Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Plastid genome evolution in partially and fully mycoheterotrophic eudicots Darby, Hayley 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_february_darby_hayley.pdf [ 26.23MB ]
JSON: 24-1.0223133.json
JSON-LD: 24-1.0223133-ld.json
RDF/XML (Pretty): 24-1.0223133-rdf.xml
RDF/JSON: 24-1.0223133-rdf.json
Turtle: 24-1.0223133-turtle.txt
N-Triples: 24-1.0223133-rdf-ntriples.txt
Original Record: 24-1.0223133-source.json
Full Text

Full Text

PLASTID GENOME EVOLUTION IN PARTIALLY AND FULLY MYCOHETEROTROPHIC EUDICOTS by  Hayley Darby  B.A., Portland State University, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Botany)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2015  © Hayley Darby, 2015 ii  Abstract Plastid-genome evolution following photosynthesis loss is characterized by substantial change, contrasting with strong conservation in most photosynthetic land plants. Common features of reduced plastid genomes across diverse heterotrophic lineages point to a predictable trajectory of genome degradation, but this has been only partly tested. Here I document the molecular evolution of plastid genomes belonging to several mycoheterotroph lineages in Ericaceae, Gentianaceae and Polygalaceae, which include several independent origins of mycoheterotrophy in eudicot angiosperms that span different time scales since photosynthesis loss. I used next-generation and Sanger sequencing techniques to assemble complete plastomes or gene sets for comparative analyses of gene content and genome structure, and phylogenomic inference. I also sequenced several partially mycoheterotrophic and fully autotrophic relatives. Patterns of gene loss in mycoheterotroph plastomes are generally consistent with a previously hypothesized trajectory of change, starting with the loss of plastid NAD(P)H dehydrogenase before full loss of photosynthesis, and ending (here) with substantial reduction in genes involved in the translation apparatus and other nonphotosynthetic functions. Several retentions (delayed losses) of subunit genes for plastid-encoded polymerase, plastid ATP synthase and Rubisco are also consistent with hypothesized secondary (nonphotosynthetic) functions for these complexes. Two within-genus comparisons (for Epirixanthes in Polygalaceae and Voyria in Gentianaceae) demonstrate substantially different levels of genome degradation, consistent with heterogeneity in rates of genome change after a given origin of full mycoheterotrophy. Mycoheterotrophs in two families (Ericaceae, Polygalaceae) have extensive genome rearrangement compared to most land plants, contrasting with near colinearity in mycoheterotrophic members of Gentianaceae (despite sometimes extensive genome reduction in the latter). However, these contrasting patterns are iii  apparently not associated with transitions to mycoheterotrophy, as photosynthetic relatives in Ericaceae and Polygalaceae are also substantially rearranged—or with inverted repeat loss (evident in Epirixanthes pallida, Polygalaceae), as autotrophic Polygala retains its inverted repeats. Phylogenomic inferences of core eudicot phylogeny made using the retained genes are generally well supported and robust to a variety of phylogenetic approaches, and are also congruent with recent phylogenetic studies in each mycoheterotrophic family.     iv  Preface  All steps of this work were conducted predominantly by me, but with the assistance of others as follows: Vivienne Lam (University of British Columbia) was responsible for DNA extraction for four samples: Exochaenium oliganthum (Gentianaceae), Voyria clavata (Gentianaceae), and Voyria caerulea (Gentianaceae); Marybel Soto Gomez (University of British Columbia) prepared DNA libraries for two samples: Exochaenium oliganthum and Voyria clavata. Additional thanks are due to Vincent S.F.T. Merckx (Naturalis Biodiversity Center), G. Beatty and J. Provan (Queen’s University Belfast), J.R. Abbott (University of Florida), K. Neubig (University of Florida), S. Stefanovi! (University of Toronto, Mississagua), and R. Bertin (College of the Holy Cross) who kindly provided multiple plant samples as DNA or silica dried specimens. David Tack (University of British Columbia) and Daisie Huang (University of British Columbia) provided bioinformatics scripting assistance.      v  Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents .......................................................................................................................... v List of Tables ............................................................................................................................... vii List of Figures ............................................................................................................................. viii Acknowledgements ....................................................................................................................... x Chapter 1: Introduction ................................................................................................................................. 1 Chapter 2: Materials and Methods ................................................................................................................ 5 2.1 Taxonomic sampling ........................................................................................................................ 5 2.2 DNA isolation and library preparation ........................................................................................... 5 2.3 De novo contig assembly, plastid gene annotation and plastome reconstruction .......................... 6 2.4 Whole-plastome rearrangements ..................................................................................................... 8 2.5 Concatenated alignment construction ............................................................................................. 8 2.6 Phylogenetic inference .................................................................................................................. 10 Chapter 3: Results ....................................................................................................................................... 13 3.1 Plastome characteristics ............................................................................................................... 13 3.2 Gene content .................................................................................................................................. 16 3.3 Plastid phylogenomics of mycoheterotrophic eudicots ................................................................. 19 Chapter 4: Discussion .................................................................................................................................. 21 4.1 Plastid phylogenomics of eudicot mycoheterotrophs .................................................................... 21 4.2 Models of plastid genome degradation in heterotrophic plants .................................................... 22 4.3 Loss and retention of plastid gene products .................................................................................. 23 vi  4.4 Structural rearrangement and the inverted repeat ........................................................................ 37 4.5 Conclusion ..................................................................................................................................... 41 Bibliography ................................................................................................................................ 60 Appendices ................................................................................................................................... 79   vii  List of Tables Table 1 Specimen source information……………………………………………………43 Table 2 Plastid gene content across newly sequenced taxa of Gentianaceae, Polygalaceae,  and Ericaceae…………….………………………………………………………44 Table 3 Species with fully assembled plastomes…………………………………………46 Table 4 Inverted repeat boundary shifts in eudicot mycoheterotrophs and autotrophic relatives…………………………………………………………………………..47 Table S1 Accession information for publically available plastomes included in the angiosperm matrix………...……………………………………………………..79 Table S2 List of primer sequences used to close gaps and verify overlapping contigs……83 Table S3 Partitioning scheme, DNA substitution models and partition subsets resulting from partition-finder analyses………………………………………………..…103 Table S4 Species with partially assembled plastid genomes……………………………..109       viii  List of Figures Figure 1.  Linearized plastome maps of photosynthetic and mycoheterotrophic representatives of Ericaceae, Gentianaceae and Polygalaceae…………………..48 Figure 2.  Pairwise Mauve-based alignments of Nicotiana tabacum with autotrophic representatives of Polygalaceae, Gentianaceae and Ericaceae. ………….……...50  Figure 3.  Mauve-based alignments of Gentianaceae plastomes ……………………….…..52 Figure 4.  Mauve-based alignments of Polygalaceae and Ericaceae plastomes ...….…..…..54 Figure 5.  Angiosperm phylogeny inferred in a likelihood analysis of 82 plastid coding regions using the GxC partitioning scheme based on an ORF-only alignment; portion the tree showing rosid relationships……………………………………..56 Figure 6.  Angiosperm phylogeny inferred in a likelihood analysis of 82 plastid coding regions using the GxC partitioning scheme based on an ORF-only alignment; portion of the tree showing asterid relationships. …...…………………………..58 Figure S1.  Circular plastome map of Polygala arillata (Polygalaceae).…………………...110 Figure S2.  Circular plastome map of Epirixanthes pallida (Polygalaceae).……………….112 Figure S3.  Circular plastome map of Exacum affine (Gentianaceae). ……………………..114 Figure S4.  Circular plastome map of Exochaenium oliganthum (Gentianaceae).………….116 Figure S5.  Circular plastome map of Bartonia virginica (Gentianaceae).…………………118 Figure S6.  Circular plastome map of Obolaria virginica (Gentianaceae)..………………..120 Figure S7.  Circular plastome map of Voyria clavata (Gentianaceae) .….…………………122 Figure S8.  Linearized plastome map of the draft partial assembly of Epirixanthes elongata (Polygalaceae).…………………………………..………….…...….…..………124  ix  Figure S9.  Angiosperm phylogeny inferred in an unpartitioned likelihood analysis of 82 plastid genes based on an ORF-only alignment..………………..………….......126 Figure S10.  Angiosperm phylogeny inferred in a likelihood analysis of 78 translated plastid genes (ORF-only) using the gene partitioning scheme……………..…………..128  Figure S11.  Angiosperm phylogeny inferred in a parsimony analysis of 82 plastid coding regions based on an ORF-only alignment....………………..………….……….130 Figure S12.  Angiosperm phylogeny inferred in a likelihood analysis of 82 plastid genes that includes putative pseudogenes....…………....………………..………….……..132      x  Acknowledgements I would like to start by offering my thanks to my supervisor Dr. Sean Graham who gave me a challenging and stimulating research topic. With his guidance I have gained a deeper understanding of many aspects of plant evolution and phylogenetics. He has taught me to think critically and to value clearly communicated research.   I would also like to extend my appreciation to my committee members Dr. Mary Berbee and Dr. Jeanette Whitton. Their thought-provoking questions have directed me to explore ideas that I may not have otherwise considered.  I thank my current and former lab-mates, Marybel Soto Gomez, Qianshi Lin, Isabel Marques, David Bell, Wesley Gerelle, Vivienne Lam and Greg Ross who taught me laboratory protocols and analytical methods, and engaged me with exciting botanical discussions. Special thanks to Greg Ross who made preparing DNA libraries fun, to Marybel Soto Gomez for providing invaluable feedback in preparing for talks, and to Vivienne Lam for teaching me everything she knows about plastome evolution in peculiar plants.  Special thanks are owed to my parents, who encourage me to do pretty much whatever I want, academically and otherwise.   And thanks to my darling J.P. for making all of my breaks from school fun and for being my emissary while I’m away from home. Most of all, I thank him for his enduring support.  xi  To my parents, who gave me a spot in the garden.   1 Chapter 1: Introduction The plastid genome (plastome) of photosynthetic land plants is generally highly conserved in gene content and order, length and overall architecture (reviewed in Palmer, 1985; Wicke et al., 2011). It typically codes for ~110-120 unique genes, and its ~120-160 kb length is quadripartite in structure: a subset of duplicated genes are located in inverted-repeat (IR) regions of variable length across taxa that separate two asymmetrical single-copy regions. The latter regions are referred to as the large and small single copy (LSC and SSC) regions, respectively. Published plastome sequences of heterotrophic plants depart in some or all of these characteristics, in a lineage-dependent manner. For example, the plastomes of the mycoheterotroph Petrosavia stellaris (Petrosaviaceae) and the obligate holoparasite Conopholis americana (Orobanchaceae) have reduced length, gene content and an atypical gene order due to rearrangements (Wicke et al., 2013; Logacheva et al., 2014), while that of Sciaphila densiflora (Triuridaceae) is highly reduced in gene content while retaining nearly complete colinearity with its close photosynthetic relatives (Lam et al., 2015). Although they have heterogeneous patterns of gene loss and genome rearrangement, comparative analysis of genome evolution in different mycoheterotrophic lineages may allow us to make broad generalizations on the effect of photosynthesis loss on plastome molecular evolution (e.g., Barrett and Davis, 2012; Barrett et al. 2014). Mycoheterotrophy is a plant nutritional strategy that is distinct from direct plant parasitism, and is referred to as “full” mycoheterotrophy when photosynthesis has been lost. Fully mycoheterotrophic plants are completely dependent on fungal partners for their nutritional needs. Parasitic plants use haustoria to penetrate and parasitize the tissues of green plants, but mycoheterotrophs attract and consume fungal hyphae in modified root systems (Leake and Cameron, 2010; Merckx, Freudenstein, et al., 2013) The hyphae may belong to fungi involved in   2 mycorrhizal networks (these mycoheterotrophic plants thus indirectly parasitize the green-plant partners of mycorrhizal fungi), or in a few cases belong to saprophytic fungi (Bidartondo, 2005). Although relatively rare in terms of species number (less than 1% of land-plant species are full-blown heterotrophs), plant parasites and full mycoheterotrophs have evolved repeatedly across land-plant phylogeny (Merckx, 2013). There are 514 known species of fully mycoheterotroph plants, representing an estimated 46 or 47 independent losses of photosynthesis. Of these, a minimum of seven origins of full mycoheterotrophy (representing 47 species) are known in the core eudicots (Merckx et al., 2013a), where full mycoheterotrophy has evolved independently in three families (Ericaceae, Gentianaceae and Polygalaceae). In addition, partial mycoheterotrophs (plants that both photosynthesize and derive some nutrition from fungal partners) are known in Ericaceae, Gentianaceae and possibly also Polygalaceae (Tedersoo et al., 2007; Zimmer et al., 2007; Hynson et al., 2009; Cameron and Bolin, 2010; Merckx et al., 2013a). Mycoheterotrophic eudicots associate with arbuscular mycorrhiza-forming glomeromycete fungi in Polygalaceae and Gentianaceae, and ectomycorrhizal basidiomycete and ascomycete fungi in Ericaceae (Hynson and Bruns, 2009; Merckx, Freudenstein, et al., 2013).  With the exception of Ericaceae (see Braukmann and Stefanovi!, 2012), plastid genome evolution in eudicot mycoheterotrophs has not been explored. Examining independent losses of photosynthesis in these lineages of plants would be useful to more fully understand the breadth of plastome evolution in plants, and would provide counterpoints for recently published plastid genomes produced for monocot and liverwort mycoheterotrophs (Wickett et al., 2008; Delannoy et al., 2011; Logacheva et al., 2011, 2014; Barrett and Davis, 2012; Barrett et al., 2014; Lam et al., 2015; Schelkunov et al., 2015). Using evidence from published plastome sequences of heterotrophic plants and known functions of plastid genes, Barrett and Davis (2012) and Barrett   3 et al. (2014) proposed models for plastid genome degradation during or following the transition to a heterotrophic lifestyle. Their closely related ratchet-like models begin with the loss of plastid NAD(P)H genes, likely before the loss of photosynthesis in partial mycoheterotrophs, followed by concerted degradation of photosynthesis genes and the plastid-encoded RNA polymerase (‘PEP,’ which transcribes most photosynthesis genes, Hajdukiewicz et al., 1997; reviewed in Yagi and Shiina, 2014). Later-stage plastid genome gene loss or degradation apparently involves plastid ATP synthase loci (which appear to be retained after the initial loss of photosynthesis; Knauf and Hachtel, 2002; Wickett et al., 2008; Barrett et al., 2014; Logacheva et al., 2014), followed by genes involved in the plastid genetic apparatus and other non-photosynthetic functions. The degradation is ratchet-like because genes are assumed to not re-evolve once lost. Thus, the extent of degradation in mycoheterotrophs may correlate with the degree and recency of dependence on heterotrophic nutrition. The primary objective of my study is to survey plastid genome evolution in eudicot mycoheterotroph lineages, to use these new data in comparative analyses of gene content and genome structure, and for use in phylogenetic inference to place taxa in the context of core eudicot relationships. I used next-generation (NGS) and Sanger sequencing techniques to assemble complete circle plastomes for mycoheterotroph plants that represent three of the estimated seven origins of full mycoheterotrophy that have occurred in eudicots. Within Gentianaceae, I included Exochaenium oliganthum as an example of a recent loss of photosynthesis (estimated to have occurred within the last three million years, Merckx et al., 2013b). Chlorophyllous populations have also been reported for it (Kissling, 2012); chlorophyll retention has also been noted in full mycoheterotrophs such as Cymbidium macrorhyzon (Merckx et al., 2013a) and Corallorhiza spp.(Cummings and Welschmeyer, 1998; Barrett et al., 2014),   4 and does not necessarily reflect retention of photosynthesis. In contrast, loss of photosynthesis in Voyria dates to at least 31 million years ago, based on a crown-age dating for this fully mycoheterotrophic lineage (Merckx et al., 2013b). I included representatives of the single origin of full mycoheterotrophy in Polygalaceae, the exclusively non-photosynthetic genus Epirixanthes, which has an estimated crown age of 14 million years (Mennes et al., 2015b). I also included several partial mycoheterotrophs (species from two genera each in Gentianaceae and Ericaceae), and green relatives for all three families (published sequences for Ericaceae and new sequences in Gentianaceae and Polygalaceae) to provide close points of genomic comparison.  This sampling allowed me to explore plastome evolution over a range of different time scales, across taxa of different evolutionary histories and degrees of heterotrophy, and involving homologous (within-genus) and non-homologous losses of photosynthesis (losses between genera here). I used these new data to address the following specific questions: (1) Do plastid genomes evolve in a predictable manner after the transition to heterotrophy, as proposed by Barrett and Davis (2012) and Barrett et al. (2014)? (2) Do we see any unexpected retention of photosynthetic genes in full mycoheterotrophs, which I think point to secondary functions for them in the plastid? (3) Are plastid genes retrieved from heterotrophs useful in plastid-genome scale phylogenetic inference? (4) What (if any) structural rearrangements are associated with the origins of mycoheterotrophy in these taxa?   5 Chapter 2: Materials and Methods  2.1  Taxonomic sampling I sampled two species representing the single origin of full mycoheterotrophy in Polygalaceae (Epirixanthes), three species representing two of the estimated four origins in Gentianaceae (Exochaenium and Voyria), and four partially mycoheterotrophic species, two each from Ericaceae and Gentianaceae. I also sampled at least two putatively fully autotrophic taxa in Polygalaceae (Polygala and Salomonia), an autotrophic Gentianaceae (Exacum), and included several publicly available plastid genomes of autotrophic members of Ericaceae, allowing comparisons between heterotrophic and autotrophic relatives in each case (see Table 1). The full taxon sampling includes sequences from 69 taxa retrieved from GenBank and 91 from the larger matrix presented in Ruhfel et al. (2014), and represents multiple lineages of monocots, magnoliids and other angiosperms (Amborellales, Nymphaeales, Austrobaileyales). My taxon sampling within eudicots includes a single representative for each available family across the core eudicots, with denser sampling in lineages that are more closely related to Polygalaceae, Gentianaceae and Ericaceae (Table S1). It also includes all available eudicot plastid genomes from heterotrophs (parasitic plants belonging to Orobanchaceae, Convolvulaceae and Santalales) and carnivorous plants (members of Lentibulariaceae).  2.2 DNA isolation and library preparation I prepared sampled species for whole-genome shotgun-sequencing on the Illumina HiSeq 2000 platform (Illumina, Inc., San Diego, USA) to retrieve complete plastid genome sequences. I first extracted DNA from silica-dried tissue samples using the method of Doyle and Doyle (1987).   6 Several samples (Orthilia secunda, Pyrola minor, Epirixanthes elongata and Salomonia cantoniensis) were provided by collaborators as DNA extractions. I prepared genomic DNA libraries using three kits (KAPA Library Preparation Kit, KAPA Biosystems, Wilmington, USA; Nugen Ovation Ultralow Library systems, NuGEN, San Carlos, USA; Bioo NextFlex Rapid sequencing kit, Bioo Scientific, Austin, USA), following manufacturer protocols for each kit, using genomic DNA sheared to 400 bp fragments with a Covaris sonicator (model: S220, Woburn, Massachusetts, USA) as a starting point. I confirmed that the libraries met a minimum concentration of 0.5 ng/ul and were in a 500-600 bp size range, by using a Qubit fluorometer (Qubit 2.0 Fluorometer, Life Technologies, Thermo Fisher Scientific, Waltham, USA) and Bioanalyzer (2100 Bioanalyzer, Agilent Technologies, Santa Clara, United States), respectively. Sample concentrations were then quantified on an iQ5 real-time qPCR system (Illumina DNA standard kit, KAPA Biosystems, Boston, USA; Bio-Rad Laboratories, Inc., Hercules, USA) and sequenced as 100 bp paired-end reads, on multiplexed Illumina runs (Cronn et al., 2008) that included 10 to 39 samples per lane.    2.3 De novo contig assembly, plastid gene annotation and plastome reconstruction The multiplexed Illumina sequence reads were sorted by taxon using CASAVA 1.8.2 (Illumina Inc., San Diego, California, USA). I performed de novo assemblies on each sample using CLC Genomics Workbench v. 6.5.1 (CLC bio, Aarhus, DK), selecting all contigs larger than 500 bp and at least 10X coverage. I then used a custom Perl script (Daisie Huang, University of British Columbia) to BLAST (Altschul et al., 1990) contigs against local databases of three reference plastomes (Gentianales: Asclepias syriaca, NC_022432.1; Fabales: Glycine max, NC_007942.1; Ericacales: Arbutus unedo, JQ067650), in order to identify and remove mitochondrial and   7 nuclear contigs. For Pyrola minor and Voyria caerulea, I annotated and isolated individual plastid genes in CLC-produced contigs using DOGMA (Wyman et al., 2004), manually inspecting gene and exon boundaries in Sequencher 4.2.2. (Gene Codes Corporation, Ann Arbor, US) using Arbutus unedo (JQ067650) or Asclepias syriaca (NC_022432.1) as reference sequences, respectively. For all other taxa, I assembled CLC-produced contigs into full or nearly full circular plastomes, by bridging gaps and confirming contig overlap using Sanger-based DNA sequencing. I designed custom primers for amplification and Sanger sequencing using Primer3 (Untergrasser et al. 2007; Koressaar and Remm 2007) (see Table S2 for primer sequences), performing amplifications using Phusion High-Fidelity DNA Polymerase (Thermo Fisher Scientific, USA) and sequencing using BigDye Terminator v.3.1 (Applied Biosystems, Inc. Foster City, USA). I performed amplification following the general methodology in Graham and Olmstead (2000) with minor modifications: (1) initial denaturation at 98° C for 5 min; (2) 40 cycles of the following: denaturation at 98° C for 20 s, annealing at 60° C for 30 s, extension at 72° C for 2 min; (3) final extension at 72° C for 5 min. For cycle sequencing, I followed the methodology in Graham and Olmstead (2000) for 25 cycles, with some modifications: (1) denaturation at 96° C for 10 s; (2) annealing at 50° C for 5 s; (3) extension at 60°C for 4 min. Sequencing reactions were run on an Applied Biosystems 3730S 48-capillary DNA analyzer (Applied Biosystems, Inc., Foster City, USA). I produced final whole or partial plastome sequences by assembling Illumina contigs and Sanger sequences in Sequencher 4.2.2, and deduced and annotated gene and exon boundaries using DOGMA and Sequencher, using Asclepias syriaca, Asclepias nivea (NC_022431), Glycine max (NC_007942.1) or Arbutus unedo as reference sequences. I used OGDRAW (Lohse et al., 2013) to prepare plastome figures.    8 2.4 Whole-plastome rearrangements I used Mauve 3.2.1 (Darling et al., 2004) to predict gene-order rearrangement in the plastomes of mycoheterotrophs with respect to photosynthetic relatives, omitting the second copy of the inverted repeated for these analyses. This program identifies regions of homology shared between at least two sequences in an alignment (called locally colinear blocks; LCBs) using a combination of string-matching, local alignment and breakpoint analysis, and positions LCBs using progressive alignment (CLUSTALW; Thompson et al., 1994). Minimum string lengths (‘seeds lengths’) and LCB calculation parameters can be optimized by the user: I used a seed length of 21 bp to minimize spurious matches, and allowed minimum LCBs to be calculated automatically.  2.5 Concatenated alignment construction I performed alignments on individual genes, excluding introns and intergenic regions (and initially included pseudogenes, see below) to prepare a final fully concatenated matrix. I compiled the plastid gene sets I generated (Table 1) with a set of taxa chosen from a publicly available green-plant-wide matrix (Ruhfel et al. 2014) and plastid-genome sequences available from GenBank (Supplementary Table S1). To do this I exported new sequences in FASTA format, and generated single-gene, multi-taxon files using custom Python scripts (Dave Tack, University of British Columbia). These files represent 78 protein-coding and four ribosomal DNA (rDNA) loci. Missing genes for individual taxa (see below) in individual alignments were represented as blanks. The protein-coding set of genes includes the loci typically present in angiosperms, but I excluded ycf1 due to alignment difficulty. For each gene I produced automated DNA sequence alignments using MAFFT (Katoh et al., 2002), inspecting the output   9 and manually adjusting it where necessary, following alignment criteria laid out in Graham et al., (2000). I performed these alignment steps (automated alignment and manual adjustment) using Mesquite v. 3.03+ and v. 3.4 (Maddison and Maddison, 2014, 2015). I used the default settings for MAFFT, although for the gene ycf2 I used the ‘linsi’ option, a more computationally intensive and thorough search approach. I removed introns from split genes, and staggered difficult-to-align regions, as described in Saarela and Graham (2010). Genes obtained from the Rufhel et al. (2014) matrix were pre-trimmed in various ways (i.e., at their 5’- and 3’-ends, and for introns and poorly-aligned regions).  I combined these individual gene alignments into a single concatenated matrix, and prepared two versions of it. One version excluded all or most pseudogenes (see below). This combined ‘ORF-only’ matrix (ORF = open reading frame) comprised 81,732 bp (for reference, derived from 57,507 bp sequence data in Exochaenium oliganthum). I also translated the 78 protein-coding genes in the ORF-only matrix using Mesquite, and constructed a concatenated 25,528 amino-acid residue matrix from this. For newly sequenced taxa, the ORF-only concatenated matrix included four genes with a single reading frame interruption compared to reference taxa (Table 2), which I retained as they may reflect sequencing errors or RNA edit sites e.g.(e.g. Freyer et al., 1997; Kugita et al., 2003; Hoffmann et al., 2009); thus, this matrix may include several genes with recent loss of function. However, in three of these four genes, other subunit genes have multiple reading frame interruptions; and so a more likely situation is that there is a lag in the accumulation of reading frame interruptions in the subunits with only a single interruption. I therefore retained a version of the concatenated matrix that included these and other more obvious pseudogenes. I used this to assess whether their inclusion had an effect on phylogenetic inference. Where the 5’- or 3’-end of a putative pseudogene was not readily   10 alignable, I trimmed this portion from the alignment. I based the pseudogene status of published sequences on their respective GenBank annotations (Table S1). This 84,567 bp matrix that included pseudogenes was derived from 66,820 bp sequence data for Exochaenium oliganthum, for reference. To ensure that copy-paste or other editing errors were not introduced during data compilation or manual alignment adjustment, I examined the DNA matrix using the following approaches. I excluded all taxa except those retrieved from the Ruhfel et al. (2014) matrix (91 taxa remained) and re-aligned these sequences using MAFFT against all sequences for the corresponding taxon set in the original matrix, for all genes simultaneously (distinguishing realigned and original data in the taxon names). I then ran a heuristic parsimony analysis of this 182-taxon matrix. I consistently found that the original and realigned sequences were sister taxa, and had no differences in terminal branch length between them. For all other taxa, I exported concatenated gene sequences for each taxon and used Sequencher 4.2.2. (Gene Codes Corporation, Ann Arbor, US) to compare them to the original individual taxon files. No obvious editing errors were found using these two error-checking methods.   2.6 Phylogenetic inference I analyzed the ORF-only data using maximum likelihood and parsimony methods. I ran a heuristic parsimony search in PAUP version 4.0a145 (Swofford, 2003), using tree-bisection-reconnection (TBR) branch swapping, with 10 random stepwise addition replicates, and holding one tree at each step. I performed several different ML searches using RAxML v. 7.4.2 (Stamatakis, 2006), conducting 20 independent searches for the best tree in each case. I also performed bootstrapping analyses to assess the strength of branch support for trees (Felsenstein,   11 1985). For the parsimony analysis, I ran 500 bootstrap replicates with 10 random stepwise addition replicates. For the ML analyses, I ran 500 rapid bootstrap replicates (Stamatakis et al., 2008) using the same DNA or amino-acid substitution models and partitioning schemes used in searches for the best tree (see below). I considered branches with 95% or better bootstrap support as well-supported, and branches with <70% bootstrap support poorly-supported, following Zgurski et al. (2008). All nucleotide ML analyses were performed on the CIPRES portal (Miller et al., 2010). The amino-acid analysis was performed using the RAxML graphical front-end interface (Silvestro and Michalak, 2012). For DNA-based ML analyses, I ran both unpartitioned and partitioned analyses. The latter considered codon positions within each protein-coding gene (‘GxC’ or gene by codon partitioning scheme). To decide on the partitions for the GxC analysis, I designated an initial 238 partitions for the concatenated matrix (derived from the first, second or third codon positions of each protein-coding gene, and four unique partitions representing the four rDNA loci). I then used PartitionFinder version 1.1.1. (Lanfear et al., 2012) to pool partitions that did not have significantly different substitution models or model parameters using the Bayesian Information selection criterion (BIC). For this analysis, branch lengths were linked and only the substitution models implemented in RAxML were explored using the relaxed hierarchical clustering algorithm, as described in Lanfear et al. (2014). I searched the top 5% of schemes expected to improve likelihood scores. I ran a partition-finder analysis of the version of the concatenated matrix with pseudogenes included, in the same manner. I also ran a partition-finder analysis for the concatenated amino-acid matrix using PartitionFinderProtein version 1.1.1 (Lanfear et al., 2012), starting with the 78 protein-coding genes, and otherwise using the settings described above. The ORF-only DNA matrix yielded a partition-scheme with 64 final partitions, and   12 recovered the GTR+ " or GTR+I+" DNA substitution models as the best fit for individual data partitions (Table S3). The version of the concatenated matrix that included obvious pseudogenes yielded a partition-scheme with 67 final partitions, and recovered the GTR+ " or GTR+I+" DNA substitution models as the best fit for individual partitions. PartitionFinder also identified GTR+ " as the best DNA substitution model for the unpartitioned ML analysis of the matrix. The partition-finder analysis of the amino-acid matrix found 37 partitions, with best models that included variants of the JTT, JTTF, CPREV, MTMAM or LG substitution models (see Table S3 final data partitioning schemes). I applied the optimal models for each data partition in the various partitioned likelihood analyses.   13 Chapter 3: Results 3.1 Plastome characteristics I assembled complete, circular plastome sequences for seven species, including three full mycoheterotrophs (Exochaenium oliganthum, Voyria clavata, and Epirixanthes pallida), two partial mycoheterotrophs (Bartonia virginica and Obolaria virginica) and two autotrophs (Exacum affine and Polygala arillata) (Table 3, Figs. S1-S7). I also recovered a nearly complete assembly for a partial mycoheterotroph (Orthilia secunda, which likely has only a single gap; Fig. 1, Table S4). Four others are presented here only as gene sets based on more incomplete assemblies, including two full mycoheterotrophs (Voyria caerulea and Epirixanthes elongata), one partial mycoheterotroph (Pyrola minor) and an autotroph (Salomonia cantoniensis) (Table S4).  3.1.1 Polygalaceae The largest new plastome belongs to Polygala arillata (Polygalaceae) (Table 3), with a length of 164,747 bp. It also has the largest inverted repeat (IR) region among those recovered here (36,168 bp, comprising 23 genes that extend from rpl2 to ndhI; Fig. 1, S1, Table 4). For comparison, the IR region of Exacum affine (in Gentianaceae) comprises 20 genes and is 26,239 bp in length, spanning from a point 300 bp into rps3 to 1,086 bp into ycf1. The plastome of Epirixanthes pallida is intermediate among the full mycoheterotrophs presented here in terms of plastome length and gene content. Epirixanthes pallida is the sole fully assembled species that has lost an IR; however, it retains two ~12 kb direct repeats composed of genes found in the IR of P. arillata (Figs. 1, S2). A partial assembly for Ep. elongata suggests that it has a very reduced plastome and several repeated regions based on depth of sequencing (Fig. S8). Sectors   14 of the assembly had read depth varying eight-fold: the lowest coverage contig (~200X read depth) includes loci for trnE-UUC,  trnY-GUA and matK, and the highest coverage contigs span the rDNA operon (~1700X read depth). A partial assembly of Salomonia cantoniensis (an autotrophic member of Polygalaceae; not shown) is consistent with it having a quadripartite structure. Polygala arillata and S. cantoniensis have three copies of trnQ-UUG. These disjunct genomic locations may have resulted from a translocation or a series of inversions, as a single copy is found adjacent to the RNA polymerase operon in the large single copy region (the ancestral arrangement) and two copies are located in the inverted repeat regions; two copies are found in Ep. pallida, one in each direct repeat, and the LSC copy of the gene is not present (Figs. 1, S2).   3.1.2 Gentianaceae The plastome of the full mycoheterotroph Exochaenium oliganthum is comparable in size to autotrophic Exacum affine, and is slightly larger than the plastomes of the two partial mycoheterotroph species, Bartonia virginica and Obolaria virginica. It also retains more genes with intact open reading frames than the latter two species (Table 3). Bartonia virginica and O. virginica have substantially smaller small single copy (SSC) regions than Exa. affine and Exo. oliganthum, which may be attributed to gene loss and shifts in SSC/IR boundaries (Table 3, 4 Figs. 1, S3-S6). The smallest, fully assembled plastome I recovered in the current study belongs to the fully mycoheterotrophic Voyria clavata, which is 31,724 bp in length and has 25 unique genes with uninterrupted reading frame, specifically four rDNA genes, four tRNA genes and 17 protein coding genes (Table 3). This species retains as single-copy genes 13 of the 20 genes   15 found in the IR of Exa. affine, but has a novel IR region corresponding to a block of five genes located in the large single copy (LSC) of Exa. affine (Figs. 1, S7).    3.1.3 Ericaceae Although incomplete, the plastome of the partial mycoheterotroph Orthilia secunda appears to be comparable in length and gene content to partial mycoheterotrophs in Gentianaceae (Table S4, Fig. 1). The partial assembly of O. secunda is consistent with a quadripartite structure, and it retains a larger SSC region than Arbutus unedo (Fig. 1, Table 4).   3.1.4 Mauve-based inferences of genome rearrangement Gene order in autotrophic relatives of mycoheterotrophs is modified in Polygalaceae (Polygala arillata and Salomonia cantoniensis, the latter based on incomplete assemblies; not shown) and Ericaceae (Arbutus unedo; Martínez-Alberola et al., 2013), compared to the putative ancestral angiosperm gene order (Jansen et al., 2007). The latter order is represented here by Exacum affine (Gentianaceae) and tobacco (Nicotiana tabacum, NC_001879), which have the same gene order (Fig. 2). Ignoring often substantial deletions, this ancestral gene order has been largely conserved in the fully mycoheterotrophic Gentianaceae examined here (Voyria and Exochaenium; Fig. 3). Using Mauve alignment, I identified three colinear blocks among Gentianaceae sequences, comprising the large single copy (LSC) region through ycf2 in the inverted repeat (IR), the IR region after ycf2 through ndhD in the small single copy (SSC), and the rest of the SSC, respectively. There are no rearrangements (which would generally appear as crossed lines in the figure) apart from simple inversions. In comparison to gene order in Exacum, two colinear blocks are homologous but in reverse orientation: a block composed of IR genes is   16 reversed in Voyria clavata, and in Bartonia there is an inverted three-gene block in its contracted SSC (Fig. 3).  The mycoheterotrophic Polygalaceae and Ericaceae are substantially more rearranged (Fig. 4). Eleven and thirteen colinear blocks are identified in Polygalaceae and Ericaceae Mauve-based alignments of full mycoheterotrophs compared to their autotrophic relatives, respectively. In Polygalaceae, rearrangements and inversions are distributed across the plastome of Epirixanthes pallida. A ~12 kb region of the IR in P. arillata is directly repeated in Ep. pallida (‘b’ in Fig. 4). Rearrangements are concentrated in the LSC in Ericaceae; the IR of O. secunda is a single colinear block. What is reconstructed as an inversion in the SSC of Orthilia secunda may be better accounted for as an expansion of IR into the SSC in Arbutus unedo (see Fig. 1).   3.2 Gene content Gene retentions, losses and putative pseudogenizations for protein-coding loci are discussed in more detail below for each family.  3.2.1 Polygalaceae Most genes coding for subunits of the plastid NAD(P)H complex, photosystems I and II, and cytochrome b6/f complex are lost or interrupted in the plastome of Epirixanthes pallida, but plastid-encoded subunits of the ATP synthase complex and rbcL (which codes for the large subunit of Rubisco) have been retained (Table 2). All genes of the plastid-encoded RNA polymerase (PEP) are interrupted by premature stop codons. Epirixanthes pallida retains all 30 plastid-encoded transfer RNA genes. Although not a complete circle, I recovered 21 unique genes with uninterrupted reading frame in the assembly of Ep. elongata, specifically three   17 rDNA, five transfer RNA and 13 protein coding (Table 2, 3). I also did not retrieve two rDNA loci in the gene set of Salomonia cantoniensis. Loss of rDNA loci is not documented in any plant, regardless of trophic status, so I presume these taxa retain these small genes but that they were not assembled into the Illumina contigs. The gene coding for the ATP-dependent caseinolytic protease (clpP), a small subunit ribosomal protein (rps16) and a large subunit ribosomal protein (rpl22) have been deleted in the plastomes of Polygala arillata and Ep. pallida, and translation initiation factor A (infA) is a pseudogene in both species (Table 2). These genes were not recovered in the gene sets of Salomonia cantoniensis or Ep. elongata. It is not clear if accD, the gene that codes for the beta subunit of acetyl-CoA carboxylase, is retained in Polygalaceae. I recovered a ~1200 bp open reading frame in P. arillata, S. cantoniensis and Ep. pallida, and a ~500 bp truncated putative pseudogene in Ep. elongata: accD lacks introns, and is ~1400-1600 bp in Gentianaceae and Fabaceae, for comparison. BLAST searches using the 1395 bp accD locus from Ceratonia siliqua (Fabaceae, NC_026678) matched only to subregions of the intact reading frame in P. arillata, S. cantoniensis and Ep. pallida, recovering matches for 49%, 52% and 46% of the query length, respectively. Protein-translated BLAST searches yielded similar match lengths.   3.2.2 Gentianaceae The autotrophic Exacum affine retains open reading frames for all loci typically found in angiosperm plastomes (Table 2). All photosynthesis-related genes have been deleted in the plastome of Voyria clavata, except for a truncated rbcL pseudogene (Table 3). Although not a complete circle, all photosynthesis genes retrieved from the assembly of Voyria caerulea (nine genes) have interrupted reading frames except for a single locus encoding a subunit of the ATP   18 synthase complex. Voyria clavata retains four transfer RNA loci, and I recovered 13 transfer RNA loci in the V. caerulea gene set. MatK, which codes for the group IIa intron maturase (MATK) is also deleted from the V. clavata plastome, and I did not recover it in the partial assembly of V. caerulea. The two Voyria plastomes do, however, retain several intact genes with group IIa introns (i.e., clpP, rpl2, and rps12) (Table 2). There are no gene deletions in the plastome of Exochaenium oliganthum, but there are reading frame interruptions in several genes that code for key components of photosystems I and II (Table 2). These include psaA, which has no detectable start codon and multiple premature stop codons, and psbA, which has a single nucleotide deletion resulting in a frame shift. The third exon of the photosystem I assembly protein, ycf3, is also deleted in this species, and there are multiple reading-frame interruptions in the sequence of ccsA. The plastid NAD(P)H-dehydrogenase (ndh) loci in the full mycoheterotrophs and in the two partial mycoheterotroph species, B. virginica and O. virginica: all have interrupted reading frames or deletions, for at least some of the genes (Table 2). Genes related directly to photosynthesis (photosystems I and II, the cytochrome b6/f complex, rbcL and ATP synthesis) all have intact reading frames in the two partial mycoheterotroph species, with two exceptions in O. virginica. First, the gene coding for a component of photosystem II, psbM, is deleted in this species, and second, the c-type cytochrome biogenesis protein, ccsA, may also be a pseudogene for it, as it has a single base deletion resulting in a frame shift.    3.2.3 Ericaceae  Most ndh genes are interrupted by premature stop codons and non-triplet indels in Orthilia secunda and Pyrola minor. All other photosynthesis genes are retained with uninterrupted reading frames. It is not clear whether accD has been retained in O. secunda and P. minor (accD   19 is a pseudogene in the autotroph Arbutus unedo, for reference; Table 2). I recovered what appears to be a ~850 bp fragment of the 3’ end of accD with an uninterrupted reading frame in O. secunda, and a ~1,600 bp ORF in P. minor. BLAST searches using the 1542 bp accD locus from a close relative in which accD is clearly retained (Camellia crapnelliana, NC_024541.1) match only to subregions of the intact reading frame in O. secunda and P. minor, recovering matches for 64% and 34% of the query length, respectively. Protein-translated BLAST searches yielded similar match lengths. ClpP, a pseudogene in A. unedo, was not recovered from P. minor or O. secunda.     3.3 Plastid phylogenomics of mycoheterotrophic eudicots I inferred no major topological differences in core eudicot relationships across the various analyses (Figs. 5-6, S9-S12). Ericaceae, Gentianaceae, and Polygalaceae comprised monophyletic lineages in all phylogenetic analyses, with consistently strong bootstrap support (Figs. 5-6, S9-S12). Within Polygalaceae, I recovered a clade comprising Epirixanthes and Salomonia as the sister group of Polygala, with strong support across all analyses (Figs. 5, S9-S12). A clade comprising Polygalaceae and Fabaceae, the only representatives of Fabales here, was recovered with strong support in all analyses (Figs. 5, S9-S12).  In Gentianaceae, Exochaenium and Exacum are inferred to be sister taxa, the partial mycoheterotrophs Obolaria and Bartonia are sister groups, and the two species of Voyria also formed a clade, all with strong support (Figs. 6, S9-S12). In the ORF-only ML analyses, I inferred Exochaenium-Exacum to be the sister group of Obolaria-Bartonia, but with poor support (Figs. 6, S9-S11). In the ML analysis that included obvious pseudogenes, Exochaenium-Exacum is the sister group of a clade comprising Obolaria-Bartonia and Voyria, with strong   20 support (Fig. S12). I recovered two equally parsimonious trees that differed in whether Voyria or Exochaenium-Exacum was the sister-group to Obolaria-Bartonia, and these relationships collapsed in the strict consensus (Figs. S11). The order Gentianales is monophyletic (considering the three of five families included here): Rubiaceae were inferred to be the sister group to Gentianaceae and Apocynaceae at the current taxon sampling, with strong support across analyses (Figs. 6, S9-S12).  Within Ericaceae, Pyrola and Orthilia are consistently strongly supported as sister groups across all analyses (Figs. 6, S9-S12), as are Rhododendron and Vaccinium. The position of Arbutus differed between the DNA and amino-acid based analyses. In the DNA-based analyses, Pyrola-Orthilia is the sister group of Arbutus, an arrangement with moderate to strong bootstrap support (98-100% ML; 81% for parsimony), and this overall clade is the sister group of a clade comprising Rhododendron and Vaccinium, also with strong support (Figs. 6, S9, S11-S12). In contrast, in the amino-acid based likelihood analysis, Arbutus is instead inferred to be the sister group of a clade comprising Pyrola-Orthilia and Rhododendron-Vaccinium, although this arrangement had poor support (Figs. S10). The order Ericales is inferred to be monophyletic (with only four of ~20 families sampled) with strong support across analyses (Figs. 6, S9-S12). Within Ericales, I recovered a clade comprising Ericaceae and Actinidiaceae as the sister group to Theaceae at the current taxon sampling, with Primulaceae then the sister group of the clade formed by those three families; this arrangement had strong support across analyses (Figs. 6, S9-S12).    21 Chapter 4: Discussion  4.1 Plastid phylogenomics of eudicot mycoheterotrophs Plastid genomes have only recently begun to be used for phylogenetic inference with full mycoheterotrophs, because it was assumed that too many genes (or the entire genome) would be lost to allow this, or that retained genes would be evolving too rapidly (e.g., Cronquist, 1988, p 467; Merckx et al., 2009). Rate elevation can be problematic if it leads to long-branch attraction in phylogenetic inference (Felsenstein, 1978; Hendy and Penny, 1989). I did not perform a formal rate analysis here, although the mycoheterotrophs examined here appear to have comparable rates of evolution to other eudicots, or moderately elevated rates (based on visual comparison of branch lengths to their sister groups, and to other close green relatives in the same or related orders of eudicots; Figs. 5, 6, see also Figs. S9-S12). Recent phylogenetic studies using retained plastid gene sets demonstrate that even highly reduced and rapidly evolving plastid genomes allow inferences of phylogenetic relationships for mycoheterotrophs that are well supported and consistent with studies based on genes from mitochondria or the nucleus (e.g., for Corsiaceae and Triuridaceae; Lam et al., 2015; Mennes et al., 2015a). The phylogenetic inferences made here are congruent with other studies using non-plastid data (for Ericaceae, Kron et al., 2002, Braukmann and Stefanovi!, 2012; for Gentianaceae, Merckx et al., 2013b; for Polygalaceae, Bello et al., 2012, Mennes et al., 2015b) where there are overlapping sets of taxa, and disagree only where one or both studies have poor branch support, as with the sister group of Pyroleae (Pyrola and Orthilia here) in Ericaceae (Bidartondo and Bruns, 2001; Kron et al., 2002; Braukmann and Stefanovi!, 2012), and the family-level arrangement of Exaceae (represented by Exacum and Exochaenium) versus Voyrieae (represented by Voyria) (Merckx et al., 2013b). In   22 addition, my phylogenetic inferences are generally not affected by the use of different phylogenetic criteria (parsimony and likelihood), the use of partitioned vs. unpartitioned likelihood analysis, or by the use of DNA vs. amino-acid substitution models (Figs. 5-6, S9-S12). It also does not seem to matter whether pseudogenes are included in analysis or not (cf. Figs. 5-6, S9-S11 and S12), although in a few cases my data resolve relationships that are unclear elsewhere, such as whether Exaceae or Voyrieae is the sister group to Gentianeae (Figs. 6, S9-S12). A clade comprising Exaceae and Gentianeae was recovered as the sister group to Voyrieae for all analyses where pseudogenes were excluded, but with weak support (Figs. 6, S9-S11). The partitioned analysis that included pseudogenes resolved Exaceae as sister-group to Gentianeae and Voyrieae with strong support (Fig. S12), which is congruent with inferences made by Merckx et al. (2013b) using a non-plastid data set.  4.2 Models of plastid genome degradation in heterotrophic plants  As the need to acquire nutrition via photosynthesis declines and ceases in heterotrophs, purifying selection to maintain genes with protein products involved in the photosynthetic apparatus should be relaxed and eventually released. Barrett and Davis (2012) and Barrett et al. (2014) developed two closely related models of plastid genome evolution in heterotrophic plants that predict an ordered series of plastid gene loss and genome reduction, proposing that the extent of plastome reduction is correlated with the degree and recency of dependence on non-photosynthetically derived nutrition. Once photosynthetic function is lost (and photosynthesis genes begin to be lost or degraded), other associated genes may follow, such as the plastid-encoded RNA polymerase genes that are thought to be necessary for photosynthesis-related gene expression (Hajdukiewicz et al., 1997; Zhelyazkova et al., 2012). The most reduced plastomes   23 may eventually begin to lose genes with roles in the plastid genetic apparatus and other non-photosynthetic metabolic (‘housekeeping’) functions, as the importance of the plastid organelle for plant survival diminishes. Eventually, most housekeeping functions may be lost, streamlined or replaced by analogous functions provided by non-homologous genes residing in other genomic compartments (or homologous but successfully transferred genes), with only a core of genetic apparatus genes retained in the service of residual but essential non-photosynthetic plastid-encoded genes (e.g., Barbrook et al., 2006; Delannoy et al., 2011).  The partial and full mycoheterotrophs that I sequenced in the eudicots display nearly the full range of large-scale genome modifications. Two partially mycoheterotrophic taxa (Obolaria and Bartonia, Gentianaceae) have nearly full-sized plastid genomes (~146 kb) but have extensive degradation in NAD(P)H dehydrogenase genes, which has apparently also happened in the two partially mycoheterotrophic Ericaceae based on the plastid gene sets that I was able to recover (Tables 2, 3). Focusing on the full mycoheterotrophs, Exochaenium oliganthum (Gentianaceae) has a genome size typical of green plants (~151 kb) with minimal detectable pseudogenization, Epirixanthes pallida (Polygalaceae) has a more reduced genome (~94 kb) with nearly all genes with protein products involved in the light reactions of photosynthesis either degraded or lost but many still retained as pseudogenes, and Voyria clavata (Gentianaceae) has a substantially truncated genome (~31 kb), with nearly all photosynthesis and many housekeeping genes also lost (Tables 2, 3).   4.3 Loss and retention of plastid gene products Below I briefly discuss the significance of gene losses and retentions in these different lineages in terms of the protein complexes and other gene products that the plastid loci code for. I present   24 each in the approximate order of loss proposed by the Barrett and Davis (2012) and Barrett et al. (2014). It should be noted that the fully mycoheterotrophic lineages examined here likely each represent evolutionarily independent losses of photosynthesis, at least concerning the genera Exochaenium and Voyria (both Gentianaceae) and Epirixanthes (Polygalaceae) (Merckx and Freudenstein, 2010; Merckx et al., 2013a; Merckx et al., 2013b; Mennes et al., 2015a)  . However, I also examined two species each in two fully mycoheterotrophic genera, Voyria (Gentianaceae) and Epirixanthes (Polygalaceae). In both genera, it is most parsimonious to assume that the two species in them are the result of a common loss of photosynthesis (one loss in each genus), as all other species in each genus are also fully mycoheterotrophic (Merckx et al., 2013a). Therefore, both genera provide an opportunity to examine the different rates and possibly different routes of genome degradation that follow an homologous origin of heterotrophy, as has been done elsewhere for Orobanchaceae (Wicke et al., 2013), Epipogium (Schelkunov et al., 2015). and Corallorhiza (Barrett et al., 2014) (the latter are distinct lineages of full mycoheterotrophs in Orchidaceae). In addition, Bartonia and Obolaria in Gentianaceae, and Orthilia and Pyrola in Ericaceae, each provide examples of pairs of related taxa in the early stages of mycoheterotrophy (all four taxa appear to be both photosynthetic and partially mycoheterotrophic based on isotopic evidence; see Cameron and Bolin, 2010 for Gentianaceae, and Tedersoo et al., 2007 and Zimmer et al., 2007 for Ericaceae). It is not known whether partial mycoheterotrophy is homologous within each of these pairs, although Bartonia and Obolaria may both be closely related to each other within Gentianaceae (both belong to subtribe Swertiinae; Struwe, 2014), as are Orthilia and Pyrola in Ericaceae (both belong to tribe Pyroleae; Kron et al., 2002). .    25 4.3.1 Loss of NAD(P)H dehydrogenase in early-transitional mycoheterotrophs The plastid NAD(P)H dehydrogenase complex is associated with cyclic electron transport and is thought to provide protection from photooxidative damage (Martín and Sabater, 2010; Shikanai, 2015). The complex may be nonessential or less essential in the absence of environmental stress (e.g. light, nutrient or CO2; Peltier and Cournac, 2002). A functional plastid NAD(P)H complex would not be needed in non-photosynthetic plants, and so it is not surprising that it is functionally lost in all full mycoheterotrophs (e.g., Table 3). However, all plastomes of partially heterotrophic plants (hemiparasites and partial mycoheterotrophic) sequenced to date also exhibit pseudogenization or loss of all or some ndh genes, supporting the loss or at least non-functionality of the NAD(P)H dehydrogenase complex. These include photosynthetic hemiparasites in Convolvulaceae (Funk et al., 2007; McNeal et al., 2007), Santalales (Petersen et al., 2015), and Orobanchaceae (Wicke et al., 2013), and partial mycoheterotrophs in Orchidaceae (Zimmer et al., 2008; Barrett et al., 2014). For mycoheterotrophs, the commonality of this loss in both partially and fully heterotrophic taxa led to the hypothesis that ndh genes are the initial functional group (and thus their gene products the first protein complex) to be lost or degraded before full mycoheterotrophy, and thus before the loss of photosynthesis (e.g. Barrett and Davis, 2012; Wicke et al., 2013; Barrett et al., 2014). This is consistent with what I found in Ericaceae and Gentianaceae, as the four partial mycoheterotrophs that I surveyed all have degradation of the genes coding for the plastid NAD(P)H dehydrogenase complex, despite the retention of all or most of the other plastid-encoded genes. The loss or non-functionality of this complex in partial (photosynthetic) mycoheterotrophs may reflect less photooxidative stress in understory plants that do not obtain all of their nutrition from sunlight (Barrett et al., 2014). Stable isotope signatures for the partial mycoheterotrophs and congeners sequenced here are enriched in 15N   26 and 13C, but at an intermediate level between full mycoheterotrophs and autotrophs, pointing to incomplete reliance on fungal nutrition (Tedersoo et al., 2007; Zimmer et al., 2007; Cameron and Bolin, 2010). Degradation of the NAD(P)H complex may have occurred repeatedly in different lineages of photosynthetic orchids (Wu et al., 2010; Yang et al., 2013; Kim et al., 2015; Ruhlman et al., 2015). It remains to be shown how many of these plants are partial mycoheterotrophs at maturity, although isotopic evidence suggests this is the case in several orchids (Gebauer and Meyer, 2003; Bidartondo et al., 2004; Tedersoo et al., 2007; Zimmer et al., 2007, 2008).  Some authors have proposed functional replacement by nuclear copies of plastid ndh genes as a possible explanation for plastid-encoded NAD(P)H degradation (e.g. Braukmann et al., 2009; Wu et al., 2010; Blazier et al., 2011). However, Ruhlman et al. (2015) found no evidence of expressed nuclear copies of plastid-encoded ndh genes or functional nuclear-encoded components in taxa where plastid-encoded subunits are lost or degraded (although a secondary nuclear-encoded plastid ndh complex has been hypothesized in at least Arabidopsis; Peltier and Cournac, 2002, see Wicke et al. 2011). It should also be noted that other photosynthetic (and likely non-mycoheterotrophic) lineages of plants have also experienced loss of this plastid protein complex (i.e., Gnetales and Pinaceae, Braukmann et al., 2009; Wu et al., 2009; some Geraniaceae, Blazier et al., 2011; some Lentibulariaceae Wicke et al., 2014; four lineages of Alismatales, Iles et al., 2013; Peredo et al., 2013; Ross et al., 2015; some Cactaceae, Sanderson et al., 2015). Thus, loss or non-functionality of the complex is not necessarily an indicator that a transition to heterotrophy has occurred, or is likely to happen.     27 4.3.2 Loss and retention of photosynthesis-related genes Full mycoheterotrophs generally have degraded or lost plastid-encoded genes related to photosynthesis (Wicke et al. 2011), comprising the photosystem I and II complexes and assembly factors (ycf3 and ycf4), cytochrome b6/f complex, Rubisco and CO2 uptake (cemA), and ATP synthase. This is the case even for taxa in the relatively early stages of plastome reduction, such as the coralroot orchids Corallorhiza (Orchidaceae) and the liverwort Aneura mirabilis (Aneuraceae) (Wickett et al., 2008; Barrett and Davis, 2012; Barrett et al., 2014). In Gentianaceae, pseudogenization within these genes is minimal in Exochaenium oliganthum, but there is a near-complete loss of these genes in Voyria (Table 3). Most photosynthesis-related genes are deleted or have reading frame interruptions in Epirixanthes pallida (Polygalaceae), consistent with loss of photosynthetic function; however, six plastid-encoded ATP synthase genes and the Rubisco large subunit (rbcL) have been retained in this species with open reading frames (Table 2). These genes were not recovered in the gene set assembled for Ep. elongata, pointing to a loss of this complex compared to its congener (this needs to be confirmed by completing the plastid genome for this species). The ATP synthase genes and rbcL are also retained as open reading frames in Exo. oliganthum (Gentianaceae), although their retention here is may be less surprising given the relatively minor extent of photosynthesis gene reduction in its plastome, and the recency of loss of photosynthesis in it (Merckx et al., 2013b).  Although a lag is expected before reading frames are interrupted in photosynthesis genes following the initial functional loss of photosynthesis (Leebens-Mack and DePamphilis, 2002) the retention of ATP synthase genes and Rubisco in multiple independent mycoheterotrophic lineages after other photosynthesis genes are degraded is noteworthy, and points to probable secondary (non-photosynthetic) functions for them (Wickett et al., 2008). Their retention in some   28 Gentianaceae and Polygalaceae here adds to the comparable published cases in non-photosynthetic representatives of Corallorhiza, in the liverwort Aneura mirabilis, and in the monocot Petrosavia stellaris (Petrosaviaceae) (Wickett et al., 2008; Barrett and Davis, 2012; Barrett et al., 2014; Logacheva et al., 2014); note, though, that the Corallorhiza species lack an open reading frame for rbcL. The plastid-encoded ATP synthase genes are also retained in some holoparasitic Cuscuta species (Convolvulaceae, their retention there may reflect cryptic photosynthesis during seedling establishment; Machado and Zetsche, 1990) and in some representatives of holoparasitic Orobanchaceae, and intact rbcL genes have also been identified in non-photosynthetic representatives in both families (Delavault et al., 1995; Randle and Wolfe, 2005; Funk et al., 2007; McNeal et al., 2007; Wicke et al., 2013). The complete suite of ATP synthase genes are also found in the plastome of a heterotrophic alga, where they are apparently transcribed (Knauf and Hachtel, 2002). Although ATP synthase is directly involved in the production of ATP used in the carbon fixing reactions of photosynthesis (reviewed in Walker, 2012), repeated retention of these genes in some heterotrophs prompted Barrett and Davis (2012) and Barrett et al. (2014) to propose that ATP synthase genes are at least initially retained after the loss of photosynthesis. They may therefore act as a landmark for an intermediate level of genome degradation. Rubisco may follow the same general pattern of delayed loss. However, as there may be no linkage between the proposed secondary functions of these protein complexes (see below), we propose that these two complexes may subsequently be lost in either order.  An explanation for the retention of (putatively) functional ATP synthase has not yet been put forward, but a continued need for plastid ATP production from a non-photosynthetically driven proton gradient, or a need for ATP hydrolysis in heterotroph plastids, have both been proposed (Wicke et al., 2013). Involvement in additional metabolic pathways may also explain   29 why some non-photosynthetic heterotrophs retain a putatively functional rbcL gene (see McNeal et al., 2007; Wickett et al., 2008; Wicke et al., 2013; Logacheva et al., 2014). In addition to its primary role in the Calvin cycle, Rubisco is known to catalyze a glycolysis-bypassing lipid synthesis pathway in white turnip (Brassica napus, Brassicaceae), although this reaction is thought to require functioning photosynthetic machinery (Schwender et al., 2004). Rubisco is also involved in the production of serine and glycine via the glycolate pathway of the C2 cycle (Tolbert, 1997) and is expressed at low levels in the non-photosynthetic seeds of the castor bean (Ricinus communis, Euphorbiaceae), although its function there is unclear (Osmond et al., 1975). It would be worthwhile to determine whether these or related biosynthetic pathways are maintained in heterotrophs with (otherwise) degraded photosynthesis genes.   In the partial mycoheterotrophs considered here (Table 2), the retention of all or most of the photosynthetic genes and isotopic evidence are both consistent with retention of a functional photosynthetic apparatus. However, in Obolaria virginica (Gentianaceae), two genes with products involved in photosynthesis are lost (psbM) or have reading frame interruption (ccsA). Despite these losses, isotopic evidence and visible photosynthetic tissue support the retention of photosynthesis in this species (Cameron and Bolin, 2010). Barrett et al. (2014) also found interruption of reading frames in psbM (and psaI) in putatively partial mycoheterotrophic species of Corallorhiza. These two genes have roles in the assembly and stability of photosystems II and I, respectively. In psbM deficient mutants, the movement of electrons around the PSII complex and stability of component dimers is diminished, but functional (Umate et al., 2007; Kawakami et al., 2011). Xu et al. (1995) demonstrated that psaI provides structural stability to photosystem I, but defective mutants had only marginally declined efficiency in the affected photosystem. As   30 such, it may be possible for O. virginica and Corallorhiza partial mycoheterotrophs to photosynthesize to some degree in the absence of these subunits.  The c-type cytochrome biogenesis protein (ccsA) is widely retained in the plastomes of photosynthetic plants (see Fajardo et al. (2013) for a possible exception), but Peterson et al. (2015) found that the gene was pseudogenized in the photosynthetic hemiparasite Viscum alba. The protein product ccsA is responsible for heme attachment to c-type cytochromes, which are essential components of the photosynthetic electron transport chain (reviewed in Wicke et al., 2011). Mutations in this gene in Chlamydomonas reinhardtii resulted in non-photosynthetic phenotypes, which was attributed to the failure to synthesize some forms of cytochromes (Xie and Merchant, 1996). However, Saint-Marcoux et al. (2009) demonstrated that heme delivery to b6-type cytochromes is mediated by a different protein, and suggested that cytochrome b6f may be assembled in the absence of functional ccsA. CcsA is among the few photosynthesis genes with reading frame interruptions in the plastome of the full mycoheterotroph Exo. oliganthum (Gentianaceae), suggesting that it may be lost relatively early in plastid genome degradation (Table 2).   4.3.3 Loss of plastid-encoded RNA polymerase (PEP) genes Plastid-encoded RNA polymerase (PEP) is coded for by four rpo genes in the plastid genome (Table 3), and is thought to perform the majority of transcription, at least in photosynthetic leaves (Zhelyazkova et al., 2012; reviewed in Liere et al., 2011). The complex may not be essential when photosynthesis genes are lost in full mycoheterotrophs, and nuclear-encoded RNA polymerase (NEP) may perform plastid gene transcription for non-photosynthetic genes that are usually or partly transcribed by PEP (most plastid genes have NEP and PEP promoters,   31 Liere et al., 2011). For example, functional replacement by nuclear gene products has been given as an explanation for the loss of plastid-encoded RNA polymerase (PEP) genes in ‘holoparasitic’ (but cryptically photosynthetic) Cuscuta species (reviewed in Krause, 2008). Berg et al. (2004) demonstrated that nuclear-encoded RNA polymerase (NEP) performs the plastid gene transcription that is usually performed by PEP in Cuscuta species. Barrett and Davis (2012) initially proposed a two-stage loss (photosynthesis genes and then PEP genes), but Barrett et al. (2014), instead proposed the concerted loss of photosynthesis genes (excluding ATP synthase) and PEP, as they found no evidence that full mycoheterotrophs with recent lost of photosynthesis have retained uninterrupted rpo genes with degraded (or lost) photosynthesis genes (see also Wicke et al., 2013). However, the plastome of Exo. oliganthum (Gentianaceae) provides a probable example of the latter (photosynthesis genes degrading before PEP), as its rpo genes are retained and are still present in open reading frame (Figs. 1, S4; Table 2). Thus, this provides initial support for the two-stage hypothesis proposed by Barrett and Davis (2012). My finding should be followed up with a functional study of PEP gene activity in Exo. oliganthum.  4.3.4 Loss of ribosomal protein and tRNA genes Land-plant plastomes encode some of the components of the plastid translational machinery, forming complete complexes with nuclear-encoded products. Complete plastid ribosomes are formed by 58-62 ribosomal proteins and four ribosomal RNA subunits, and among these the plastomes of land plants commonly encode all four ribosomal RNA subunits (rDNA genes) and 21 ribosomal proteins (12 small subunit or rps genes; 9 large subunit or rpl genes) (Palmer, 1985; Wicke et al., 2011; Sugiura, 2014). Ribosomal DNA loci are highly conserved (Palmer, 1985; Harris et al., 1994), and they are retained in all sequenced heterotrophic plant plastomes to   32 date, including the completely assembled mycoheterotrophs presented here (see Barrett et al., 2014; Lam et al., 2015; Petersen et al., 2015; Table 2). Plastid-encoded ribosomal protein genes are rarely lost in autotrophs (but see Jansen et al., 2007, 2011; Fajardo et al., 2013; Martínez-Alberola et al., 2013), but some have been deleted or found as pseudogenes in the relatively more degraded plastomes of some heterotrophic plants (Delannoy et al., 2011; Wicke et al., 2013; Lam et al., 2015; Schelkunov et al., 2015). Exochaenium oliganthum (Gentianaceae) retains all 21 plastid-encoded ribosomal proteins in open reading frame, and Ep. pallida (Polygalaceae) a slightly smaller set of 18, although it has two losses in common with autotrophic relatives. A set of eleven plastid-encoded ribosomal proteins is retained across mycoheterotrophic plants: rps2, 3, 4, 7, 8, 11, 14 and rpl2, 14, 16 and 36 (see Lam et al., 2015). These genes are retained in all fully assembled plastomes presented here, plus an additional four: rps12, 18, 19 and rpl20 (Table 2).  Land-plant plastomes generally retain loci for 30 transfer RNA (tRNA or trn genes). A complete, or nearly complete, set of these is retained in heterotrophs with relatively less degraded plastomes (e.g. Aneura mirabilis, Wickett et al., 2008; Petrosavia stellaris, Logacheva et al., 2014). Exochaenium oliganthum (Gentianaceae) and Ep. pallida (Polygalaceae) retain complete sets, although the latter has lost many more photosynthesis genes than the former (Table 3). The loss of many transfer RNA genes is typical of highly reduced mycoheterotrophs and parasites (Delannoy et al., 2011; Wicke et al., 2013; Lam et al., 2015; Schelkunov et al., 2015). The two Voyria (Gentianaceae) species and Ep. elongata (Polygalaceae) retain few of the thirty transfer RNA genes normally coded for by the plastome (Tables 3, S4: four tRNA genes in V. caerulea, 13 in V. clavata and five in Ep. elongata, although note that the latter two are based on incomplete assemblies). The loss of some tRNA genes might be compensated for by import   33 from the cytosol (e.g. Alkatib et al., 2012) or ‘superwobbling’ (Rogalski et al., 2008), although a few of them may not be replaceable by either means (Barbrook et al., 2006). Among these is trnE-UUC, whose gene product (glutamyl tRNA) has a secondary role outside translation, in heme biosynthesis (Jahn et al., 1992), and possibly in the regulation of nuclear-encoded plastid RNA synthase (NEP) (Hanaoka et al., 2005, but see Bohne et al., 2009). Barbrook et al. (2006) proposed that the interaction of glutamyl tRNA with multiple enzymes involved in the production of heme makes a replacement by a cytosolic product unlikely. Howe and Purton, (2007) gave a related explanation for the retention of plastid-encoded formylmethionyl-tRNA (trnfM-CAU), which has a role in initiating translation in plastids and possibly some mitochondria (Barbrook et al., 2006). It has been suggested that the need to be recognized by multiple enzymes limits the likelihood of replacement (Barbrook et al., 2006; Howe and Purton, 2007; Delannoy et al., 2011); presumably this would require independent adjustment to replacement, by each enzyme that interacts with the tRNA. Barbrook et al. (2006) proposed that the indispensability of these two plastid-encoded transfer RNAs could explain the retention of plastomes in non-photosynthetic organisms, which they call the “essential tRNAs hypothesis.” In total, four tRNA genes are retained in all species presented here: trnW-CCA, trnI-CAU, trnfM-CAU, trnE-UUC (Table 2). The latter three are retained in all sequenced heterotrophic plants to date, and lend support to the essential tRNA hypothesis.   4.3.5 Loss of other plastid genes of known and unknown function Plastids are not just photosynthetic organelles. They are the site of additional essential cellular functions including fatty acid, amino acid and tetrapyrrole biosynthesis, pigment production and the conversion of inorganic nitrogen to useful forms (reviewed in Ernes and Neuhaus, 2005). In   34 addition to genes involved in the plastid genetic apparatus, loci that are retained in the most degraded heterotrophic plastomes encode proteins with roles in essential non-photosynthetic metabolism, including protein turnover and import or intron removal. MATK, the only plastid-encoded group IIa intron maturase (Zoschke et al., 2010), is coded for by matK, a locus retained in nearly all plant plastid genomes. However, the matK gene has been deleted from the plastome of the mycoheterotrophic orchids Rhizanthella gardneri, Epipogium aphylla and E. roseum, and from some holoparasitic Cuscuta species (Funk et al., 2007; McNeal et al., 2007; Delannoy et al., 2011; Braukmann et al., 2013; Schelkunov et al., 2015). Voyria clavata (Gentianaceae) has lost matK, and I also did not recover it in V. caerulea (based on the gene set assembled for this without a full circular genome). It is also likely a pseudogene in Ep. elongata (Polygalaceae), as a non-triplet deletion ~900 bp into the reading frame results in a frame shift. In Cuscuta, the loss of matK coincides with the loss of all group IIa introns (McNeal et al., 2009). In contrast, Epipogium and Rhizanthella retain loci with group IIa introns, and at least two of these genes (rpl2 and rps12) are thought to be targeted by MATK (Zoschke et al., 2010, although the retention of the third exon of rps12, and its MATK targeted intron, is uncertain in Epipogium; Schelkunov et al., 2015). This parallels the situation in Voyria and Ep. elongata where rpl2 and rps12 are retained with group IIa introns intact. Delannoy et al. (2011) demonstrated in Rhizanthella that rpl2 is correctly spliced, suggesting that an alternative splicing factor facilitates intron removal from their RNA transcripts. Furthermore, rpl2 is one of the plastid-encoded ribosomal protein loci retained across heterotrophic land plants, and therefore is likely functional in Voyria and Ep. elongata. My finding could be followed up with selection tests to ascertain whether these genes in matK-deleted plastomes are under the same selective regime as homologous genes in matK-retaining plastomes.   35 The gene coding for the beta subunit of Acetyl-CoA carboxylase (accD) is retained in all sequenced plastomes of heterotrophic plants. The protein product is assembled with nuclear-encoded subunits to form a complex that catalyzes the formation of essential components of fatty acids (Ohlrogge and Browse, 1995; Sasaki and Nagano, 2004). However, losses have been documented in autotrophs, where functional transfer of plastid-encoded accD to the nucleus has been proposed and demonstrated in some lineages (Straub et al., 2011; Rousseau-Gueutin et al., 2013; Sabir et al., 2014). All Gentianaceae plastomes sampled here retain the genes coding for the beta subunit of Acetyl-CoA carboxylase (accD), including the very reduced plastome of Voyria clavata (Gentianaceae) and gene set of V. caerulea (Table 2). In the gene set of Ep. elongata (Polygalaceae) I recovered a ~500 bp truncated accD that is likely a pseudogene. The functional status of the accD gene is otherwise unclear in the Polygalaceae and Ericaceae representatives presented here, regardless of trophic category. Only subregions (~30-60%) of the open reading frames recovered in autotroph and mycoheterotroph representatives of these families presented here BLAST to the homologous gene in relatives, where accD is clearly retained (Ceratonia siliqua, Fabaceae, and Camellia crapnelliana, Theaceae). In contrast C. siliqua and C. crapnelliana BLAST to Nicotiana tabacum (Solanaceae) with 94% and 99% query cover, respectively. A regulatory role has been proposed to explain the general retention of the plastid-encoded subunit of accD (Bungard, 2004; Delannoy et al., 2011), but the nuclear relocation of the gene in some autotrophs suggests that this plastome-specific role is not essential, in at least some lineages. Nevertheless, the retention of long open-reading frames for accD-like genes in these two families, despite substantial sequence change, suggests the retention of function of some kind, which warrants further investigation.   36 ClpP is a plastid-encoded subunit of the Clp protease (or ATP-dependent caseinolytic protease), which has roles in protein turnover and processing, but has also been linked to isopyrenoid and tetrapyrrole biosynthesis, and fibrillins (lipid-body stabilizing molecules) (Kim et al., 2009; Stanne et al., 2009; Krause, 2012). The gene is found in most, but not all heterotrophic plant plastomes, and is considered essential for plant development (Kuroda and Maliga, 2003); a gene with a similar protein product (clpC) is retained in the reduced plastome of apicomplexan parasites (reviewed in Sato, 2011). As with accD, this gene is deleted from the plastomes of several lineages of autotrophs, where a nuclear gene presumably codes for the protein, although this has not been demonstrated (Jansen et al., 2007; Straub et al., 2011). All Gentianaceae plastomes retain the clpP locus, but it is deleted from the plastomes of fully assembled Polygalaceae, and was not recovered in the genes sets of Ericaceae representatives.   The plastid-encoded translation initiation factor infA has been lost independently many times in land plants (Wicke et al. 2011), and transfer to the nucleus has been demonstrated in several eudicot (asterid and fabid) lineages lacking the plastid locus (Millen et al., 2001; Jansen et al., 2007). Loss of infA may be associated with heterotrophy in Gentianaceae, as it is retained in autotrophic Exacum but lost (or found with reading frame interruptions) in four of the five heterotrophs (Table 3). The reading frame is uninterrupted in Ericaceae representatives, and thus may be functional. Fully assembled Polygalaceae retain infA loci with multiple reading frame interruptions, and I did not recover the gene in the gene sets of Salomonia cantoniensis or Ep. elongata.  Ycf1 and ycf2 are large hypothetical chloroplast reading frames for which reading frame interruptions have lethal consequences in tobacco (Drescher et al., 2000), yet their precise functions remain uncertain. These loci are retained in most land plants and some heterotrophs,   37 but have been deleted or pseudogenized in a few autotrophic lineages (Downie et al., 1994; Jansen et al., 2007). Sequence similarity in the binding domain of ycf1 to genes in the CDC48 family prompted Wolfe (1994) to suggest a cell membrane-related function for the gene. Recently Kikuchi et al. (2013) demonstrated association of its gene product with a nuclear-encoded inner-envelope membrane translocon complex (TOC/TIC machinery), and Nakai (2015) proposed renaming ycf1 as tic214 (but see de Vries et al, 2015). Less is known about ycf2, but drought-stress expression profiling suggests a role in water-use efficiency (Ruiz-Nieto et al., 2015). I recovered the ycf1and ycf2 loci with open reading frames in all fully assembled Gentianaceae species except Voyria: ycf1 is deleted from V. clavata and may or may not be retained in V. caerulea (as I recovered an incomplete gene without reading frame interruption). Ycf2 is severely truncated in both Voyria species (Table 2). Both genes are likely functional in autotrophic Polygalaceae, but there are truncated, probable pseudogenes of them in Ep. pallida and I did not recover either locus in Ep. elongata. Braukmann and Stefanovi! (2012) noted lack of ycf2 probe hybridization in Arbutoideae and Pyroleae, based on a survey of plastome gene content in Ericaceae. I did not recover a ycf2 locus in Orthilia, and the gene is absent in Arbutus unedo (Martínez-Alberola et al., 2013). However, I did recover a severely truncated locus in the Pyrola gene set.   4.4 Structural rearrangement and the inverted repeat Sequenced plastid genomes of heterotrophs (parasites and mycoheterotrophs) have a range of levels of genome rearrangement, from those that are essentially colinear with green relatives despite gene loss (e.g. Sciaphila densiflora, Triuridaceae, Lam et al. 2015; Corallorhiza spp., Orchidaceae, Barrett et al. 2014; Aneura mirabilis, Aneuraceae, Wickett et al., 2009; Epifagus   38 virginiana, Orobanchaceae Wolfe et al., 1992) to highly rearranged ones (e.g. Orobanche crenata, Orobanchaceae, Wicke et al. 2013; Petrosavia stellaris, Petrosaviaceae, Logacheva et al., 2014), and including several intermediate levels of rearrangement (e.g., some Cuscuta, Convolvulaceae, Funk et al., 2007; some Orobanchaceae, Wicke et al., 2013). It is not well understood whether there are general processes affecting genome structure in mycoheterotrophs, and so the newly sequenced genomes here provide additional independent data points for addressing this issue.  Gene loss is associated with considerable changes in genome structure in many of the mycoheterotroph genomes included here (Fig. 1, Table 2), and inverted repeat boundaries have also shifted in some cases (Figs. 1, S1-S7, Table 4). Inverted repeat boundary shifts are reasonably common at the IR/large single copy boundary in autotrophic lineages, but shifts at the IR/small single copy boundary (as found here for autotrophs Polygala and Arbutus; note that their inverted repeats extend four and eleven genes further into what is the small single copy region in Nicotiana, for example, Figs. 2, Table 4) are less common (Zhu et al., 2015).  Setting aside frequent genome compaction due to gene loss, and the typically minor shifts in IR boundaries, the completely (or nearly completely) sequenced plastid genomes of the eudicot mycoheterotrophs examined here (Fig. 1, Table 4) generally do not appear to be evolving in a substantially different manner to their closest green relatives. For example, in Gentianaceae, all four partially and fully mycoheterotrophic taxa (two partial mycoheterotrophs, Bartonia and Obolaria, and two full mycoheterotrophs, Exochaenium and Voyria) have colinear or nearly colinear genomes with an autotrophic member of Gentianaceae (Exacum; Fig. 3), that in turn is colinear with the plastome of tobacco (Nicotiana, Solanaceae; Fig. 2), which has a gene order that is similar to most other angiosperms (Palmer, 1985; Palmer and Stein, 1986; Jansen et al.,   39 2007). One major rearrangement in Voyria clavata concerns the boundaries of the inverted repeats (and hence single copy regions), which no longer span genes found in the inverted repeat regions of other members of the family. Voyria also has a single inversion compared to the other Gentianaceae (the right-hand LCB in Fig. 3). Nonetheless, gene order is otherwise largely conserved (one of the Voyria inverted repeat copies falls in a locally colinear block, LCB; note that the other is not shown in Fig. 3), and this minimal pattern of genome restructuring is comparable to that of Sciaphila (Triuridaceae; Lam et al., 2015), Rhizanthella (Orchidaceae; Delannoy et al., 2011) and Epifagus (Orobanchaceae; Wolfe et al., 1992) in terms of having retained colinearity despite extensive gene loss. I inferred multiple plastid genome rearrangements in Epirixanthes (Polygalaceae) and Orthilia (Ericaceae) compared to their close green relatives (Polygala and Arbutus, respectively; Figs. 3). However, in both cases, their green relatives also have fairly substantial rearrangements compared to tobacco (Nicotiana, Solanaceae; Fig. 2), so it may difficult to distinguish any effects of mycoheterotrophy from other processes leading to genome rearrangement in these taxa. In Ericaceae, two fully assembled plastomes have been published (Arbutus unedo, Martínez-Alberola et al., 2013; Vaccinium macrocarpon, Fajardo et al., 2013). Both show multiple major rearrangements in comparison to tobacco (see Arbutus unedo in Fig. 2). Unusually, the inverted repeat regions of Arbutus and Vaccinium have expanded to encompass nearly all of the ancestral small single copy region in both autotrophs. Martínez-Alberola et al. (2013) noted that among 37 asterid plastomes sampled, only Arbutus unedo, Vaccinium macrocarpon and two other species had tandem repeats larger than 150 bp (‘megasatellites’), which are associated with rearrangement in pathogenic yeast (Thierry et al., 2008). Dispersed repeats may contribute to plastid genome rearrangements in other taxa (Downie and Palmer, 1992; Cosner et al., 1997; Cai   40 et al., 2008; Haberle et al., 2008), and it is possible that dispersed repeat proliferation is a characteristic of Ericaceae plastomes, including the mycoheterotrophs here, though I did not attempt to characterize this possibility here. However, given the relatively modest level of plastome degradation observed in Orthilia secunda and the similar level of rearrangement found in fully autotrophic Ericaceae (Fig. 2), it is likely that the number of rearrangements are attributable to shared characteristics of the family, and not to trophic status. Dispersed repeats may also explain substantial plastid genome rearrangements in some taxa in an inverted-repeat-lacking clade (IRLC) of legumes (Fabaceae), the completely autotrophic sister group of Polygalaceae (e.g., Cai et al. 2008; Schwarz et al., 2015) although it has also been suggested that the lack of an inverted repeat also contributes to genome instability (Palmer et al., 1987; Milligan et al., 1989; Cai et al., 2008; Sabir et al., 2014). Outside this clade of legumes, other members of Fabaceae are largely conserved in plastid genome structure, although several inversions and gene losses have been documented in subfamily Papillionoideae (Schwarz et al., 2015). The loss of the plastid inverted repeat (IR) in Epirixanthes, and its switch to a mycoheterotrophic nutritional mode, may not contribute to genome rearrangement in this taxon (Figs. 4, S2), as Polygala arillata is autotrophic and retains an inverted repeat, and yet also has substantial rearrangements compared to Nicotiana (Fig. 2), Nonetheless, the loss of an IR in Epirixanthes provides an intriguing parallel to the IRLC in the sister group of Polygalaceae. The gain of a single large direct repeat in Epirixanthes is also unusual and noteworthy (Figs. 4, S2). Large repeats are thought to be selected against as destabilizing elements in plastomes that cause aberrant recombination (Gray et al., 2009; Maréchal and Brisson, 2010). As with Ericaceae, it would be useful to explore the possibility that dispersed repeats have contributed to genome rearrangements in autotrophic and mycoheterotrophic Polygalaceae.   41  4.5 Conclusion A rationale for the retention of genetic apparatus genes is the continued need to express plastid genes with putatively essential roles that are not involved in photosynthesis (e.g., accD, clpP, trnE) (Delannoy et al., 2011; Krause, 2008). As independent losses of accD and clpP have occurred in multiple lineages of photosynthetic plants, including species sampled here, the endpoint of plastome reduction may vary by lineage in a manner that is unrelated to heterotrophy. Some essential plastid genes may not be readily replaceable in non-photosynthetic plants (e.g., trnE), and it is not yet clear if any land plants have completely lost their plastomes (see Molina et al., 2014, for a possible exception), although this is known in some heterotrophic protists (Janou#kovec et al., 2015). The patterns of gene loss characterized here are generally consistent with the trajectory hypothesized by Barrett and Davis (2012) model: plastid NAD(P)H dehydrogenase is likely lost before the loss of photosynthesis in partially mycoheterotrophic plants, most photosynthesis genes are then lost after the initial switch to full mycoheterotrophy, and plastid-encoded RNA polymerase genes are lost next. ATP synthase subunit genes and rbcL appear to repeatedly linger after the loss of photosynthesis, likely because of secondary non-photosynthetic roles that they play in the plastid. I propose here that they may be lost in either order after the loss of most photosynthesis genes (this is a modification of the hypothesis of Barrett and Davis, 2012). In the late stages of full mycoheterotrophy, multiple genes in the plastid translation apparatus are lost from the plastome (the most extreme example here is Voyria, Gentianaceae), although a core set of ribosomal protein, rDNA and tRNA genes is retained in all mycoheterotrophs examined here. Other non-photosynthetic genes may be lost in a more sporadic manner in the later stages of gene loss, and may include some surprising losses   42 (e.g., of matK, given that some group IIa introns are retained). Future work should follow these observations up with physiological studies to assess gene-product functionality (for example to determine the possible functions of plastid ATP synthase and Rubisco in the full mycoheterotrophs that retain them). The full and partial mycoheterotrophs sampled here also vary considerably in terms of plastome size and gene content, from extremely reduced to only marginally degraded, and from substantially rearranged plastomes to those that are nearly colinear with green relatives. These differences do not appear to be related to the loss of photosynthesis or the loss of the plastid inverted repeat regions. Because substantial diversity was uncovered among close relatives that represent the same loss of photosynthesis (in Voyria and Epirixanthes), it would be useful to continue sampling in these genera and other eudicot mycoheterotrophs. Despite gene loss and moderate rate elevation, the plastid gene sets recovered here are shown to be useful for inferring phylogenetic relationships of the mycoheterotrophic eudicots.  43 Table 1. Specimen source information; herbarium abbreviations follow Thiers (2015)     Specimen voucher Trophic status1 Family Species [Collector number (herbarium)]  _________________________________________________________________________________________________ Full MH Gentianaceae Exochaenium oliganthum (Gilg.) Kissling Sainge s.n. (YA) Full MH Gentianaceae Voyria caerulea Aubl. Merckx 244 (L)  Full MH Gentianaceae Voyria clavata Splitg. Merckx 224 (L) Full MH Polygalaceae Epirixanthes elongata Blume Hsu 17814 (FLAS) Full MH Polygalaceae Epirixanthes pallida T. Wendt Merckx & Mennes CM001 (L) Partial MH Ericaceae Orthilia secunda (L.) House No voucher1 Partial MH Ericaceae Pyrola minor L. No voucher2 Partial MH Gentianaceae Bartonia virginica  (L.) Britton, Sterns & Poggenb. Bertin 6708 (MASS) Partial MH Gentianaceae Obolaria virginica L. Stefanovic SS-04-103 (TRT) Full autotroph Gentianaceae Exacum affine Balf.f. ex Regel Darby s.n. (UBC) Full autotroph Polygalaceae Polygala arillata Buch.-Ham. ex D. Don. Larsen 46516 (FLAS) Full autotroph Polygalaceae Salomonia cantoniensis Lour. Nosuro 9830009 (FLAS) 1 See Beatty and Provan (2010)  2 See Beatty et al. (2010) ! ""!Table 2. Plastid gene content across newly sequenced taxa of Gentianaceae, Polygalaceae and Ericaceae (the Arbutus unedo plastome is from Martínez-Alberola et al., 2013). Full mycoheterotrophs are bolded, and partial mycoheterotrophs are underlined. An asterisk (*) indicates that a full plastid genome was assembled. Genes with open-reading frames are indicated by ‘+’ (incompletely recovered genes with open reading frames by ‘(+)’). Gene absence (loci for which remnants could not be detected in full genomes, or that could not be retrieved in plastid gene set assemblies) is indicated with a dash (‘-’). Probable pseudogenes (loci with multiple internal stop codons, see text) are indicated as ‘!’. Loci with single reading frame interruption included in ORF-only matrix (there are four) are indicated with ‘#’. Genes found intact in all fully assembled species are indicated in bold font.     Gentianaceae   Polygalaceae  Ericaceae              Gentianaceae              Polygalaceae  Ericaceae  *Exacum affine *Bartonia virginica *Obolaria virginica *Exochaenium oliganthum Voyria caerulea *Voyria clavata *Polygala arillata Salomonia cantoniensis *Epirixanthes pallida Epirixanthes elongata *Arbutus unedo Orthilia secunda Pyrola minor  *Exacum affine *Bartonia virginica *Obolaria virginica *Exochaenium oliganthum Voyria caerulea *Voyria clavata *Polygala arillata Salomonia cantoniensis *Epirixanthes pallida Epirixanthes elongata *Arbutus unedo Orthilia secunda Pyrola minor NAD(P)H dehydrogenase              psbL + + + + - - + + - - + + + ndhA + - ! ! - - + + - - + ! ! psbM + + - + - - + + + - + + + ndhB + ! ! + ! - + + ! - + + + psbN + + + + - - + + - - + + + ndhC + - ! + - - + + ! - + + + psbT + + + + - - + + - - + + + ndhD + ! ! ! - - + + - - + ! ! psbZ + + + + - - + + ! - + + + ndhE + - ! ! - - + + - - + ! # PSI assembly factors              ndhF + - ! ! - - + + ! - + ! ! ycf3 + + + ! - - + + ! - + + + ndhG + - ! + - - + + - - + ! ! ycf4 + + + + - - + + - - + + + ndhH + ! ! ! - - + + ! - + ! ! Cytochrome b6/f complex              ndhI + - - + - - + + - - + ! ! petA + + + + - - + + - - + + + ndhJ + ! - # - - + + - - + # ! petB + + + + - - + (+) - - + + + ndhK + - ! ! - - + + ! - + + + petD + + + + ! - + + ! - + + + Photosystem (PS) I               petG + + + + - - + + + - + + + psaA + + + ! - - + (+) ! - + + + petL + + + + - - + + - - + + + psaB + + + + ! - + (+) ! - + + + petN + + + + - - + + - - + + + psaC + + + + - - + + ! - + + + Rubisco              psaI + + + + - - + + - - + + + rbcL + + + + - ! + + + - + + + psaJ + + + + - - + + + - + + + ATP synthase              Photosystem (PS) II              atpA + + + + ! - + + + - + + + psbA + + + ! - - + + ! - + + + atpB + + + + - - + + + - + + + psbB + + + + - - + (+) - - + + + atpE + + + + - - + + + - + + + psbC + + + + ! - + + ! - + + + atpF + + + + ! - + + + - + + + psbD + + + + ! - + + ! - + + + atpH + + + + + - + + + - + + + psbE + + + + - - + + - - + + + atpI + + + + ! - + + + - + + + psbF + + + + - - + + - - + + + Other photosynthesis proteins              psbH + + + + - - + + - - + + + cemA + + + + - - + + - - + + + psbI + + + + - - + + ! - + + + ccsA + + ! ! - - + + ! - + + (+) psbJ + + + + - - + + - - + + +               psbK + + + + - - + + - - + + +               ! "#! *Exacum affine *Bartonia virginica *Obolaria virginica *Exochaenium oliganthum Voyria caerulea *Voyria clavata *Polygala arillata Salomonia cantoniensis *Epirixanthes pallida Epirixanthes elongata *Arbutus unedo Orthilia secunda Pyrola minor  *Exacum affine *Bartonia virginica *Obolaria virginica *Exochaenium oliganthum Voyria caerulea *Voyria clavata *Polygala arillata Salomonia cantoniensis *Epirixanthes pallida Epirixanthes elongata *Arbutus unedo Orthilia secunda Pyrola minor RNA Polymerase              Ribosomal DNA genes              rpoA + + + + ! - + + ! - + + + rrn4.5 + + + + + + + - + + + + + rpoB + + + + - - + (+) ! - + + + rrn5 + + + + + + + - + - + + + rpoC1 + + + + - - + + ! - + + + rrn16 + + + + (+) + + + + + + + + rpoC2 + + + + - - + + ! - + + + rrn23 + + + + + + + (+) + + + + + Proteins of other function              Transfer RNA genes              accD + + + + (+) + + + + ! ! (+) + trnA-UGC + + + + - - + + + - + + + clpP + + + + + + - - - - ! - - trnC-GCA + + + + + - + + + - + + + infA + ! ! + - ! ! - ! - + + + trnD-GUC + + + + + - + + + + + + + matK + + + + - - + + + # + + + trnE-UUC + + + + + + + + + + + + + Proteins of unknown function              trnF-GAA + + + + + - + + + - + + + ycf1 + + + + (+) - + (+) ! - ! - - trnfM-CAU + + + + + + + + + - + + + ycf2 + + + + ! ! + ! ! - - - ! trnG-GCC + + + + - - + + + - + + + Ribosomal proteins              trnG-UCC + + + + (+) - + + + - + + + rpl2 + + + + + + + + + + + (+) + trnH-GUG + + + + - - + + + - + + + rpl14 + + + + + + + + + + + + + trnI-CAU + + + + + + + + + + + + + rpl16 + + + + + + + + + + + + + trnI-GAU + + + + - - + + + - + + + rpl20 + + + + + + + + + - + + + trnK-UUU + + + + - - + + + - + + + rpl22 + + + + ! ! - - - - + + + trnL-CAA + + + + + - + + + - + + + rpl23 + + + + ! ! + + + - + ! ! trnL-UAA + + + + - - + + + - + + + rpl32 + + + + - - + + ! - + + + trnL-UAG + + + + - - + + + - + + + rpl33 + ! + + ! - + + + - + + + trnM-CAU + + + + - - + + + - + + + rpl36 + + + + - + + + + + + + + trnN-GUU + + + + + - + + + - + + - rps2 + + + + + + + + + + + + + trnP-UGG + + + + + - + + + - + + + rps3 + + + + + + + + + + + + + trnQ-UUG + + + + - - + + + - + + + rps4 + + + + + + + + + + + + (+) trnR-ACG + + + + - - + - + - + + + rps7 + + + + + + + + + (+) + + + trnR-UCU + + + + - - + + + - + + + rps8 + + + + + + + + + (+) + + + trnS-GCU + + + + - - + + + - + + + rps11 + + + + + + + + + ! + + + trnS-GGA + + + + - - + + + - + + + rps12 + + + + + + + + + (+) + + + trnS-UGA + + + + - - + + + - + + + rps14 + + + + + + + (+) + + + + + trnT-GGU + + + + - - + + + - + + + rps15 + + + + - - + + + + + + + trnT-UGU + + + + - - + + + - + + + rps16 + - ! + - - - - - - ! + + trnV-GAC + + + + + - + (+) + - + + + rps18 + + + + + + + + + + + + + trnV-UAC + + + + - - + + + - + + + rps19 + + + + - + + + + + + - + trnW-CCA + + + + + + + + + + + + +               trnY-GUA + + + + + - + + + + + + +                                                                                                                                                                                                        46 Table 3. Species with fully assembled plastid genomes. Gent. = Gentianaceae; Poly. = Polygalaceae.          No. genes  No. No.   No. raw     with intact  rDNA tRNA  Family Species reads X-Cov1 Length (bp) LSC (bp) SSC (bp) IR (bp) reading frame2  genes genes ____________________________________________________________________________________________________________  Gent. Bartonia virginica 18,522,856 657.23 145,525 80,530 3,491 30,752  65 4 30 Gent. Exacum affine 10,824,214 1248.89 154,164 83,770 17,916 26,239  79 4 30 Gent. Exochaenium oliganthum 18,882,270 143.99 151,797 81,921 17,512 26,182  68 4 30 Gent. Obolaria virginica 17,075,844 152.25 145,825 79,411 10,014 28,158  64 4 30 Gent. Voyria clavata 14,927,268 917.83 31,724 18,603 9,987 1,567  17 4 4 Poly. Epirixanthes pallida 27,608,108 322.54 96,420 n.a. n.a. n.a.  29 4 30 Poly. Polygala arillata 9,171,790 355.35 164,747 83,668 8,743 36,168  76 4 30  1 Mean depth of coverage (based on remapping original reads to fully assembled plastome sequence) 2 Genes found in the inverted repeat are counted once        47 Table 4. Inverted repeat (IR) boundary shifts in eudicot mycoheterotrophs and autotrophic relatives. Following Zhu et al. (2015) the last full gene included in the IR at the SSC and LSC boundaries is indicated (genes that are partially duplicated in IR are not shown here, but see Fig.  1). Numbers in parentheses indicate the number of genes that have been expanded (exp.) into (+) or contracted (cont.) out from (-) the ancestral angiosperm IR boundaries, compared to autotrophic relatives.        IR/SSC  IR/LSC     boundary  boundary    (No. genes  (No. genes  Family Species  Trophic status exp./cont.)   exp./cont.)  _       __  Ericaceae Arbutus unedo autotroph trnL (+11)  trnI-CAU (-2) Ericaceae Orthilia secunda partial MH trnL (+4)  trnI-CAU (-2) Gentianaceae Exacum affine autotroph trnN-GUU  rpl22 (+2) Gentianaceae Obolaria virginica partial MH trnN-GUU  rpl2  Gentianaceae Bartonia virginica partial MH rps15 (+2)  rpl2  Gentianaceae Exochaenium oliganthum full MH trnN-GUU  rpl22 (+2) Gentianaceae Voyria clavata full MH rps11 (n.a.)  rps8 (n.a.) Polygalaceae Polygala arillata autotroph ndhI (+4)  rpl2 Polygalaceae Epirixanthes pallida full MH n.a.  n.a.      48 Figure 1. Linearized plastome maps of photosynthetic and mycoheterotrophic representatives of Ericaceae (the Arbutus unedo plastome is from Martínez-Alberola et al., 2013), Gentianaceae and Polygalaceae. Blue horizontal bars are 10 kb increments. The inverted repeat region is indicated with grey bars. Genes are colour-coded by function (see caption). Genes with introns are indicated with an asterisk (*), and putative pseudogenes in red text. The Orthilia secunda draft assembly is a single contig, split into two fragments to match the orientation of the other plastome maps; the arrow points to the true ends of the assembly. A 12 kb direct repeat in Epirixanthes pallida is indicated with blue bars.                 ! "#!!  50 Figure 2. Pairwise Mauve-based alignments of Nicotiana tabacum (NC_001879) with autotrophic representatives of Polygalaceae, Gentianaceae and Ericaceae. A linear map of the N. tabacum reference sequence appears first. A single copy of the inverted repeat region was included in each comparison. Coloured blocks are homologous regions with shared gene order between two or more genomes, referred to as ‘locally colinear blocks’ (LCB).  LCBs appearing above the central line are colinear and in the same orientation as the reference sequence. LCBs below align in reverse complement. Coloured lines link blocks of homology shared between taxa.                ! "#!!  52 Figure 3. Mauve-based alignments of Gentianaceae plastomes (a linear map of autotrophic Exacum affine appears first for reference; this genome is colinear with Nicotiana tabacum, see Fig. 2). A single copy of the inverted repeat region was included in each comparison. Coloured blocks are homologous regions with shared gene order between two or more genomes, referred to as ‘locally colinear blocks’ (LCB).  LCBs appearing above the central line are colinear and in the same orientation as the reference sequence. LCBs below align in reverse complement. Coloured lines link blocks of homology shared between taxa.                 ! "#!!  54 Figure 4. Mauve-based alignments of Polygalaceae and Ericaceae plastomes (a linear map of autotrophic Polygala arillata, Polygalaceae, or Arbutus unedo, Ericaceae, appears above the respective comparisons, for reference; these genomes are rearranged compared to Nicotiana tabacum, see Fig. 2). A single copy of the inverted repeat region was included in each comparison. Coloured blocks are homologous regions with shared gene order between two or more genomes, referred to as ‘locally colinear blocks’ (LCB).  LCBs appearing above the central line are colinear and in the same orientation as the reference sequence. LCBs below align in reverse complement. Coloured lines link blocks of homology shared between taxa. Polygalaceae: ‘a’ and ‘c’ are regions of P. arillata sequence that are deleted from E. pallida plastome; ‘b’ corresponds to the inverted repeat region of P. arillata, which appears in direct repeat in the E. pallida plastome. Ericaceae: ‘d’-‘g’ are intragenic regions that do not align under the set parameters.             ! ""!!  56 Figure 5. A portion of angiosperm phylogeny inferred in a likelihood analysis of 82 plastid coding regions using the “GxC” partitioning scheme based on an ‘ORF-only’ alignment (see text and Table S3 for details); this portion of the tree shows rosid relationships. Eudicot families where mycoheterotrophy has evolved are indicated in blue. Log likelihood score of best tree: -1,506,318.267. Bootstrap support values are indicated beside branches; thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. The scale bar indicates estimated substitutions per site.                  ! "#!!  58 Figure 6. A portion of angiosperm phylogeny inferred in a likelihood analysis of 82 plastid coding regions using the “GxC” partitioning scheme based on an ‘ORF-only’ alignment (see text and Table S3 for details); this portion of the tree shows asterid relationships. Eudicot families where mycoheterotrophy has evolved are indicated in blue. Log likelihood score of best tree: -1,506,318.267.  Bootstrap support values are indicated beside branches; thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. The scale bar indicates estimated substitutions per site.                  ! "#!!  60 Bibliography ALKATIB, S., T.T. FLEISCHMANN, L.B. SCHARFF, and R. BOCK. 2012. Evolutionary constraints on the plastid tRNA set decoding methionine and isoleucine. Nucleic Acids Research 40: 6713–6724. ALTSCHUL, S.F., W. GISH, T. PENNSYLVANIA, and U. PARK. 1990. Basic Local Alignment Search Tool. Journal of Molecular Biology 215: 403–410. BARBROOK, A.C., C.J. HOWE, and S. PURTON. 2006. Why are plastid genomes retained in non-photosynthetic organisms? Trends in Plant Science 11: 101–108. BARRETT, C.F., and J.I. DAVIS. 2012. The plastid genome of the mycoheterotrophic Corallorhiza striata (Orchidaceae) is in the relatively early stages of degradation. American Journal of Botany 99: 1513–1523. BARRETT, C.F., J. V. FREUDENSTEIN, J. LI, D.R. MAYFIELD-JONES, L. PEREZ, J.C. PIRES, and C. SANTOS. 2014. Investigating the path of plastid genome degradation in an early-transitional clade of heterotrophic orchids, and implications for heterotrophic angiosperms. Molecular Biology and Evolution 31: 3095–3112. BEATTY, G.E., M. PHILIPP, and J. PROVAN. 2010. Unidirectional hybridization at a species’ range boundary: implications for habitat tracking. Diversity and Distributions 16: 1–9. BEATTY, G.E., and J. PROVAN. 2010. Refugial persistence and postglacial recolonization of North America by the cold-tolerant herbaceous plant Orthilia secunda. Molecular Ecology 19: 5009–5021. BELLO, M.A., P.J. RUDALL, and J. A. HAWKINS. 2012. Combined phylogenetic analyses reveal interfamilial relationships and patterns of floral evolution in the eudicot order Fabales. Cladistics 28: 393–421.   61 BERG, S., K. KRAUSE, and K. KRUPINSKA. 2004. The rbcL genes of two Cuscuta species, C. gronovii and C. subinclusa, are transcribed by the nuclear-encoded plastid RNA polymerase (NEP). Planta 219: 541–6. BIDARTONDO, M.I. 2005. The evolutionary ecology of myco-heterotrophy. New Phytologist 167: 335–352. BIDARTONDO, M.I., and T.D. BRUNS. 2001. Extreme specificity in epiparasitic Monotropoideae (Ericaceae): widespread phylogenetic and geographical structure. Molecular Ecology 10: 2285–2295. BIDARTONDO, M.I., B. BURGHARDT, G. GEBAUER, T.D. BRUNS, and D.J. READ. 2004. Changing partners in the dark: isotopic and molecular evidence of ectomycorrhizal liaisons between forest orchids and trees. Proceedings of the Royal Society B: Biological Sciences 271: 1799–1806. BLAZIER, C.C., M.M. GUISINGER, and R.K. JANSEN. 2011. Recent loss of plastid-encoded ndh genes within Erodium (Geraniaceae). Plant Molecular Biology 76: 263–272. BOHNE, A., A. WEIHE, and T. BÖRNER. 2009. Transfer RNAs inhibit Arabidopsis phage-type RNA polymerases. 63–69. BRAUKMANN, T., M. KUZMINA, and S. STEFANOVI!. 2013. Plastid genome evolution across the genus Cuscuta (Convolvulaceae): Two clades within subgenus Grammica exhibit extensive gene loss. Journal of Experimental Botany 64: 977–989. BRAUKMANN, T., and S. STEFANOVI!. 2012. Plastid genome evolution in mycoheterotrophic Ericaceae. Plant Molecular Biology 79: 5–20. BRAUKMANN, T.W.A., M. KUZMINA, and S. STEFANOVI!. 2009. Loss of all plastid ndh genes in Gnetales and conifers: extent and evolutionary significance for the seed plant phylogeny.   62 Current Genetics 55: 323–337. BUNGARD, R. A. 2004. Photosynthetic evolution in parasitic plants: Insight from the chloroplast genome. BioEssays 26: 235–247. CAI, Z., M. GUISINGER, H.G. KIM, E. RUCK, J.C. BLAZIER, V. MCMURTRY, J. V. KUEHL, ET AL. 2008. Extensive reorganization of the plastid genome of Trifolium subterraneum (Fabaceae) is associated with numerous repeated sequences and novel DNA insertions. Journal of Molecular Evolution 67: 696–704. CAMERON, D.D., and J.F. BOLIN. 2010. Isotopic evidence of partial mycoheterotrophy in the Gentianaceae: Bartonia virginica and Obolaria virginica as case studies. American Journal of Botany 97: 1272–1277. COSNER, M.E., R.K. JANSEN, J.D. PALMER, and S.R. DOWNIE. 1997. The highly rearranged chloroplast genome of Trachelium caeruleum (Campanulaceae): multiple inversions, inverted repeat expansion and contraction, transposition, insertions/deletions, and several repeat families. Current Genetics 31: 419–429. CRONN, R., A. LISTON, M. PARKS, D.S. GERNANDT, R. SHEN, and T. MOCKLER. 2008. Multiplex sequencing of plant chloroplast genomes using Solexa sequencing-by-synthesis technology. Nucleic Acids Research 36: e122. CRONQUIST, A. 1988. The Evolution and Classification of Flowering Plants. 2nd ed. The New York Botanical Garden, Bronx, New York. CUMMINGS, M.P., and N.A. WELSCHMEYER. 1998. Pigment composition of putatively achlorophyllous angiosperms. Plant Systematics and Evolution 210: 105–111. DARLING, A.C.E., B. MAU, F.R. BLATTNER, and N.T. PERNA. 2004. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome Research 14: 1394–403.   63 DELANNOY, E., S. FUJII, C. COLAS DES FRANCS-SMALL, M. BRUNDRETT, and I. SMALL. 2011. Rampant gene loss in the underground orchid Rhizanthella gardneri highlights evolutionary constraints on plastid genomes. Molecular Biology and Evolution 28: 2077–2086. DELAVAULT, P., V. SAKANYAN, and P. THALOUARN. 1995. Divergent evolution of two plastid genes, rbcL and atpB, in a non-photosynthetic parasitic plant. Plant Molecular Biology 29: 1071–9. DOWNIE, S.R., D.S. KATZ-DOWNIE, K.H. WOLFE, P.J. CALIE, and J.D. PALMER. 1994. Internal plasticity and multiple gene loss during angiosperm evolution. Current Genetics 25: 367–378. DOWNIE, S.R., and J.D. PALMER. 1992. Use of chloroplast DNA rearrangments in reconstructing plant phylogeny. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular Systematics of Plants, 14–35. Springer US, New York. DOYLE, J., and J. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. The Phytochemical Bulletin 19: 11–15. DRESCHER, A., R. STEPHANIE, T. CALSA, H. CARRER, and R. BOCK. 2000. The two largest chloroplast genome-encoded open reading frames of higher plants are essential genes. Plant Journal 22: 97–104. ERNES, M.J., and H.E. NEUHAUS. 2005. Metabolism and transport in non-photosynthetic plastids. Journal of Experimental Botany 48: 1995–2005. FAJARDO, D., D. SENALIK, M. AMES, H. ZHU, S. A. STEFFAN, R. HARBUT, J. POLASHOCK, ET AL. 2013. Complete plastid genome sequence of Vaccinium macrocarpon: Structure, gene content, and rearrangements revealed by next generation sequencing. Tree Genetics and Genomes 9: 489–498.   64 FELSENSTEIN, J. 1978. Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zoology 27: 401–410. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. FREYER, R., M.C. KIEFER-MEYER, and H. KÖSSEL. 1997. Occurrence of plastid RNA editing in all major lineages of land plants. Proceedings of the National Academy of Sciences of the United States of America 94: 6285–6290. FUNK, H.T., S. BERG, K. KRUPINSKA, U.G. MAIER, and K. KRAUSE. 2007. Complete DNA sequences of the plastid genomes of two parasitic flowering plant species, Cuscuta reflexa and Cuscuta gronovii. BMC Plant Biology 7: 45. GEBAUER, G., and M. MEYER. 2003. 15N and 13C natural abundance of autotrophic and myco-heterotrophic orchids provides insight into nitrogen and carbon gain from fungal association. New Phytologist 160: 209–223. GRAHAM, S.W., and R.G. OLMSTEAD. 2000. Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms. American Journal of Botany 87: 1712–1730. GRAHAM, S.W., P. A. REEVES, A.C.E. BURNS, and R.G. OLMSTEAD. 2000. Microstructural changes in noncoding chloroplast DNA: interpretation, evolution, and utility of indels and inversions in basal angiosperm phylogenetic inference. International Journal of Plant Sciences 161: S83–S96. GRAY, B.N., B. A AHNER, and M.R. HANSON. 2009. Extensive homologous recombination between introduced and native regulatory plastid DNA elements in transplastomic plants. Transgenic Research 18: 559–572. HABERLE, R.C., H.M. FOURCADE, J.L. BOORE, and R.K. JANSEN. 2008. Extensive rearrangements   65 in the chloroplast genome of Trachelium caeruleum are associated with repeats and tRNA genes. Journal of Molecular Evolution 66: 350–361. HAJDUKIEWICZ, P.T., L.A. ALLISON, and P. MALIGA. 1997. The two RNA polymerases encoded by the nuclear and the plastid compartments transcribe distinct groups of genes in tobacco plastids. The EMBO Journal 16: 4041–8. HANAOKA, M., K. KANAMARU, M. FUJIWARA, H. TAKAHASHI, and K. TANAKA. 2005. Glutamyl-tRNA mediates a switch in RNA polymerase use during chloroplast biogenesis. EMBO Reports 6: 545–550. HARRIS, E.H., J.E. BOYNTON, and N.W. GILLHAM. 1994. Chloroplast ribosomes and protein synthesis. Microbiological Reviews 58: 700–754. HENDY, M.D., and D. PENNY. 1989. A framework for the quantitative study of evolutionary trees. Systematic Zoology 38: 297–309. HOFFMANN, S., C. OTTO, S. KURTZ, C.M. SHARMA, P. KHAITOVICH, J. VOGEL, P.F. STADLER, and J. HACKERMÜLLER. 2009. Fast mapping of short sequences with mismatches, insertions and deletions using index structures. PLoS Computational Biology 5: e1000502. HOWE, C.J., and S. PURTON. 2007. The little genome of apicomplexan plastids: its raison d’etre and a possible explanation for the “delayed death” phenomenon. Protist 158: 121–133. HYNSON, N. A, and T.D. BRUNS. 2009. Evidence of a myco-heterotroph in the plant family Ericaceae that lacks mycorrhizal specificity. Proceedings of the Royal Society B: Biological Sciences 276: 4053–4059. HYNSON, N. A, K. PREISS, G. GEBAUER, and T.D. BRUNS. 2009. Isotopic evidence of full and partial myco-heterotrophy in the plant tribe Pyroleae (Ericaceae). The New phytologist 182: 719–26.   66 ILES, W.J.D., S.Y. SMITH, and S.W. GRAHAM. 2013. A well-supported phylogenetic framework for the monocot order Alismatales reveals multiple losses of the plastid NADH dehydrogenase complex and a strong long branch effect. In P. Wilkin, and S. J. Mayo [eds.], Early Events in Monocot Evolution, 1–19. Cambridge University Press. JAHN, D., E. VERKAMP, and D. SÖLL. 1992. Glutamyl-transfer RNA: a precursor of heme and chlorophyll biosynthesis. Trends in Biochemical Sciences 17: 215–218. JANOU!KOVEC, J., D. V. TIKHONENKOV, F. BURKI, A.T. HOWE, M. KOLÍSKO, A.P. MYLNIKOV, and P.J. KEELING. 2015. Factors mediating plastid dependency and the origins of parasitism in apicomplexans and their close relatives. Proceedings of the National Academy of Sciences 112: 10200–10207. JANSEN, R.K., Z. CAI, L. A RAUBESON, H. DANIELL, C.W. DEPAMPHILIS, J. LEEBENS-MACK, K.F. MÜLLER, ET AL. 2007. Analysis of 81 genes from 64 plastid genomes resolves relationships in angiosperms and identifies genome-scale evolutionary patterns. Proceedings of the National Academy of Sciences of the United States of America 104: 19369–19374. JANSEN, R.K., C. SASKI, S.-B. LEE, A.K. HANSEN, and H. DANIELL. 2011. Complete plastid genome sequences of three Rosids (Castanea, Prunus, Theobroma): evidence for at least two independent transfers of rpl22 to the nucleus. Molecular Biology and Evolution 28: 835–47. KATOH, K., K. MISAWA, K. KUMA, and T. MIYATA. 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Research 30: 3059–3066. KAWAKAMI, K., Y. UMENA, M. IWAI, Y. KAWABATA, M. IKEUCHI, N. KAMIYA, and J.-R. SHEN. 2011. Roles of PsbI and PsbM in photosystem II dimer formation and stability studied by deletion mutagenesis and X-ray crystallography. Biochimica et Biophysica Acta (BBA) -   67 Bioenergetics 1807: 319–325. KIKUCHI, S., J. BÉDARD, M. HIRANO, Y. HIRABAYASHI, M. OISHI, M. IMAI, M. TAKASE, ET AL. 2013. Uncovering the protein translocon at the chloroplast inner envelope membrane. Science 339: 571–4. KIM, H.T., J.S. KIM, M.J. MOORE, K.M. NEUBIG, N.H. WILLIAMS, W.M. WHITTEN, and J.-H. KIM. 2015. Seven new complete plastome sequences reveal rampant independent loss of the ndh gene family across orchids and associated instability of the inverted repeat/small single-copy region boundaries. Plos ONE 10: e0142215. KIM, J., A. RUDELLA, V. RAMIREZ RODRIGUEZ, B. ZYBAILOV, P.D.B. OLINARES, and K.J. VAN WIJK. 2009. Subunits of the plastid ClpPR protease complex have differential contributions to embryogenesis, plastid biogenesis, and plant development in Arabidopsis. The Plant Cell 21: 1669–1692. KISSLING, J. 2012. Taxonomy of Exochaenium and Lagenias: two resurrected genera of tribe Exaceae (Gentianaceae). Systematic Botany 37: 238–253. KNAUF, U., and W. HACHTEL. 2002. The genes encoding subunits of ATP synthase are conserved in the reduced plastid genome of the heterotrophic alga Prototheca wickerhamii. Molecular Genetics and Genomics 267: 492–7. KRAUSE, K. 2008. From chloroplasts to “cryptic” plastids: evolution of plastid genomes in parasitic plants. Current Genetics 54: 111–121. KRAUSE, K. 2012. Plastid genomes of parasitic plants: A trail of reductions and losses. In Bullerwell C.E. [ed.], Organelle Genetics, 79–103. Springer-Verlag, Berlin Heidelberg. KRON, K.A., W.S. JUDD, P.F. STEVENS, D.M. CRAYN, and A.A. ANDERBERG. 2002. Phylogenetic classification of Ericaceae: molecular and morpholgical evidence. The Botanical Review 68:   68 335–423. KUGITA, M., Y. YAMAMOTO, T. FUJIKAWA, T. MATSUMOTO, and K. YOSHINAGA. 2003. RNA editing in hornwort chloroplasts makes more than half the genes functional. Nucleic Acids Research 31: 2417–2423. KURODA, H., and P. MALIGA. 2003. The plastid clpP1 protease gene is essential for plant development. Nature 425: 86–89. LAM, V.K.Y., M. SOTO GOMEZ, and S.W. GRAHAM. 2015. The highly reduced plastome of mycoheterotrophic Sciaphila (Triuridaceae) is colinear with its green relatives and is under strong purifying selection. Genome Biology and Evolution 7: 2220–2236. LANFEAR, R., B. CALCOTT, S.Y.W. HO, and S. GUINDON. 2012. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution 29: 1695–1701. LANFEAR, R., B. CALCOTT, D. KAINER, C. MAYER, and A. STAMATAKIS. 2014. Selecting optimal partitioning schemes for phylogenomic datasets. BMC Evolutionary Biology 14: 82. LEAKE, J.R., and D.D. CAMERON. 2010. Physiological ecology of mycoheterotrophy. New Phytologist 185: 601–605. LEEBENS-MACK, J., and C. DEPAMPHILIS. 2002. Power analysis of tests for loss of selective constraint in cave crayfish and nonphotosynthetic plant lineages. Molecular biology and evolution 19: 1292–1302. LIERE, K., A. WEIHE, and T. BÖRNER. 2011. The transcription machineries of plant mitochondria and chloroplasts: composition, function, and regulation. Journal of Plant Physiology 168: 1345–1360. LOGACHEVA, M.D., M.I. SCHELKUNOV, M.S. NURALIEV, T.H. SAMIGULLIN, and A. A. PENIN.   69 2014. The plastid genome of mycoheterotrophic monocot Petrosavia stellaris exhibits both gene losses and multiple rearrangements. Genome Biology and Evolution 6: 238–246. LOGACHEVA, M.D., M.I. SCHELKUNOV, and A. A. PENIN. 2011. Sequencing and analysis of plastid genome in mycoheterotrophic orchid Neottia nidus-avis. Genome Biology and Evolution 3: 1296–1303. LOHSE, M., O. DRECHSEL, S. KAHLAU, and R. BOCK. 2013. OrganellarGenomeDRAW--a suite of tools for generating physical maps of plastid and mitochondrial genomes and visualizing expression data sets. Nucleic Acids Research 41: 1–7. MACHADO, M. A., and K. ZETSCHE. 1990. A structural, functional and molecular analysis of plastids of the holoparasites Cuscuta reflexa and Cuscuta europaea. Planta 181: 91–96. MADDISON, W.P., and D.R. MADDISON. 2014. Mesquite: a modular system for evolutionary analyses. Version 3.03+. Available at: MADDISON, W.P., and D.R. MADDISON. 2015. Mesquite: a modular system for evolutionary analyses. Version 3.04. Available at: MARÉCHAL, A., and N. BRISSON. 2010. Recombination and the maintenance of plant organelle genome stability. New Phytologist 186: 299–317. MARTÍN, M., and B. SABATER. 2010. Plastid ndh genes in plant evolution. Plant Physiology and Biochemistry 48: 636–645. MARTÍNEZ-ALBEROLA, F., E.M. DEL CAMPO, D. LÁZARO-GIMENO, S. MEZQUITA-CLARAMONTE, A. MOLINS, I. MATEU-ANDRÉS, J. PEDROLA-MONFORT, ET AL. 2013. Balanced gene losses, duplications and intensive rearrangements led to an unusual regularly sized genome in Arbutus unedo chloroplasts. PLoS ONE. http://doi:10.1371/journal.pone.0079685 MCNEAL, J.R., K. ARUMUGUNATHAN, J. V KUEHL, J.L. BOORE, and C.W. DEPAMPHILIS. 2007.   70 Systematics and plastid genome evolution of the cryptically photosynthetic parasitic plant genus Cuscuta (Convolvulaceae). BMC Biology 5: 55. MCNEAL, J.R., J. V. KUEHL, J.L. BOORE, J. LEEBENS-MACK, and C.W. DEPAMPHILIS. 2009. Parallel loss of plastid introns and their maturase in the genus Cuscuta. PLoS ONE 4: 1–8. MENNES, C.B., V.K.Y. LAM, P.J. RUDALL, S.P. LYON, S.W. GRAHAM, E.F. SMETS, and V.S.F.T. MERCKX. 2015. Ancient Gondwana break-up explains the distribution of the mycoheterotrophic family Corsiaceae (Liliales). Journal of Biogeography 42: 1123–1136. MENNES, C.B., M.S. MOERLAND, M. RATH, E.F. SMETS, and V.S.F.T. MERCKX. 2015. Evolution of mycoheterotrophy in Polygalaceae: the case of Epirixanthes. American Journal of Botany 102: 598–608. MERCKX, V., F.T. BAKKER, S. HUYSMANS, and E. SMETS. 2009. Bias and conflict in phylogenetic inference of myco-heterotrophic plants: a case study in Thismiaceae. Cladistics 25: 64–77. MERCKX, V., and J. V FREUDENSTEIN. 2010. Evolution of mycoheterotrophy in plants: a phylogenetic perspective. New Phytologist 185: 605–609. MERCKX, V.S.F.T. 2013. Mycoheterotrophy: an introduction. In V. S. F. T. Merckx [ed.], Mycoheterotrophy: The Biology of Plants Living on Fungi, 1–17. Springer Science and Business Media, New York. MERCKX, V.S.F.T., J. V. FREUDENSTEIN, B. CRANDALL-STOTLER, M.J. CHRISTENHUSZ, R.E. STOTLER, N.J. WICKETT, H. MAAS-VAN DE KAMER, and P.J.M. MAAS. 2013. Taxonomy and Classification. In V. S. F. T. Merckx [ed.], Mycoheterotrophy:The Biology of Plants Living on Fungi, 19–101. Springer Science and Business Media, New York. MERCKX, V.S.F.T., J. KISSLING, H. HENTRICH, S.B. JANSSENS, C.B. MENNES, C.D. SPECHT, and E.F. SMETS. 2013. Phylogenetic relationships of the mycoheterotrophic genus Voyria and   71 the implications for the biogeographic history of Gentianaceae. American Journal of Botany 100: 712–721. MILLEN, R.S., R.G. OLMSTEAD, K.L. ADAMS, J.D. PALMER, N.T. LAO, L. HEGGIE, T. A KAVANAGH, ET AL. 2001. Many parallel losses of infA from chloroplast DNA during angiosperm evolution with multiple independent transfers to the nucleus. The Plant Cell 13: 645–658. MILLER, M. A, W. PFEIFFER, and T. SCHWARTZ. 2010. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. 2010 Gateway Computing Environments Workshop, GCE 2010:1–8. MILLIGAN, B.G., J.N. HAMPTON, and J.D. PALMER. 1989. Dispersed repeats and structural reorganization in subclover chloroplast DNA. Molecular Biology and Evolution 6: 355–368. MOLINA, J., K.M. HAZZOURI, D. NICKRENT, M. GEISLER, R.S. MEYER, M.M. PENTONY, J.M. FLOWERS, ET AL. 2014. Possible loss of the chloroplast genome in the parasitic flowering plant Rafflesia lagascae (Rafflesiaceae). Molecular Biology and Evolution 31: 793–803. NAKAI, M. 2015. YCF1: A green TIC: response to the de Vries et al. commentary. The Plant cell 27: 1834–1838. OHLROGGE, J., and J. BROWSE. 1995. Lipid biosynthesis. Plant Cell 7: 957–970. OSMOND, C.B., T. AKAZAWA, and H. BEEVERS. 1975. Localization and properties of ribulose diphosphate carboxylase from castor bean endosperm. Plant Physiology 55: 226–230. PALMER, J.D. 1985. Comparative organization of chloroplast genomes. Annual Review of Genetics 19: 325–354. PALMER, J.D., B. OSORIO, J. ALDRICH, and W.F. THOMPSON. 1987. Chloroplast DNA evolution among legumes: loss of a large inverted repeat occurred prior to other sequence   72 rearrangements. Current Genetics 11: 275–286. PALMER, J.D., and D.B. STEIN. 1986. Conservation of chloroplast genome structure among vascular plants. Current Genetics 10: 823–833. PELTIER, G., and L. COURNAC. 2002. Chlororespiration. Annual Review of Plant Biology 53: 523–550. PEREDO, E.L., U.M. KING, and D.H. LES. 2013. The plastid genome of Najas flexilis: adaptation to submersed environments is accompanied by the complete loss of the NDH complex in an aquatic angiosperm. PloS ONE 8: e68591. PETERSEN, G., A. CUENCA, and O. SEBERG. 2015. Plastome evolution in hemiparasitic mistletoes. Genome Biology and Evolution 7: 2520–2532. RANDLE, C.P., and A.D. WOLFE. 2005. The evolution and expression of rbcL in holoparasitic sister-genera Harveya and Hyobanche (Orobanchaceae). American Journal of Botany 92: 1575–1585. ROGALSKI, M., D. KARCHER, and R. BOCK. 2008. Superwobbling facilitates translation with reduced tRNA sets. Nature, Structural & Molecular Biology 15: 192–198. ROSS, T.G., C.F. BARRETT, M. SOTO, V.K.Y. LAM, C.L. HENRIQUEZ, D.H. LES, J.I. DAVIS, ET AL. 2015. Plastid phylogenomics and molecular evolution of Alismatales. Cladistics1–19. ROUSSEAU-GUEUTIN, M., X. HUANG, E. HIGGINSON, M. AYLIFFE, A. DAY, and J.N. TIMMIS. 2013. Potential functional replacement of the plastidic acetyl-CoA carboxylase subunit (accD) gene by recent transfers to the nucleus in some angiosperm lineages. Plant Physiology 161: 1918–29. RUHFEL, B.R., M. A GITZENDANNER, P.S. SOLTIS, D.E. SOLTIS, and J.G. BURLEIGH. 2014. From algae to angiosperms - inferring the phylogeny of green plants (Viridiplantae) from 360   73 plastid genomes. BMC Evolutionary Biology 14: 23. RUHLMAN, T. A, W.-J. CHANG, J.J. CHEN, Y.-T. HUANG, M.-T. CHAN, J. ZHANG, D.-C. LIAO, ET AL. 2015. NDH expression marks major transitions in plant evolution and reveals coordinate intracellular gene loss. BMC Plant Biology 15: 1–9. RUIZ-NIETO, J.E., C.L. AGUIRRE-MANCILLA, J.A. ACOSTA-GALLEGOS, J.C. RAYA-PÉREZ, E. PIEDRA-IBARRA, J. VÁZQUEZ-MEDRANO, and V. MONTERO-TAVERA. 2015. Photosynthesis and chloroplast genes are involved in water-use efficiency in common bean. Plant Physiology and Biochemistry 86: 166–173. SAARELA, J.M., and S.W. GRAHAM. 2010. Inference of phylogenetic relationships among the subfamilies of grasses (Poaceae: Poales) using meso-scale exemplar-based sampling of the plastid genome. Botany 88: 65–84. SABIR, J., E. SCHWARZ, N. ELLISON, J. ZHANG, N. A BAESHEN, M. MUTWAKIL, R. JANSEN, and T. RUHLMAN. 2014. Evolutionary and biotechnology implications of plastid genome variation in the inverted-repeat-lacking clade of legumes. Plant Biotechnology Journal 12: 743–54. SAINT-MARCOUX, D., F.A. WOLLMAN, and C. DE VITRY. 2009. Biogenesis of cytochrome b6 in photosynthetic membranes. Journal of Cell Biology 185: 1195–1207. SANDERSON, M.J., D. COPETTI, A. BURQUEZ, E. BUSTAMANTE, J.L.M. CHARBONEAU, L.E. EGUIARTE, S. KUMAR, ET AL. 2015. Exceptional reduction of the plastid genome of saguaro cactus (Carnegiea gigantea): loss of the ndh gene suite and inverted repeat. American Journal of Botany 102: 1115–1127. SASAKI, Y., and Y. NAGANO. 2004. Plant acetyl-CoA carboxylase: structure, biosynthesis, regulation, and gene manipulation for plant breeding. Bioscience, Biotechnology, and Biochemistry 68: 1175–1184.   74 SATO, S. 2011. The apicomplexan plastid and its evolution. Cellular and Molecular Life Sciences 68: 1285–1296. SCHELKUNOV, M., V. SHTRATNIKOVA, M. NURALIEV, M. -A. SELOSSE, A. PENIN, and M. LOGACHEVA. 2015. Exploring the limits for reduction of plastid genomes: a case study of the mycoheterotrophic orchids Epipogium aphyllum and Epipogium roseum. Genome Biology and Evolution evv019: 1–29. SCHWARZ, E.N., T.A. RUHLMAN, J.S.M. SABIR, N.H. HAJRAH, N.S. ALHARBI, A.L. AL-MALKI, C.D. BAILEY, and R.K. JANSEN. 2015. Plastid genome sequences of legumes reveal parallel inversions and multiple losses of rps16 in papilionoids. Journal of Systematics and Evolution 53: 458–468. SCHWENDER, J., F. GOFFMAN, J.B. OHLROGGE, and Y. SHACHAR-HILL. 2004. Rubisco without the Calvin cycle improves the carbon efficiency of developing green seeds. Nature 432: 779–782. SHIKANAI, T. 2015. Chloroplast NDH: A different enzyme with a structure similar to that of respiratory NADH dehydrogenase. Biochimica et Biophysica Acta. SILVESTRO, D., and I. MICHALAK. 2012. raxmlGUI: a graphical front-end for RAxML. Organisms Diversity & Evolution 12: 335–337. STAMATAKIS,  A. 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690. STAMATAKIS, A., P. HOOVER, and J. ROUGEMONT. 2008. A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology 57: 758–771. STANNE, T.M., L.L.E. SJÖGREN, S. KOUSSEVITZKY, and A.K. CLARKE. 2009. Identification of   75 new protein substrates for the chloroplast ATP-dependent Clp protease supports its constitutive role in Arabidopsis. The Biochemical Journal 417: 257–68. STRAUB, S.C.K., M. FISHBEIN, T. LIVSHULTZ, Z. FOSTER, M. PARKS, K. WEITEMIER, R.C. CRONN, and A. LISTON. 2011. Building a model: developing genomic resources for common milkweed (Asclepias syriaca) with low coverage genome sequencing. BMC Genomics 12: 211. STRUWE, L. 2014. Classification and evolution in the family Gentianaceae. In J. J. Rybcynski, M. R. Davey, and A. Mikula [eds.], The Gentianaceae - Vol. 1: Characterization and Ecology, 1–12. Springer-Verlag, Berlin Heidelberg. SUGIURA, M. 2014. Plastid mRNA translation. In P. Maliga [ed.], Chloroplast Biotechnology: Methods and Protocols, Methods in Molecular Biology, 73–91. Springer Science and Business Media, New York. TEDERSOO, L., P. PELLET, U. KÕLJALG, and M. A. SELOSSE. 2007. Parallel evolutionary paths to mycoheterotrophy in understorey Ericaceae and Orchidaceae: ecological evidence for mixotrophy in Pyroleae. Oecologia 151: 206–217. THIERRY, A., C. BOUCHIER, B. DUJON, and G.-F. RICHARD. 2008. Megasatellites: a peculiar class of giant minisatellites in genes involved in cell adhesion and pathogenicity in Candida glabrata. Nucleic Acids Research 36: 5970–82. THIERS, B. 2015. Index Herbariorium: A global directory of public herbaria and associated staff. New York Botanical Garden’s Virtual Herbarium. Available at: THOMPSON, J.D., D.G. HIGGINS, and T.J. GIBSON. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific   76 gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673–4680. TOLBERT, N.E. 1997. The C2 oxidative photosynthetic carbon cycle. Annual Review of Plant Physiology and Plant Molecular Biology 48: 1–25. UMATE, P., S. SCHWENKERT, I. KARBAT, C. DAL BOSCO, L. MLC! OCHVÁ , S. VOLZ, H. ZER, ET AL. 2007. Deletion of PsbM in tobacco alters the QB site properties and the electron flow within photosystem II. Journal of Biological Chemistry 282: 9758–9767. DE VRIES, J., F.L. SOUSA, B. BÖLTER, J. SOLL, and S.B. GOULD. 2015. YCF1: A Green TIC? The Plant Cell 27: 1827–1833. WALKER, J.E. 2013. The ATP synthase: the understood, the uncertain and the unknown. Biochemical Society Transactions 41: 1–16. WICKE, S., K.F. MÜLLER, C.W. DE PAMPHILIS, D. QUANDT, N.J. WICKETT, Y. ZHANG, S.S. RENNER, and G.M. SCHNEEWEISS. 2013. Mechanisms of functional and physical genome reduction in photosynthetic and nonphotosynthetic parasitic plants of the broomrape family. The Plant Cell 25: 3711–25. WICKE, S., B. SCHÄFERHOFF, C.W. DEPAMPHILIS, and K.F. MÜLLER. 2014. Disproportional plastome-wide increase of substitution rates and relaxed purifying selection in genes of carnivorous Lentibulariaceae. Molecular Biology and Evolution 31: 529–545. WICKE, S., G.M. SCHNEEWEISS, C.W. DEPAMPHILIS, K.F. MÜLLER, and D. QUANDT. 2011. The evolution of the plastid chromosome in land plants: gene content, gene order, gene function. Plant Molecular Biology 76: 273–297. WICKETT, N.J., Y. ZHANG, S.K. HANSEN, J.M. ROPER, J. V. KUEHL, S. A. PLOCK, P.G. WOLF, ET AL. 2008. Functional gene losses occur with minimal size reduction in the plastid genome of the parasitic liverwort Aneura mirabilis. Molecular Biology and Evolution 25: 393–401.   77 WOLFE, K.H. 1994. Similarity between putative ATP-binding sites in land plant plastid ORF2280 proteins and the FtsH/CDC48 family of ATPases. Current Genetics 25: 379–383. WOLFE, K.H., C.W. MORDEN, and J.D. PALMER. 1992. Function and evolution of a minimal plastid genome from a nonphotosynthetic parasitic plant. Proceedings of the National Academy of Sciences of the United States of America 89: 10648–10652. WU, C.S., Y.T. LAI, C.P. LIN, Y.N. WANG, and S.M. CHAW. 2009. Evolution of reduced and compact chloroplast genomes (cpDNAs) in gnetophytes: selection toward a lower-cost strategy. Molecular Phylogenetics and Evolution 52: 115–124. WU, F.-H., M.-T. CHAN, D.-C. LIAO, C.-T. HSU, Y.-W. LEE, H. DANIELL, M.R. DUVALL, and C.-S. LIN. 2010. Complete chloroplast genome of Oncidium Gower Ramsey and evaluation of molecular markers for identification and breeding in Oncidiinae. BMC Plant Biology 10: 68. WYMAN, S.K., R.K. JANSEN, and J.L. BOORE. 2004. Automatic annotation of organellar genomes with DOGMA. Bioinformatics 20: 3252–3255. XIE, Z., and S. MERCHANT. 1996. The plastid-encoded ccsA gene is required for heme attachment to chloroplast c-type cytochromes. Journal of Biological Chemistry 271: 4632–4639. XU, Q., D. HOPPE, V.P. CHITNIS, W.R. ODOM, J. A GUIKEMA, and P.R. CHITNIS. 1995. Mutational analysis of photosystem I polypeptides in the cyanobacterium Synechocystis sp. PCC 6803. Targeted inactivation of psaI reveals the function of psaI in the structural organization of psaL. The Journal of Biological Chemistry 270: 16243–50. YAGI, Y., and T. SHIINA. 2014. Recent advances in the study of chloroplast gene expression and its evolution. Frontiers in Plant Science 5: 61. YANG, J.-B., M. TANG, H.-T. LI, Z.-R. ZHANG, and D.-Z. LI. 2013. Complete chloroplast genome   78 of the genus Cymbidium: lights into the species identification, phylogenetic implications and population genetic analyses. BMC Evolutionary Biology 13: 84. ZGURSKI, J.M., H.S. RAI, Q.M. FAI, D.J. BOGLER, J. FRANCISCO-ORTEGA, and S.W. GRAHAM. 2008. How well do we understand the overall backbone of cycad phylogeny? New insights from a large, multigene plastid data set. Molecular Phylogenetics and Evolution 47: 1232–1237. ZHELYAZKOVA, P., C.M. SHARMA, K.U. FORSTNER, K. LIERE, J. VOGEL, and T. BORNER. 2012. The primary transcriptome of barley chloroplasts: numerous noncoding RNAs and the dominating role of the plastid-encoded RNA polymerase. The Plant Cell 24: 123–136. ZHU, A., W. GUO, S. GUPTA, W. FAN, and J.P. MOWER. 2015. Evolutionary dynamics of the plastid inverted repeat: the effects of expansion, contraction, and loss on substitution rates. New Phytologist. ZIMMER, K., N. A. HYNSON, G. GEBAUER, E.B. ALLEN, M.F. ALLEN, and D.J. READ. 2007. Wide geographical and ecological distribution of nitrogen and carbon gains from fungi in pyroloids and monotropoids (Ericaceae) and in orchids. New Phytologist 175: 166–175. ZIMMER, K., C. MEYER, and G. GEBAUER. 2008. The ectomycorrhizal specialist orchid Corallorhiza trifida is a partial myco-heterotroph. New Phytologist 178: 395–400. ZOSCHKE, R., M. NAKAMURA, K. LIERE, M. SUGIURA, T. BÖRNER, and C. SCHMITZ-LINNEWEBER. 2010. An organellar maturase associates with multiple group II introns. Proceedings of the National Academy of Sciences of the United States of America 107: 3245–3250.    79 Appendices Supplementary Table S1. Accession information for publically available plastomes included in the angiosperm matrix (for all others, see Ruhfel et al., 2014).   Order Family Species Accession #  ______________________________________________________________________________  Apiales Araliaceae Aralia undulata Hand.-Mazz NC_022810  Asterales Asteraceae Artemisia frigida Eichw. NC_020607  Asterales Asteraceae Centaurea diffusa Lam. NC_024286  Asterales Campanulaceae Campanula takesimana Nakai NC_026203  Asterales Campanulaceae Hanabusaya asiatica (Nakai) NC_024732   Nakai   Brassicales Brassicaceae Raphanus sativus L. NC_024469  Caryophyllales Polygonaceae Fagopyrum esculentum Moench NC_010776  Ericales Actinidiaceae Actinidia chinensis Planch. NC_026690  Ericales Actinidiaceae Actinidia deliciosa (A.Chev.) NC_026691    C.F. Lian & A.R.Ferguson  Ericales Ericaceae Vaccinium macrocarpon Aiton NC_019616  Ericales Primulaceae Ardisia polysticta  Miq. NC_021121  Ericales Primulaceae Lysimachia coreana Nakai NC_026197  Ericales Primulaceae Primula poissonii Franch. NC_024543  Ericales Theaceae Camellia crapnelliana Tutcher NC_024541  Fabales Fabaceae Acacia ligulata Benth. NC_026134  Fabales Fabaceae Apios americana Medik. NC_025909  Fabales Fabaceae Arachis hypogaea L. NC_026676  Fabales Fabaceae Ceratonia siliqua L. NC_026678    80 Order Family Species Accession #  ______________________________________________________________________________  Fabales Fabaceae Haematoxylum brasiletto NC_026679    H.Karst. Fabales Fabaceae Libidibia coriaria (Jacq.) NC_026677    Schltdl. Fabales Fabaceae Lupinus albus L. NC_026681  Fabales Fabaceae Prosopis glandulosa Torr. NC_026683  Fabales Fabaceae Tamarindus indica L. NC_026685  Gentianales Apocynaceae Asclepias syriaca L. NC_022432  Gentianales Apocynaceae Catharanthus roseus (L.) G.Don NC_021423    cultivar Pacifica Punch Halo Gentianales Apocynaceae Echites umbellatus Jacq. NC_025655  Gentianales Apocynaceae Oncinotis tenuiloba Stapf NC_025657  Gentianales Apocynaceae Pentalinon luteum NC_025658    (L.) B.F.Hansen & Wunderlin Gentianales Apocynaceae Rhazya stricta Decne. NC_024292   Geraniales Melianthaceae Melianthus villosus Bolus NC_023256  Geraniales Vivianiaceae Viviania marifolia Cav. NC_023259  Lamiales Acanthaceae Andrographis paniculata NC_022451    (Burm.f.) Nees    Lamiales Lamiaceae Ajuga reptans L. NC_023102  Lamiales Lamiaceae Premna microphylla Turcz. NC_026291  Lamiales Lentibulariaceae Genlisea margaretae Hutch. NC_025652  Lamiales Lentibulariaceae Pinguicula ehlersiae Speta NC_023463    & F. Fuchs    81 Order Family Species Accession #  ______________________________________________________________________________  Lamiales Lentibulariaceae Utricularia gibba L. NC_021449  Lamiales Lentibulariaceae Utricularia macrorhiza Leconte NC_025653  Lamiales Orobanchaceae Cistanche deserticola Y.C.Ma NC_021111  Lamiales Orobanchaceae Cistanche phelypaea (L.) Cout. NC_025642  Lamiales Orobanchaceae Epifagus virginiana (L.) NC_001568   W.P.C. Barton  Lamiales Orobanchaceae Lindenbergia philippensis NC_022859    (Cham. & Schltdl.) Benth. Lamiales Orobanchaceae Orobanche californica Cham NC_025651   & Schltdl.  Lamiales Orobanchaceae Orobanche crenata Forssk. NC_024845  Lamiales Orobanchaceae Orobanche gracilis Sm. NC_023464  Lamiales Orobanchaceae Phelipanche purpurea (Jacq.) NC_023132    Sojak  Lamiales Orobanchaceae Phelipanche ramosa (L.) Pomel  NC_023465  Lamiales Orobanchaceae Schwalbea americana L. NC_023115  Lamiales Scrophulariaceae Boulardia latisquama NC_025641    F.W.Schultz  Lamiales Scrophulariaceae Scrophularia takesimensis NC_026202   Nakai  Malphigiales Chrysobalanaceae Parinari campestris Aubl. NC_024067  Malphigiales Salicaceae Salix interior Rowlee NC_024681  Malvales Malvaceae Gossypium anomalum Wawra NC_023213   & Peyr.    82 Order Family Species Accession #  ______________________________________________________________________________  Myrtales Myrtaceae Eucalyptus aromaphloia Pryor NC_022396   & J.H.Willis  Pandanales Cyclanthaceae Carludovica palmata Ruiz NC_026786    & Pav. Proteales Proteaceae Macadamia integrifolia Maiden NC_025288   & Betche  Rosales Moraceae Morus mongolica (Bureau) NC_025772   C.K.Schneid.    Rosales Rosaceae Fragaria chiloensis Auct. NC_019601  Rosales Rosaceae Fragaria virginiana Mill. NC_019602  Sapindales Meliaceae Azadirachta indica A.Juss. NC_023792  Sapindales Sapindaceae Sapindus mukorossi Gaertn. NC_025554  Saxifragales Altingiaceae Liquidambar formosana Hance NC_023092  Saxifragales Crassulaceae Sedum sarmentosum Bunge NC_023085  Saxifragales Paeoniaceae Paeonia obovata Maxim. NC_026076  Saxifragales Penthoraceae Penthorum chinense Pursh NC_023086  Trochodendrales Trochodendraceae Tetracentron sinense Oliv. NC_021425          83 Supplementary Table S2. List of primer sequences used to close gaps and verify overlapping contigs.   Taxon Primer name Primer sequence (5’ to 3’) Primer pair          Ericaceae Orthilia secunda Orsec_291F GTAAAGGGGGTCTGGGAAAA Orsec_291R  Orsec_291R TTCCTATTTCTTCGCGTTCG Orsec_291F  Orsec_34L CCCCCTTCTATCCACACCTT Orsec_501L  Orsec_501L CTCTGGCCTCTCAGGAATTG Orsec_34L  Orsec_81R TGATGTGGAAATTGGCTCTG Orsec_L1R  Orsec_A1F TCTACCCTTTCCCGTAAGTTGA Orsec_A1R  Orsec_A1R GGAAGGGGTTAAGTGCAACA Orsec_A1F  Orsec_A2R CCCGGTTCAATTGTAATGATG Orsec_O2F  Orsec_B1F TTCCGAGATGGAACTCTTGC Orsec_B1R  Orsec_B1R CAACGAAAGTGACCACGAGA Orsec_B1F  Orsec_D1F CGGCATGCCATCTTCTAAA Orsec_D1R, Orsec_J1R  Orsec_D1R CCAATAATCCAATTGTTCAATCA Orsec_D1F  Orsec_F1F CCAAGGGCTCAAGAATAAACC Orsec_F1R  Orsec_F1R TCCATGATACAGCAGAGCAGA Orsec_F1F   84 Taxon Primer name Primer sequence (5’ to 3’) Primer pair            Orsec_I1F AAATTGCTTTGGGTCGTTTG Orsec_I1R  Orsec_I1R AATCCCAATGAAAAGGCAGA Orsec_I1F  Orsec_J1R TGCAACATTGTTAACTCGAGGA Orsec_D1F  Orsec_L1R GCTGCTTGGCCTGTAGTAGG Orsec_81R  Orsec_M1R TGCTCAAACAATCCCAATCA Orsec_O1R  Orsec_O1R CCAATGGCGTTGGCTACTAT Orsec_M1R  Orsec_O2F TGGACAATGAGGAAGACTGC Orsec_A2R Gentianaceae Bartonia virginica Bavir_12F CCCCCAGGATCTATAATTTACTC Bavir_12R  Bavir_12R ATTGGTGAACCAGCAGATCC Bavir_12F  Bavir_19L CCTTGGGGTTATCCTGCACT Bavir_5R  Bavir_19L3 ATGTTGGGGTGAACCAGAAA Bavir_5R  Bavir_19L4 AAAAGGAGTAAGCTTGGGACA Bavir_50R2, Bavir_50R3  Bavir_200F CACGCAGAGGAACTAGGATTC Bavir_200R       85 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Bavir_200R CCTTGTTGTTCTAGTTGGATGTG Bavir_200F  Bavir_218R ACATCCGTCCCAAGGTATCA Bavir_C6R, Bavir_C8R  Bavir_219F ATCGAACCCGCATCTTCTC Bavir_219R  Bavir_219F2 GCATCGTTTCTCCTCCAAAA Bavir_219R2  Bavir_219R TCCCTTGAACCTGTGTATGAAG Bavir_219F  Bavir_219R2 AGGCGTAGGTGCTTTTCTTC Bavir_219F2  Bavir_37F ATTGCCTTGGACTTGTCGTT Bavir_37R  Bavir_37F2 TACCGGAACAAACGGCTATC Bavir_37R2  Bavir_37R CGCACACACTCTCTTTCCAA Bavir_37F  Bavir_37R2 AAGCTAACGATGCGGGTTC Bavir_37F2  Bavir_50R2 GCTATGCATGGTTCCTTGGT Bavir_19L4  Bavir_50R3 CTGCTGCTATAGAAGTTCCATCT Bavir_19L4  Bavir_5R TAGATGTCGGCCAAAAGCA Bavir_19L3, Bavir_19L  Bavir_A1F GGCCCGAGAATTGATGTGTA Bavir_A1R  Bavir_A1R TTCCCGCTGTTTTCTCATGT Bavir_A1F   86 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Bavir_B2F CGGTCCAGTAGGTCCGTAAA Bavir_B2R  Bavir_B2R CTACCACGTGGAAACGCTCT Bavir_B2F  Bavir_C1F TGGGTAACGGTATTCTGCCTA Bavir_C1R  Bavir_C1R CGTTGCGGTCGGACTCTAT Bavir_C1F  Bavir_C6R TGTTGGTAGCCCAGTTTTCC Bavir_218R  Bavir_C8R GTCCTCCCTACCCACCAATC Bavir_218R  Bavir_D1F CCTGGATACTCGGGTTCAAA Bavir_D2R  Bavir_D2R AACCCCAGGTTAAGCGAGAT Bavir_D1F  Bavir_E1F ACCTGAGAGCGGACAGCTAA Bavir_E1R  Bavir_E1R GTTGTATGCTGCGTTCGAGA Bavir_E1F  Bavir_F1F GTTTGATTCAGCGGGAGAAA Bavir_F1R  Bavir_F1R CTTTGCCAAGGAGAAGATGC Bavir_F1F Exacum affine Exaff_146F ACCTTTCCGAAGTCCTGGAG Exaff_146R  Exaff_146F2 AGATTACGCCCCTACTCTGC Exaff_146R2  Exaff_146R TCGCTATCAACTGCTTGTCC Exaff_146F   87 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Exaff_146R2 TCCACAGACGACGAAACTCT Exaff_146F2  Exaff_14F2 TCCACGTGGTAGAACCTCCT Exaff_14R2  Exaff_14F3 GTAGGCCCCCATCGTCTAGT Exaff_14R3  Exaff_14F4 ACTATAGGCGGAGCAATTCG Exaff_14R4  Exaff_14L TAGACGCCCCAGCAACTAAG Exaff_14R  Exaff_14R CACCACCAACTGTAGCAGCA Exaff_14L  Exaff_14R2 GGTCCTGAAGCACAAGGAGA Exaff_14F2  Exaff_14R3 TGCTAGGGGTGGGATATTTG Exaff_14F3  Exaff_14R4 CACCACCAACTGTAGCAGCA Exaff_14F4  Exaff_20R CAATATTCACCGGCCCAAGG Exaff_63L3  Exaff_235F AATGTATCGCCCCATCTCAA Exaff_235R  Exaff_235R GCTGGATCAACCCTTGAAAC Exaff_235F  Exaff_236F AATCGGAATCGTGGGTAGTA Exaff_236R  Exaff_236R TCAAGCTCTGGCAGATGGTA Exaff_236F  Exaff_289R TGAGTTCAACCAAGCCAACC Exaff_571R   88 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Exaff_30F CAACGAATCCGAATGTTTGA Exaff_30R  Exaff_30R GCCGATGATTTGGACGATAC Exaff_30F  Exaff_31F TGCCATGGTTCCTTACTTCG Exaff_31R  Exaff_31L CTTCTTGCTTTATCAAGGGAACAT Exaff_G1F  Exaff_31R GTGGAGAACGGAACCAAGAA Exaff_31F  Exaff_38F AGAGGGACGATTTCGTGAGA Exaff_38R  Exaff_38F2 GAAGCTCGGTAAAAGCAACG Exaff_38R2  Exaff_38R TTGAACTAGCCATCCCTTCG Exaff_38F  Exaff_38R2 CCCTGGATAAGCTTCACGAC Exaff_38F2  Exaff_47L2 CTCCTCGAAGCGATAAACGA Exaff_84R2  Exaff_49R AGCTCCACGCTTTCTTTCCT Exaff_65L2  Exaff_4F AATTCGAGTGGCTGAAGCTG Exaff_4R  Exaff_4R CAGGGTCAAGAACGACGAAT Exaff_4F  Exaff_571R CCTCCCCGTTCAGTGAATTA Exaff_289R  Exaff_63L3 TTAAAAGTTGCTCCTGCTACTCA Exaff_20R   89 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Exaff_65L2 AAGACATCACGATCCCTTGC Exaff_49R  Exaff_70F GCTCTTATGCCTGCAGAAACA Exaff_Q1f  Exaff_76F GGACGTTACCAAGGCTGAGA Exaff_76R  Exaff_76F2 TGCAGTCACTTCTTGTTTCCTG Exaff_76R2  Exaff_76R GTTGGTAACCGACCCAAAGA Exaff_76F  Exaff_76R2 GACACATAAGAGCCCGAACC Exaff_76F2  Exaff_84R2 AGACGACTGAGCCAACTTGAG Exaff_47L2  Exaff_91F3 GAGCGCGAAAAATTGAGC Exaff_91R3  Exaff_91R TGGTTGGTCATATAATCGTGCT Exaff_E1R  Exaff_91R3 GACTCGTGTTCTGGCTCGTC Exaff_91F3  Exaff_C1F TCTCACATTCGGCTAGAGCA Exaff_C1R  Exaff_C1R CCAAGGCTTTACCCCAAGAT Exaff_C1F  Exaff_D1F GGTCGAATTTTCCATCTCCA Exaff_D1R  Exaff_D1R ATCGGAGGAGTAGCTGCTGA Exaff_D1F  Exaff_E1F GTTTTGTCTAGTGCCAACAAAGG Exaff_E1R   90 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Exaff_E1R AGAAGGGGTGGAAAGTGAGG Exaff_91R, Exaff_E1F  Exaff_F1F CCCTGGAGAGATGGTTCACT Exaff_F1R  Exaff_F1R ACGACAGAAAGGGGGATTG Exaff_F1F  Exaff_G1F TAGTGGGGGAGTATGGGACA Exaff_31L, Exaff_G1R  Exaff_G1R TCCGTGTCGCTAAATATCCA Exaff_G1F  Exaff_H1F TCCGGCGTAGTTTTATACGG Exaff_H1R  Exaff_H1R ATCCACAAGTACCGGCAGAG Exaff_H1F  Exaff_K1F AAAATCGTGGTTGGGAAGG Exaff_K1R  Exaff_K1R GAGTTGACCGCCAGACCTAC Exaff_K1F  Exaff_L1F TCCTCCCGGAATAAAAGGAT Exaff_L1R  Exaff_L1R GGTTTGCCTTGGTATCGTGT Exaff_L1F  Exaff_M1F CCAAAGATCTCGGTCAGAGC Exaff_M1R  Exaff_M1R CAAGTATGGTCGTCCCCTGT Exaff_M1F  Exaff_N1F GGGTGAACGTACTCGTGAGG Exaff_N1R  Exaff_N1R GCGCTCGTGCTACAGTTAAA Exaff_N1F   91 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Exaff_O1F TCAGAAAAGGGGTGGCTCTA Exaff_O1R  Exaff_O1R TCCATCTCTCCTACCCGTTG Exaff_O1F  Exaff_Q1f CCGATTAGCCGTTGTCATTT Exaff_70F  Exaff_R1F CCACTCCAGTCGTTGCTTTT Exaff_R1R  Exaff_R1R TGGGCGGAACAGGTCTACTA Exaff_R1F  Exaff_S1F GCGTTCTTCGTCTCATCGTT Exaff_S1R  Exaff_S1R GGGGCTTCGACTCTCACATA Exaff_S1F  Exaff_T1F TTGGGGCCTCCTAAAAAGAT Exaff_T1R  Exaff_T1R GCTTAAAGTGCGGGAATATGA Exaff_T1F  Exaff_U1F GCTGGATTATTCGTCACTGC Exaff_U1R  Exaff_U1R GTCGCTTGCCTAACAATCAA Exaff_U1F  Exaff_V1F CTGAGGTACTCGGGTTCCAA Exaff_V1R  Exaff_V1R TCACCCCTTTCACTTCCTTG Exaff_V1F  Exaff_W1F TCCGCCTATAGTTCCTCGAA Exaff_W1R  Exaff_W1R CAGATTGGGGAGGAAGATCA Exaff_W1F   92 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Exaff_ycf2F ACAGACAGAGTTCGAAGGGG Exaff_ycf2R  Exaff_ycf2R TCCAGCTCCGTATCAAGGTC Exaff_ycf2F Exochaenium oliganthum Exoli_1689F CCCCTTTATTTCACCGGTTT Ex_ol_1689R  Exoli_1689R GTGTGGACCGACGGACTTAC Ex_ol_1689F  Exoli_104F CCCACAGCTTTGCTTTCAAT Exol_104R  Exoli_104R CAAAACTTCTACCCCGAGCA Exol_104F  Exoli_90F TGGGGTGATCTCGTAGTTCC Exoli_90R  Exoli_90R GCCAGGGTAAGGAAGAAAGG Exoli_90F  Exoli_A1F CAAGGTGGTCCTTGCTGATT Exoli_A1R  Exoli_A1R CGAGTCCGCTTATCTCCAAC Exoli_A1F  Exoli_B1F TCGAGCCGTGAAAAAGATTC Exoli_B1R  Exoli_B1R GCCACTACTGGTGAGCCCTA Exoli_B1F  Exoli_C1F GCTGGGGTTGCAAAATAAAA Exoli_C1R  Exoli_C1R CGGACAAAGCAAGAAGGGTA Exoli_C1F, Exoli_R1R  Exoli_D1F CACAATCTGGTTCTTGTTTCCA Exoli_D1R   93 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Exoli_D1R GGCAGAATACCGTCATCCAT Exoli_D1F  Exoli_E1F CAACTGCGAAATAGGCACAA Exoli_E1R  Exoli_E1R GAGGGGGAGTCGATTATTCC Exoli_E1F  Exoli_F1F AGAGCACGTAGGGCTTTGAA Exoli_F1R  Exoli_F1R GAAAAACTGGGTTGCGCTAT Exoli_F1F  Exoli_G1F TAGCACCATGCCAAATGTGT Exoli_G1R  Exoli_G1R TTTGCAGCTTTTGTTGTTGC Exoli_G1F  Exoli_H1F TCCCTTGCCTAACAATCAAA Exoli_H1R  Exoli_H1R GGGATCAGTTGGACCTTTGA Exoli_H1F  Exoli_I1F GCTTCCTCGTTTCACTTTGC Exoli_I1R  Exoli_I1R CCACGCGAAGGGTTTAGTTA Exoli_I1F  Exoli_J1F TAGGGCGTATCGTCCAAATC Exoli_J1R  Exoli_J1R CGGCGATAAGGTGCTAAAAG Exoli_J1F  Exoli_K1F AGCCTTTGCACAATTTGCTT Exoli_K1R  Exoli_K1R GAATGAAAGGCGTCCATTGT Exoli_K1F   94 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Exoli_L1F GCATGGGAACAGGTTCATCT Exoli_L1R  Exoli_L1R CTCTACCCAGGATCCCAACA Exoli_L1F  Exoli_M1F GATCCAACTCACATTCGGCC Exoli_M1R  Exoli_M1R AAATCCGCGGTTCCTAATGG Exoli_M1F  Exoli_N1F AAAAGCACTTGCCATTCGTT Exoli_N1R  Exoli_N1R TTCTTCTCTCCATCGGACCA Exoli_N1F  Exoli_O1F CTTCCTCAGCCAGGCAATAG Exoli_O1R  Exoli_O1R AGTTTGCGAAAGATGCAGGT Exoli_O1F  Exoli_P1F TGGACAAAGGTAAACATCTTGG Exoli_Q1F  Exoli_Q1F AATTTTTCGCAAACCCCTCT Exoli_P1F  Exoli_R1R CGACTCCTCGTGATCGACTT Exoli_C1R  Exoli_S1F CCAAAGATCTCCGTCAGAGC Exoli_S1R  Exoli_S1R TTTGGATTCAAAGCCCTACG Exoli_S1F  Exoli_T1F TGTACAAGGGCGTGCTGTAG Exoli_T1R  Exoli_T1R ACAACGTCGATGAAGACGTG Exoli_T1F   95 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Exoli_T2F GGTTACACCTCCAACCGAAA Exoli_T2R  Exoli_T2R GAAGACGTGTGGGTGCACTA Exoli_T2F  Exoli_T3F TTGGTTTACGCACGAATGAA Exoli_T3R  Exoli_T3R CAATACCCACGCCAAGAAAT Exoli_T3F  Exoli_T4F TTTCATCCACAAACGCAGAG Exoli_T4R  Exoli_T4R CATGCCCAGACGGATAAACT Exoli_T4F Obolaria virginica Obvir_1068L TGCATTCACACCATTCCAAC Obvir_E1F  Obvir_108R GAACATAGAAAGGCGGGATG Obvir_221L  Obvir_10L GGATTGCCTCACGAAATAGC Obvir_90R  Obvir_221L CGTCGGATGCTGGATATCTT Obvir_108R  Obvir_403R CTGGGTAGCTGACCCTTTGA Obvir_B1F  Obvir_90L2 AAAGAAGGATATGCTTGAAATGA Obvir_A2F, Obvir_A4F  Obvir_90R GGATTGCAAGGGTCAGTCAT Obvir_10L  Obvir_A2F CCAAAAACTGCTCAGCAACA Obvir_90L2, Obvir_A2R  Obvir_A2R CCCTCGCCCTAGGTTTTAAT Obvir_A2F, Obvir_A4F   96 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Obvir_A4F GCATCTACCATTATCCCCACA Obvir_90L2, Obvir_A2R  Obvir_B1F GCAAAGCCCTATGGGTTGTA Obvir_403R  Obvir_C1F TCAAGTCCACCACGAAGACA Obvir_C1R  Obvir_C1R GGTTGGGGATTTTGTGAAAG Obvir_C1F  Obvir_C2F GAGGAGGGCCTTGAAAAGTT Obvir_C2R  Obvir_C2R AGCAAGTCAAGTCGCACGTT Obvir_C2F  Obvir_D1F CTTGGCTTGGACAGGTCATT Obvir_D1R  Obvir_D1R GATAGCTCCATGGGCAAAAG Obvir_D1F  Obvir_E1F CCTGAAACCTTGGCACAGAT Obvir_E1R, 1069L  Obvir_E1R TGTCGAATGAGTTTGGAAAGA Obvir_E1F  Obvir_ccsAF CGATGTCAGGGCTTTTAACG Obvir_ccsAR  Obvir_ccsAR TACGATTCGTGTCGGTTCAC Obvir_ccsAF  Obvir_ycf2F CTCCAGGGATGAATCGAAAA Obvir_ycf2R  Obvir_ycf2R AGGGTGCTATTGTTCCTCCA Obvir_ycf2F Voyria clavata Vocla_21F2 CCCAATGCTGTCCTAGTTGA Vocla_21R2   97 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Vocla_21R GCAGCATCCAAAATGCCTAT Vocla_B1R  Vocla_21R2 TGTGAATTGCGCGAAAGTAG Vocla_21F2  Vocla_6L GGCTCTACTCCGGGTAAAAA Vocla_C1F  Vocla_B1F TCGATGAACGTTTGATTTTCC Vocla_B1R  Vocla_B1R TCGAAGTAACCTCCTTTGATCC Vocla_B1F, Vocla_21R  Vocla_C1F TGTAGACCCCCGAACAAAAG Vocla_6L, Vocla_C1R  Vocla_C1R AAAAGTGGCTCGGTGGTATG Vocla_C1F Polygalaceae Epirixanthes elongata Epelo_15L GTTCGAGTACCAGGCGCTAC Epelo_416R  Epelo_15R GTAGCGCCTGGTACTCGAAC Epelo_441L  Epelo_20L AGGCCTACGGGTCGTAAACT Epelo_39R  Epelo_21L TCTAGCCCCTCTGGGATGTA Epelo_847R  Epelo_21R GGGGAACTCGAATTTTTGGT Epelo_416L  Epelo_2828L TCTAAGGGTAGCCTGCTCCA Epelo_416L  Epelo_348L TGCACGGCTACACAGAAATC Epelo_416R   98 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Epelo_348R AGGGGCTCAGGACATCTCTC Epelo_39L, Epelo_847L  Epelo_39L CCGTCACACTAGGGAAGCTG Epelo_348R, Epelo 441R  Epelo_39R CATGTCAAGCCCTGGTAAGG Epelo_20L, Epelo_847L  Epelo_416L GTGGGCGTTAGAGCATTGAT Epelo_2828L, Epelo_21R  Epelo_416R CCCCCATACATGGTCTTACG Epelo_15L  Epelo_441L GGGTGATCTATCCAGGACCA Epelo_15R  Epelo_441R GCTACTGGACTCTCGCCATC Epelo_39L  Epelo_847L TCGACGAAGACGTGTAGGTG Epelo_39R, Epelo_348R  Epelo_847R GATCTCGCGGATCTTTCGAT Epelo_21L Epirixanthes pallida Eppal_2427F ATCTCCCGGATAAGCCTCAC Eppal_2427R  Eppal_2427R TGCCCTGGCTAAACCTATTG Eppal_2427F  Eppal_701F TCTTGATTGGAAGGGACACC Eppal_701R  Eppal_701R GGGCGTTAGAGCATTGAGAG Eppal_701F  Eppal_A1F CATCGGTCCACACAGTTGTC Eppal_A1R  Eppal_A1R AGCGATGGAGTTAGCAATCG Eppal_A1F   99 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Eppal_B1F TGCGTTTTGGGAGCTTCTAT Eppal_B2R  Eppal_B2R GCGCCTAACCCTATGAGTTG Eppal_B1F  Eppal_C1F GAATCCCATGAAGGACGAAA Eppal_C1R  Eppal_C1R ACGGGAATCCCCTTTATTTG Eppal_C1F  Eppal_D1F AGCATGGACCCACTCCTATG Eppal_D2R  Eppal_D2R CACATGGAGCCATCTCCTTA Eppal_D1F  Eppal_E1F TCATTCATGGGCGTTGATAA Eppal_E1R  Eppal_E1R CAGAGCGCAAGCTAGTGATG Eppal_E1F  Eppal_F1F CCGCCATCCTACCTAATGAA Eppal_F1R  Eppal_F1R CTCATCGCCTCGCTTTATCT Eppal_F1F  Eppal_G1F TTCATCGAATACGGCTTTCC Eppal_K1F, Eppal_G1R  Eppal_G1R AGGGGGAAGGGTTAAGGATT Eppal_G1F  Eppal_H2F ACGAAATCGCATTGATAGCC Eppal_I1F  Eppal_I1F TCAACCCACCCTTAGTACCG Eppal_H2F  Eppal_I1R AACTACGAGATCGCCCCTTT Eppal_J3R   100 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Eppal_J3R CGTAGTTCCTACGGGGTGAA Eppal_I1R  Eppal_K1F GGCATGGCATCTTATGAAGG Eppal_G1F  Eppal_L1F TGGAACTCCAACAGGCATAA Eppal_L2R  Eppal_L2R GGATTCAACAAAGACGGTTCA Eppal_L1F Polygala arillata Poari_2F GAATGAGGAGCCGTATGAGG Poari_2R  Poari_2R TCCCTACGAAATACCAGACGA Poari_2F  Poari_A1F TGATTGGTCGTATAATCGTGGT Poari_A1R  Poari_A1R TGGGACGTTTACCAGTGTCA Poari_A1F, Poari_C1F  Poari_B1R GCGCTAACCTTGGTATGGAA Poari_B4F  Poari_B4F GGAAATCGGCCACATTAAAA Poari_B1R  Poari_C1F TGCTGCAGCTACAAAGTGTG Poari_A1R Salomonia cantoniensis Sacan_326R CAACCGGTCGAGTAAGATGAG Sacan_347L  Sacan_347L TGCTTCTGGCCTGGATAAAC Sacan_326R  Sacan_B1F CACGGAATGTATTTGCACCA Sacan_B1R  Sacan_B1R TTGGTTCACGGGTACAACCT Sacan_B1F   101 Taxon Primer name Primer sequence (5’ to 3’) Primer pair         Sacan_C1F AGCTGTGCTGCTGCTACAAA Sacan_C1R  Sacan_C1R TGGGACGTTTACCAGTGTCA Sacan_C1F, Sacan_M1F  Sacan_D1R GGATTGAGCCGAATACAACC Sacan_Na1F  Sacan_E1F TTAGCGAATTCGTGTGCTTG Sacan_E1R   Sacan_E1R ATCGGCCAAAATAACCATGA Sacan_E1F  Sacan_F1F GCGCTAACCTTGGTATGGAA Sacan_P1R  Sacan_F1R TGGCTAGGTAAGCGTCCTGT Sacan_F1F  Sacan_G1F TCCCCATGAGTTCCAGTCTC Sacan_G1R  Sacan_G1F TCCCCATGAGTTCCAGTCTC Sacan_G1R  Sacan_G1R ATCCAGGATTTGAACGGATG Sacan_G1F  Sacan_G1R ATCCAGGATTTGAACGGATG Sacan_G1F  Sacan_H1F TCGGTTTCCATTTTGGTTGT Sacan_H1R  Sacan_H1R CTACTCAGCCCAGAGCCTTG Sacan_H1F  Sacan_I1F AAGGGGTTTCAAAAACCAAGA Sacan_I1R  Sacan_I1R CTTCGTTTGCAGCAACACTC Sacan_I1F   102 Taxon Primer name Primer sequence (5’ to 3’) Primer pair        Sacan_J1F CATGCACGGTTTTGAATGAG Sacan_J1R  Sacan_J1R TTCTTGGTTTCGTCCAGTCA Sacan_J1F  Sacan_M1F TGCTTGGTCGTATCATCGTG Sacan_C1R  Sacan_Na1F TGAACAGATCCGGTGAAAAA Sacan_D1R  Sacan_Nb1F TTTCAACTTGCTCTGCTCCT Sacan_R1F  Sacan_O1F AGGGTGTCCGTGACGTGT Sacan_O1R  Sacan_O1R AGGGGTTGTGGATACTGCTG Sacan_O1F  Sacan_P1R TGCTCTATTTCGTTCCTTGG Sacan_F1F  Sacan_R1F CCCGTTCTCTACGTTTTTGC Sacan_Nb1F  Sacan_T1F AGGCCATTTAGTCCATGTCG Sacan_T1R  Sacan_T1R CAGAAAGAGGCTGACCCAAC Sacan_T1F     103 Supplementary Table S3. Results of partition-finder analyses, summarizing final partitioning schemes and the optimal DNA or amino-acid substitution models associated with each data partition: (a) ‘ORF-only’ (open reading frame only) matrix partitioned using the ‘GxC’ (gene by codon) partitioning scheme; (b) A version of the matrix with pseudogenes included, partitioned using the GxC scheme; (c) Amino-acid matrix, partitioned by gene. Genes are indicated before the underscore; the ‘pos’ term after the underscore indicates the codon position for protein-coding genes.   Partition Best Model Partition subset             a) ORF-only (GxC scheme)  1 GTR+I+! accD_pos1, clpP_pos1 2 GTR+I+!  accD_pos2, ccsA_pos1, ndhF_pos1 3 GTR+I+! accD_pos3, atpE_pos3, infA_pos3, rpl20_pos3,    rpoC2_pos3 4 GTR+I+! atpA_pos1, atpI_pos1, petA_pos1, rpoB_pos1, rps12_pos1 5 GTR+I+! atpA_pos2, atpB_pos2, psbB_pos2, rps12_pos  6 GTR+I+! atpA_pos3, atpI_pos3, ndhK_pos3, petA_pos3, petB_pos3,     psbB_pos3, psbI_pos3, rps16_pos3 7 GTR+I+!  atpB_pos1, psbB_pos1 8 GTR+I+! atpB_pos3, atpF_pos3, ndhJ_pos3, psbC_pos3, rpl33_pos3,    rpoC1_pos3 9 GTR+I+! atpE_pos1, atpF_pos1, rpl2_pos3, rps14_pos1, rps19_pos1,    rps2_pos1 10 GTR+I+!  atpE_pos2, cemA_pos2, ndhC_pos2, ndhE_pos2,    petL_pos1, petL_pos2, psaJ_pos1, psbL_pos3, psbT_pos1, 11 GTR+I+! atpF_pos2, psbF_pos3, rpoC2_pos2, rps11_pos2 12 GTR+I+! atpH_pos1, petD_pos1, psaA_pos1, psaB_pos1, psbN_pos1   psbZ_pos1, rpoC1_pos2, rps12_pos3, rps2_pos2,    ycf4_pos2 13  GTR+!  atpH_pos2 14 GTR+I+! atpH_pos3, psbZ_pos3, rpl36_pos3, rps14_pos3,    rps18_pos3, rps4_pos 15 GTR+I+! atpI_pos2, ndhG_pos2, ndhJ_pos2, rpoB_pos2, rps14_pos2  16 GTR+I+! ccsA_pos2  17 GTR+I+! ccsA_pos3, ndhD_pos3, ndhE_pos3, petD_pos3,    psaC_pos3, rps15_pos3 18 GTR+I+! cemA_pos1, rpoC2_pos1, rps18_pos2 19 GTR+! cemA_pos3, clpP_pos3 20 GTR+I+! clpP_pos2  21 GTR+I+!  infA_pos1, rpl16_pos1, rpl20_pos1, rpl33_pos1,    rps11_pos1, rps16_pos1, rps4_pos1   104 Partition Best Model Partition subset             22 GTR+I+! infA_pos2, ndhH_pos2, petA_pos2, rpl16_pos2,    rps19_pos2 23 GTR+I+! matK_pos1 24 GTR+! matK_pos2 25 GTR+I+! matK_pos3, rpl16_pos3, rpoA_pos3 26 GTR+I+! ndhA_pos1, ndhD_pos1, psaI_pos2, psaJ_pos2,    psbH_pos1, psbJ_pos1, psbM_pos1, psbT_pos2, rps8_pos2 27 GTR+I+! ndhA_pos2, petN_pos3, psaI_pos1, psbE_pos3, psbJ_pos3,    rps16_pos2 28 GTR+I+! ndhA_pos3, ndhG_pos3, psbT_pos3, rps3_pos3, rps8_pos3 29 GTR+!  ndhB_pos1, petN_pos1, petN_pos2, psbF_pos1, psbI_pos2,    psbL_pos1, psbM_pos2, rrn5 30 GTR+I+! ndhB_pos2, psbI_pos1, psbN_pos2 31 GTR+! ndhB_pos3, ndhE_pos1, ndhJ_pos1, rpl23_pos2,    rpl23_pos3, rps7_pos3 32 GTR+I+!  ndhC_pos1, ndhH_pos1, ndhI_pos1, ndhK_pos2,    rps4_pos2 33 GTR+I+! ndhC_pos3, ndhG_pos1, psbD_pos3, psbK_pos3,    psbN_pos3, rpl20_pos2, ycf3_pos3 34 GTR+I+! ndhD_pos2, petG_pos1 35 GTR+I+! ndhF_pos2, ndhI_pos2, psbK_pos1 36 GTR+I+! ndhF_pos3 37 GTR+I+! ndhH_pos3, ndhI_pos3, rbcL_pos3, rps11_pos3 38 GTR+I+! ndhK_pos1, rpl14_pos1, rpoC1_pos1, ycf4_pos1  39 GTR+I+! petB_pos1, psbD_pos1 40 GTR+I+! petB_pos2, psbD_pos2 41 GTR+I+! petD_pos2, psbA_pos1, psbA_pos2, psbE_pos1,    psbL_pos2 42 GTR+! petG_pos2 43 GTR+I+! petG_pos3, petL_pos3, psaI_pos3, rps15_pos2 44 GTR+I+! psaA_pos2, psbE_pos2 45 GTR+I+! psaA_pos3, psaB_pos3, psbH_pos3, psbM_pos3,    rpl14_pos3, rpoB_pos3, rps2_pos3, ycf4_pos3 46 GTR+I+! psaB_pos2, psaC_pos2, psbC_pos2, psbF_pos2, psbJ_pos2,     psbK_pos2, psbZ_pos2 47 GTR+I+! psaC_pos1, rpl36_pos1, rpl36_pos2, rps7_pos1, ycf3_pos1 48 GTR+I+! psaJ_pos3 49 GTR+I+! psbA_pos3 50 GTR+I+! psbC_pos1 51 GTR+I+! psbH_pos2 52 GTR+I+! rbcL_pos1 53 GTR+I+! rbcL_pos2   105 Partition Best Model Partition subset             54 GTR+I+! rpl14_pos2, rpl23_pos1, rpl2_pos1, rpl2_pos2, rps7_pos2,    ycf3_pos2 55 GTR+! rpl22_pos1 56 GTR+I+!  rpl22_pos2 57 GTR+I+! rpl22_pos3, rpl32_pos3 58 GTR+!  rpl32_pos1, rpl32_pos2 59 GTR+I+! rpl33_pos2, rpoA_pos1, rps15_pos1, rps18_pos1,    rps3_pos1 60 GTR+I+! rpoA_pos2, rps3_pos2 61 GTR+I+! rps8_pos1 62 GTR+I+! rps19_pos3 63 GTR+! rrn16, rrn4_5 64 GTR+I+! rrn23 65 GTR+! ycf2_pos1 66 GTR+! ycf2_pos2 67 GTR+!  ycf2_pos3  b) Pseudogenes included (GxC scheme)  1 GTR+I+! accD_pos1, clpP_pos1, rpl32_pos1 2 GTR+I+! accD_pos2 3 GTR+I+! accD_pos3, infA_pos3, rpoC2_pos3 4 GTR+I+! atpA_pos1 5 GTR+I+! atpA_pos2, rpoB_pos2, rpoC1_pos2, rps2_pos2 6 GTR+I+! atpA_pos3, atpI_pos3, ndhG_pos3, ndhK_pos3,    petA_pos3, petB_pos3, psbB_pos3, psbI_pos3 7 GTR+I+! atpB_pos1, atpI_pos1, petA_pos1, petA_pos2, rpl16_pos2,    rps12_pos1 8 GTR+I+! atpB_pos2, psbB_pos2, psbF_pos2 9 GTR+I+! atpB_pos3, ndhJ_pos3 10 GTR+! atpE_pos1, atpF_pos1, infA_pos2, rpl2_pos3, rps19_pos1 11 GTR+I+! atpE_pos2, rps19_pos2, rps7_pos2 12 GTR+I+! atpE_pos3, ndhA_pos3, psbT_pos3, rpl20_pos3,    rps16_pos3, rps3_pos3, rps8_pos3 13 GTR+I+! atpF_pos2, rpoC2_pos2, rps11_pos2, rps18_pos2 14 GTR+I+! atpF_pos3, rpl32_pos2, rpl33_pos3, rpoB_pos3,    rpoC1_pos3, rps18_pos3, ycf4_pos3 15 GTR+I+! atpH_pos1, ndhC_pos2, petB_pos1, petD_pos1,    psaB_pos1, psbA_pos1, psbA_pos2, psbN_pos1 16 GTR+! atpH_pos2 17 GTR+I+! atpH_pos3, psbH_pos3, psbN_pos3, psbZ_pos3,    rpl36_pos3   106 Partition Best Model Partition subset             18 GTR+I+! atpI_pos2, petB_pos2 19 GTR+I+! ccsA_pos1, ndhF_pos1 20 GTR+I+! ccsA_pos2, ndhI_pos2 21 GTR+I+! ccsA_pos3, ndhD_pos3, ndhE_pos3, petD_pos3,    psaC_pos3, rps15_pos3 22 GTR+I+! cemA_pos1, psbM_pos1, rpl20_pos2, rpl33_pos2,    rpoC2_pos1, rps15_pos1, rps18_pos1 23 GTR+! cemA_pos2, ndhE_pos2, petL_pos1, petL_pos2,    psaJ_pos1, psbL_pos3, psbZ_pos1, ycf4_pos2 24 GTR+! cemA_pos3, clpP_pos3 25 GTR+I+! clpP_pos2, petG_pos1 26 GTR+I+! infA_pos1, rpl22_pos2, rpl33_pos1, rpoA_pos1,    rpoA_pos2, rps3_pos2 27 GTR+I+! matK_pos1 28 GTR+I+! matK_pos2, ndhG_pos1, petG_pos3, petL_pos3,    psaI_pos1, rps15_pos2 29 GTR+I+! matK_pos3, ndhI_pos3, rpoA_pos3 30 GTR+I+! ndhA_pos1, ndhD_pos1, psaI_pos2, psaJ_pos2,    psbH_pos1, psbJ_pos1, psbL_pos2, psbT_pos2, rps8_pos2 31 GTR+I+! ndhA_pos2, ndhG_pos2, ndhJ_pos2, ndhK_pos2,    rps12_pos2, rps16_pos2, rps4_pos2 32 GTR+I+! ndhB_pos1, petN_pos1, petN_pos2, psaA_pos2,    psbC_pos2, psbE_pos2, psbF_pos1, psbI_pos2, psbL_pos1,    psbM_pos2, rpl14_pos2, rrn4_5, rrn5, ycf3_pos2 33 GTR+I+! ndhB_pos2, psbI_pos1, psbN_pos2, psbZ_pos2 34 GTR+! ndhB_pos3, rpl23_pos1, rpl23_pos2, rpl23_pos3,    rps7_pos3 35 GTR+I+! ndhC_pos1, ndhH_pos1, ndhI_pos1, ndhK_pos1,  rpl14_pos1, rps11_pos1, rps12_pos3, rps14_pos2, rps16_pos1, rps4_pos1, ycf4_pos1 36 GTR+I+! ndhC_pos3, psaB_pos3, psbC_pos3 37 GTR+I+! ndhD_pos2, petD_pos2  38 GTR+! ndhE_pos1, ycf2_pos2 39 GTR+I+! ndhF_pos2, psbK_pos1, psbM_pos3 40 GTR+I+! ndhF_pos3 41 GTR+! ndhH_pos2, psbK_pos2 42 GTR+I+! ndhH_pos3, rbcL_pos3, rpl16_pos3, rps11_pos3 43 GTR+! ndhJ_pos1, psaA_pos1, rpl36_pos2, rps7_pos1 44 GTR+I+! petG_pos2, rbcL_pos2 45 GTR+! petN_pos3, psbE_pos3, psbF_pos3, psbJ_pos3 46 GTR+I+! psaA_pos3, rpl14_pos3 47 GTR+I+! psaB_pos2, psaC_pos2, psbD_pos2   107 Partition Best Model Partition subset             48 GTR+I+! psaC_pos1, rpl2_pos1, rpl2_pos2, ycf3_pos1 49 GTR+I+! psaI_pos3, psbD_pos3, psbK_pos3, rps14_pos3,    rps2_pos3, rps4_pos3, ycf3_pos3 50 GTR+I+! psaJ_pos3 51 GTR+I+! psbA_pos3 52 GTR+I+! psbB_pos1, psbC_pos1, psbE_pos1, psbT_pos1 53 GTR+I+! psbD_pos1, psbJ_pos2, rrn16 54 GTR+I+! psbH_pos2 55 GTR+I+! rbcL_pos1 56 GTR+! rpl16_pos1, rpl20_pos1 57 GTR+! rpl22_pos1 58 GTR+I+! rpl22_pos3, rpl32_pos3 59 GTR+! rpl36_pos1, ycf2_pos1 60 GTR+I+! rpoB_pos1, rpoC1_pos1, rps14_pos1, rps2_pos1 61 GTR+I+! rps3_pos1, rps8_pos1 62 GTR+I+! rps19_pos3 63 GTR+I+! rrn23 64 GTR+! ycf2_pos3  c) Amino acid scheme (partitioned by gene)  1 JTT+!+F accD, rpl22, rpl32 2 JTT+I+! atpA, atpB 3 CPREV+! atpE 4 JTT+!+F atpF, ndhG 5 CPREV+! atpH 6 JTT+!+F atpI, ndhE, ndhI, ndhJ, petA, petL, rpoB 7 JTT+I+!+F ccsA 8 JTT+I+!+F cemA, psaI, rpoA, rps15, rps18 9 CPREV+! clpP 10 JTT+! infA, rps4 11 JTT+!+F matK 12 JTT+I+!+F ndhA 13 JTT+!+F ndhB, petD, psaA, psaB, psbN, psbZ  14 JTT+I+!+F ndhC, ndhK, psbK, rpl14, rpoC1 15 JTT+I+!+F ndhD, rpoC2, rps3, rps8 16 JTT+I+!+F ndhF 17 JTT+I+! ndhH 18 JTT+I+! petB, psbE 19 CPREV+! petG, petN, psb 20 JTT+I+! psaC, psbJ 21 MTMAM+! psaJ   108 Partition Best Model Partition subset             22 CPREV+I+! psbA 23 JTT+I+! psbB, psbL, psbT 24 CPREV+I+! psbC 25 CPREV+I+! psbD 26 CPREV+! psbF, psbM, rpl36, rps12 27 JTT+I+! psbH 28 LG+I+! rbcL 29 JTT+! rpl2, ycf3 30 CPREV+! rpl16 31 CPREV+I+! rpl20 32 JTT+! rpl23, rps7 33 JTT+! rpl33, rps19 34 JTT+!+F rps2, ycf2 35 JTT+! rps11, rps16 36 JTT+! rps14 37 JTT+!+F ycf4       109 Supplementary Table S4. Species with partially assembled plastid genomes.                  No. genes  No.  No.       Combined with intact rDNA tRNA Family Species No. raw reads  No. contigs length (bp) reading frame1 genes1 genes1 ___________________________________________________________________________________________________________ Ericaceae Orthilia secunda2 15,484,936  1 145,723 66  4 30 Ericaceae Pyrola minor 8,314,630  17 127,096 68  4 30  Gentianaceae Voyria caerulea 29,597,488  9 46,826 17  4 13 Polygalaceae Epirixanthes elongata2 19,295,216  23 16,5933 133  3 5 Polygalaceae Salomonia cantoniensis2 15,896,526  5 135,290 75  2 30  1 Genes found in the inverted repeat counted once  2 Species for which additional PCR and Sanger sequencing were used to join de novo contigs into larger fragments. 3 Calculated with all possible plastid sequences, including low-depth regions (see Fig. S8)   110 Figure S1. Circular plastome map of Polygala arillata (Polygalaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   !! """! !  112 Figure S2. Circular plastome map of Epirixanthes pallida (Polygalaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick blue lines indicate direct repeat copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   ! ""#! !  114 Figure S3. Circular plastome map of Exacum affine (Gentianaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). The truncated ycf1 pseudogene is marked as ‘!’.                   ! ""#! !  116 Figure S4. Circular plastome map of Exochaenium oliganthum (Gentianaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   ! ""#! !  118 Figure S5. Circular plastome map of Bartonia virginica (Gentianaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   ! ""#! !  120 Figure S6. Circular plastome map of Obolaria virginica (Gentianaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   ! "#"! !  122 Figure S7. Circular plastome map of Voyria clavata (Gentianaceae). Genes located inside the circle are transcribed clockwise, those outside are transcribed counterclockwise. The grey circle marks the GC content; the inner circle marks a 50% threshold. Thick black lines indicate inverted repeat (IR) copies. Genes with introns are indicated with asterisks (*). Pseudogenes are marked as ‘!’.                   ! "#$! !  124 Figure S8. Linearized plastome map of the draft partial assembly of Epirixanthes elongata (Polygalaceae). Black lines below the map indicate the Sanger-connected Illumina contigs in the assembly, and the relative read depth is indicated (1X = ~200X read depth; 8X = ~eight times coverage; 4X = ~four times coverage; 2X = ~two times coverage; see main text). Arrows indicate regions of assembly where gaps and contig overlaps were respectively connected or confirmed using Sanger sequencing (not to scale; thin dashed lines are sequenced regions not represented in de novo contigs). Pseudogenes are indicated in red. Genes with introns are indicated with an asterisk (*). A ~1300 bp contig is not shown here (it has an uninterrupted copy of the 3’-rps12 and rps7; note that I did not recover the 5’-rps12 in the gene set ), as its relative connection to the main assembly has not been confirmed. Scale is in kb.              ! "#$!!  126 Figure S9. Angiosperm phylogeny inferred in an unpartitioned likelihood analysis of 82 plastid genes (ORF-only; see text and Table 3). Log likelihood score of best tree: -1,527,008.266. Bootstrap support values are indicated beside branches. Thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. Eudicot families where mycoheterotrophy has evolved are indicated in blue. The scale bar indicates estimated substitutions per site.                    !! "#$!!  128 Figure S10. Angiosperm phylogeny inferred in a likelihood analysis of 78 translated plastid genes (ORF-only) using the gene partitioning scheme (see text and Table S3 for details). Log likelihood score of best tree: -708,394.300. Bootstrap support values are indicated beside branches. Thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. Eudicot families where mycoheterotrophy has evolved are indicated in blue. The scale bar indicates estimated substitutions per residue.                  ! "#$! !  130  Figure S11. Angiosperm phylogeny inferred in a parsimony analysis of 82 plastid coding regions (ORF-only; see text and Table S3). This is one of the two shortest trees: length = 287,405 steps. Branches that collapse in the strict consensus are indicated with arrows. Bootstrap support values are indicated besides branches. Thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. Eudicot families where mycoheterotrophy has evolved are indicated in blue. The scale bar indicates the inferred number of changes.                  ! "#"!!  132  Figure S12. Angiosperm phylogeny inferred in a likelihood analysis of 82 plastid genes that includes putative pseudogenes (see text and Table S3). Log likelihood score of best tree: -1,565,108.435. Bootstrap support values are indicated beside branches. Thick lines indicate 100% bootstrap support; ‘--’ indicates <50% bootstrap support. Eudicot families where mycoheterotrophy has evolved are indicated in blue. The scale bar indicates estimated substitutions per site.  ! "##! !


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items