analytical methods (Table 1; but see Xi et al., 2014). Thus, it is unclear whether the different placements of COM are due to errors in the analyses or biological incongruence among loci. The level of incongruence within the nuclear genome also is unknown. We use COM as an exemplar to investigate phylogenetic incongruence at deep levels in angiosperm phylogeny. Specifically, we first compare phylogenetic results from chloroplast, mitochondrial, and nuclear data sets having similar taxon sampling and examine whether the results are robust to various character-coding and data-exclusion protocols. We also survey large-scale nuclear data sets of both single-copy and multi-copy genes to investigate the patterns of phylogenetic discordance within the nuclear genome and then discuss whether these patterns are consistent with incomplete lineage sorting (i.e., deep coalescence) (Maddison, 1997, Maddison and Knowles, 2006, Page and Charleston, 1998) or ancient hybridization and introgression (Chang et al., 2011, Cui et al., 2013, Linder and Rieseberg, 2004, Tsitrone et al., 2003, Zhang et al., 2014).
Section snippets
Materials and methods
Throughout this paper, to facilitate discussion, we treat COM, Fabidae, and Malvidae as three separate groups, despite current classifications that consider COM to be part of Fabidae (APG, 2009, Cantino et al., 2007).
Chloroplast, mitochondrial, and nuclear data sets
ML analyses of the chloroplast, mitochondrial, and nuclear multi-gene alignments with similar taxon sampling recover different placements of COM (Fig. 1, Fig. 2, Fig. 3). We focus on the relationships among members of Rosidae, but all the trees generated in our analyses in the present study are available as supplemental data and on Dryad (http://dx.doi.org/10.5061/dryad.7sg58).
The phylogeny based on the 82-taxon, 78-gene chloroplast data set largely agrees with conclusions from previous
Conflict among multi-locus phylogenetic analyses
In spite of much recent progress resolving the angiosperm tree of life, the phylogenetic placement of COM remains uncertain. Most previous efforts to place COM have used a variety of data sources, taxon sampling strategies, and phylogenetic methods (but see Xi et al., 2014). Therefore, it is difficult to determine if the conflicting placements of COM are due to errors or actual biological conflict among loci (Table 1). Our ML analyses of multi-gene chloroplast, mitochondrial, and nuclear data
Concluding remarks
Numerous plant systematics studies have demonstrated the promise of genomic data to resolve angiosperm relationships that were not evident in analyses with a few genes (Burleigh et al., 2011, Finet et al., 2010, Lee et al., 2011, Moore et al., 2010, Moore et al., 2011, Zeng et al., 2014). We demonstrate here that analyses of data sets with many unlinked loci can highlight the ambiguity and discordance in phylogenetic relationships and potentially reveal the complexity of angiosperm evolution.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgments
We thank Yin-Long Qiu, who contributed to the early design of this project, and Ning Zhang, who graciously provided us with the 92-taxon, 5-gene nuDNA alignment used in this study. This work was supported by the National Natural Science Foundation of China (NNSF 31270268), National Basic Research Program of China (No. 2014CB954101), Chinese Academy of Sciences Visiting Professorship for Senior International Scientists (grant number 2011T1S24), State Key Laboratory of Systematic and Evolutionary
References (111)
- et al.
Gene tree discordance, phylogenetic inference and the multispecies coalescent
Trends Ecol. Evol.
(2009) - et al.
Multigene phylogeny of the green lineage reveals the origin and diversification of land plants
Curr. Biol.
(2010) - et al.
The phylogeny of the Asteridae sensu lato based on chloroplast ndhF gene sequences
Mol. Phylogenet. Evol.
(2000) - et al.
Trees within trees: phylogeny and historical associations
Trends Ecol. Evol.
(1998) - et al.
The root of the mammalian tree inferred from whole mitochondrial genomes
Mol. Phylogenet. Evol.
(2003) - et al.
Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences
Bot. J. Linn. Soc.
(2000) - et al.
Evidence of chloroplast capture in South American Nothofagus (subgenus Nothofagus, Nothofagaceae)
Mol. Phylogenet. Evol.
(2010) - et al.
Phylogenetic analysis of asterids based on sequences of four genes
Ann. Mo. Bot. Gard.
(2001) - APG, 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants:...
- et al.
The age and diversification of the angiosperms re-revisited
Am. J. Bot.
(2010)