UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Phylogenetics and molecular evolution of Alismatales based on whole plastid genomes Ross, Thomas Gregory 2014

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

Item Metadata


24-ubc_2015_february_ross_thomas.pdf [ 4.44MB ]
JSON: 24-1.0165562.json
JSON-LD: 24-1.0165562-ld.json
RDF/XML (Pretty): 24-1.0165562-rdf.xml
RDF/JSON: 24-1.0165562-rdf.json
Turtle: 24-1.0165562-turtle.txt
N-Triples: 24-1.0165562-rdf-ntriples.txt
Original Record: 24-1.0165562-source.json
Full Text

Full Text

  PHYLOGENETICS AND MOLECULAR EVOLUTION OF ALISMATALES BASED ON WHOLE PLASTID GENOMES   by  Thomas Gregory Ross   B.Sc. The University of British Columbia, 2011   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIRMENTS FOR THE DEGREE OF    MASTER OF SCIENCE   in   The Faculty of Graduate and Postdoctoral Studies   (Botany)   THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  November 2014   © Thomas Gregory Ross, 2014  ii  ABSTRACT  The order Alismatales is a mostly aquatic group of monocots that displays substantial morphological and life history diversity, including the seagrasses, the only land plants that have re-colonized marine environments.  Past phylogenetic studies of the order have either considered a single gene with dense taxonomic sampling, or several genes with thinner sampling. Despite substantial progress based on these studies, multiple phylogenetic uncertainties still remain concerning higher-order phylogenetic relationships. To address these issues, I completed a near-genus level sampling of the core alismatid families and the phylogenetically isolated family Tofieldiaceae, adding these new data to published sequences of Araceae and other monocots, eudicots and ANITA-grade angiosperms. I recovered whole plastid genomes (plastid gene sets representing up to 83 genes per taxa) and analyzed them using maximum likelihood and parsimony approaches. I recovered a well supported phylogenetic backbone for most of the order, with all families supported as monophyletic, and with strong support for most inter- and intrafamilial relationships. A major exception is the relative arrangement of Araceae, core alismatids and Tofieldiaceae; although most analyses recovered Tofieldiaceae as the sister-group of the rest of the order, this result was not well supported. Different partitioning schemes used in the likelihood analyses had little effect on patterns of clade support across the order, and the parsimony and likelihood results were generally highly congruent. I also used the inferred phylogeny of Alismatales to study the loss of the mostly plastid-encoded NADH dehydrogenase enzyme complex in the order. This enzyme is hypothesized to be involved in mitigating photooxidative stress by inducing chlororespiration. The inclusion or exclusion of ndh  iii  pseudogenes had little impact on the main phylogenetic results. Previous work hypothesized three independent losses/pseudogenization events within the core alismatids, which I confirmed here. I also inferred an additional loss in Tofieldiaceae, the first example in unsubmerged species of Alismatales. The repeated loss of plastid NADH dehydrogenase may spur future research into the physiological bases of the loss.    iv  PREFACE All steps of this work, including sequencing and analysis was conducted predominantly by me, but with the assistance of others at the following specific locations. Jerrold Davis (Cornell University) and Craig Barrett (California State Los Angeles) were responsible for sequencing 11 of the samples I used in my analysis: Sagittaria latifolia (Alismataceae), Amphibolis griffithii (Cymodoceaceae), Ruppia polycarpa (Cymodoceaceae), Najas guadalupensis (Hydrocharitaceae), Vallisneria americana (Hydrocharitaceae), Triglochin procera (Juncaginaceae), Posidonia ostenfieldii (Posidoniaceae), Potamogeton pectinatus (Potamogetonaceae), Zanichellia palustris (Potamogetonaceae), Scheuchzeria palustris (Scheuchzeriaceae), Zostera mulleri (Zosteraceae). Vivienne Lam (UBC) was responsible for sequencing, assembly and gene annotation of  Lophiola aurea (Nartheciaceae) and Campynema lineare (Campynemataceae), and Marybel Soto Gomez (UBC) was responsible for sequencing, assembly and gene annotation of Burmannia bicolor (Burmanniaceae), Cyclanthus bipartitus (Cyclanthaceae), Freycinetia banksii (Pandanaceae), Stichoneuron caudatum (Stemonaceae) and Xerophyta retinervis (Velloziaceae). Additional thanks are due to Donald Les (University of Connecticut), Dennis Stevenson (New York Botanical Garden), John Conran (University of Adelaide), and Gitte Petersen (University of Copenhagen) who provided multiple plant samples as DNA or silica dried specimens. Bioinformatics scripting assistance was provided by David Tack (UBC) and Daisie Huang (UBC).     v  TABLE OF CONTENTS  ABSTRACT .................................................................................................................................... ii PREFACE ...................................................................................................................................... iv TABLE OF CONTENTS ................................................................................................................ v LIST OF TABLES ........................................................................................................................ vii LIST OF FIGURES ..................................................................................................................... viii ACKNOWLEDGMENTS ............................................................................................................. ix 1. INTRODUCTION ...................................................................................................................... 1 2. MATERIALS AND METHODS ................................................................................................ 4 2.1 Taxonomic sampling ............................................................................................................. 4 2.2 DNA and library preparation................................................................................................. 4 2.3 Contig assembly and plastid gene annotation ....................................................................... 6 2.4 Alignment and matrix construction ....................................................................................... 7 2.5 Error checking ....................................................................................................................... 8 2.6 Phylogenetic inference .......................................................................................................... 8 3. RESULTS ................................................................................................................................. 12 3.2 Relationships in core Alismatales ....................................................................................... 12 3.3 Loss of ndh genes ................................................................................................................ 13 4. DISCUSSION ........................................................................................................................... 14 4.1 The phylogenetic positions of Alismatales and Acorus in monocot phylogeny ................. 14 4.2 Resolving the first split in Alismatales phylogeny.............................................................. 15 4.3 Relationships in core alismatids: Higher-order relationships ............................................. 16 4.4 Relationships in core alismatids: The petaloid clade .......................................................... 17 4.5 Relationships in core alismatids: The tepaloid clade .......................................................... 19 4.6 Repeated loss of the plastid NADH dehydrogenase complex in Alismatales ..................... 22 4.7 Conclusion ........................................................................................................................... 25 FIGURES AND TABLES ............................................................................................................ 27 BIBLIOGRAPHY ......................................................................................................................... 38  vi  APPENDICES .............................................................................................................................. 47      vii  LIST OF TABLES  Table 1 Summary of bootstrap support for clades found in shortest trees for core alismatids that have less than 100% bootstrap in at least one parsimony or likelihood analysis  ..................................................................................................................................28  Table 2 Species in Alismatales with pseudogenization or loss of at least one plastid-encoded ndh subunit gene  ......................................................................................................29  Table S1  Specimen source information  ..................................................................................47  Table S2 DNA substitutions models and partitioning scheme and partition subsets resulting from two different partition analyses conducted in PartitionFinder using the BIC criterion  ....................................................................................................................51      viii  LIST OF FIGURES  Figure 1 Relationships among Araceae, core alismatids, and Tofieldiaceae inferred in parsimony and likelihood analyses  ..........................................................................30  Figure 2 Core alismatid phylogeny inferred in a likelihood analysis of 83 plastid genes using the “GxC-n” partitioning scheme  ............................................................................32  Figure 3  Parsimony reconstructions showing predicted losses of the NADH dehydrogenase complex genes in Alismatales  .................................................................................34  Figure 4  Branch support in core alismatids compared to other studies  .................................36  Figure S1  Phylogram of angiosperm phylogeny inferred from a parsimony analysis of coding regions from whole plastid genomes  .......................................................................55  Figure S2 Phylogram of angiosperm phylogeny inferred from an parsimony analysis of coding regions from whole plastid genomes in which ndh genes were entered as blanks for those species in which one or more of these was pseudogenized or lost .................57  Figure S3 Phylogram of angiosperm phylogeny inferred from a unpartitioned likelihood analysis of coding regions from whole plastid genomes  .........................................59  Figure S4  Phylogram of angiosperm phylogeny inferred from a codon-partitioned likelihood analysis of coding regions from whole plastid genomes  .........................................61  Figure S5  Phylogram of angiosperm phylogeny inferred from a gene-by-codon (GxC) partitioned likelihood analysis of coding regions from whole plastid genomes  .....63  Figure S6 Phylogram of angiosperm phylogeny inferred from a gene-by-codon (GxC) partitioned likelihood analysis of coding regions from whole plastid genomes in which ndh genes were entered as blanks for those species in which one or more of these was pseudogenized or lost ...............................................................................65      ix  ACKNOWLEDGMENTS I would like to start by offering my sincere thanks to my supervisor Dr. Sean Graham who took a chance and gave me an exciting research topic.  The experiences and knowledge gained throughout my Masters would not have been possible without his guidance and persistence in all aspects of my project.  I would also like to extend my appreciation to my committee members, Wayne Maddison and Quentin Cronk for their help throughout my thesis. I would like to thank all my lab mates that have been with me throughout most or all of my project, Will Iles, Vivienne Lam, Marybel Soto Gomez, Qianshi Lin, Erin Fenneman, Isabel Marquez, Hayley Darby, Wesley Gerelle and David Bell with special thanks to Will for all his initial and continual guidance with Alismatales, and Vivienne and Marybel for all their support on the many academic trips we took together.   I would also like to thank fellow graduate students Dave Tack for his continual (and extensive) help with computer-related issues and Kate McGrath for all the conversations that helped me conceive of new ways to tackle the variety of problems faced in my degree.  I would like to thank my parents for all the extra support, including financial support, when the need arose.  Finally, a special thanks to Charlotte Veerman for her continual emotional support through all my years at UBC.   1  1. INTRODUCTION  The order Alismatales represents one of the deepest divisions in monocot phylogeny (e.g., Davis et al., 2004; Givnish et al., 2006; Graham et al., 2006; Qiu et al., 2006; Ruhfel et al., 2014), one of sixty major clades currently recognized at the ordinal level in the angiosperms (APG 2009). It includes one of the largest and most diverse plant families (Araceae, ~3800 species; Mabberley, 1997; Henriquez et al., 2014) and the largest clade of aquatic angiosperms, a group of ~12 exclusively aquatic families that provide spectacular examples of adaptation to life in aquatic environments (Sculthorpe, 1967; Les et al., 1997), including the seagrasses, the only plant examples of reversions to life in marine habitats. The order also includes Tofieldiaceae, a small herbaceous family once considered to represent ‘primitive’ monocot stock based on combinations of vegetative and floral characteristics that were perceived to be plesiomorphic (see Tamura, 1998).  Several studies have addressed overall phylogenetic relationships in the order (Les et al., 1997, Petersen et al., 2006a; Li and Zhou, 2009; Iles et al., 2013; Les and Tippery, 2013), and there has been a large collection of studies focused on relationships within individual families and genera (reviewed in Les and Tippery, 2013). Most of these were based on one or a few genes. Although the broad outline of Alismatales phylogeny is now clear, there are also numerous cases where higher-order relationships are unclear. Examples of major uncertainties include the deepest phylogenetic split in the order (most studies recover either Araceae or Tofieldiaceae as the sister group of the remaining taxa, typically with poor support: Chase et al., 2000; Qiu et al., 2000, 2006, 2010; Davis et al., 2004; Tamura et al., 2004; Chase et al., 2006;  2  Duvall et al., 2006; Givnish et al., 2006; Graham et al., 2006; Petersen et al., 2006b; Li and Zhou, 2007; Soltis et al., 2007; Cuenca et al., 2012; von Meering and Kadereit, 2010; Azuma and Tobe, 2011; Iles et al., 2013), the relative placements of Maundiaceae, Ruppiaceae and Scheuchzeriaceae within a large clade of tepaloid alismatids (Les et al., 1997; Chen et al., 2004; Les et al., 2005; Petersen et al., 2006b; Waycott et al., 2006; Ito et al., 2010; von Meering and Kadereit, 2010; Iles et al., 2013; Les and Tippery, 2013; Sokoloff et al., 2013), and multiple aspects of relationship in the larger families, including Alismataceae, Cymodoceaceae, Hydrocharitaceae and Potamogetonaceae (Les et al., 1993; Les et al., 1997; Tanaka et al., 1997; Haynes et al., 1998; Chen et al., 2004; Lindqvist et al., 2006; Les et al., 2006; Lehtonen, 2006; Petersen et al., 2006b; Jacobson and Hedrén, 2007; Lehtonen and Myllys, 2008; Li and Zhou, 2009; Les et al., 2010; Chen et al., 2012a; Chen et al., 2012b; Petersen et al., 2014; note that many of these are genus-level studies).   Recent advances in DNA sequencing technology have made the retrieval of whole plastid genomes, or at least full sets of plastid genes, a relatively straightforward technical challenge (e.g., Cronn et al., 2008; Steele et al., 2012). As a result, phylogenetic data sets featuring this level of gene sampling are increasingly commonly used to address higher-order relationships that were unresolved or poorly supported. Recent broad-scale examples include studies of land plants as a whole (e.g., Ruhfel et al., 2014) and of other major groups of angiosperms and gymnosperms (e.g., Parks et al., 2009; Moore et al., 2010; Jansen et al., 2011), including recent studies of monocot orders (e.g., Givnish et al., 2010; Steele et al., 2012; Barrett et al., 2013; Barrett et al., 2014) and of Araceae in Alismatales (Henriquez et al., 2014). Data sets of this size offer unparalleled opportunities for revisiting higher-order relationships that were left unresolved in earlier studies, but they also present new analytical challenges related to large-scale data and  3  the diversity of component genes or character sets (data partitions), in terms of their DNA substitutional dynamics (e.g., Barrett et al., 2013).  These data sets also provide substantial information on genome composition and molecular evolution, allowing us to study patterns of gene loss in heterotrophic taxa (e.g., Barbrook, 2006; Krause et al., 2008; Wick et al., 2011; Merckx, 2013) and genome rearrangement (e.g., Cai et al., 2008; Guisinger et al., 2010; Wick et al., 2011). NADH dehydrogenase, a major plastid protein complex coded primarily by the plastid genome, has apparently been lost multiple times in Alismatales, based on predictions made by studying genes coding for several proteins in the complex (Iles et al., 2013). A recent plastid genome for a core alismatid taxon, Najas flexilis (Peredo et al., 2013) confirmed that these genes have been lost or degraded in one lineage of the core alismatids. I addressed these issues using whole plastid gene sets from a taxonomically dense sample of Alismatales, which included the bulk of the genera in Tofieldiaceae and the core Alismatales families, and a phylogenetically representative sample of lineages in Araceae (Henriquez et al., 2014). I used this large-scale phylogenomic data set to address the following specific questions: (1) Does this large-scale data set resolve current uncertainties in Alismatales phylogeny with strong support? (2) How much variation is there in patterns of branch support when different criteria are used to infer phylogenetic relationships? For example, do different data partitioning schemes for likelihood analyses matter? (3) What is the phylogenetic distrubution of taxa lacking a functional plastid NADH dehydrogenase complex, and how many losses of this complex were there in the phylogenetic history of Alismatales?     4  2. MATERIALS AND METHODS  2.1 Taxonomic sampling I included new plastid genomes for 54 species (specifically, full plastid gene sets representing the 79 protein-coding genes found in most angiosperms and four rDNA genes; Raubeson and Jansen 2005), of which I generated data for 43 species (with data for 11 species generated by C. Barrett, California State Los Angeles, and G. Davis, Cornell University). These taxa represent 49 of ~54 genera in the core alismatids (Alimatales sensu Les and Tippery, 2013), and three of ~four genera in Tofieldiaceae (Remizowa et al., 2011; Campbell and Dorr, 2012) (Appendix 1). The full taxon sampling for analysis included an additional seven monocot plastid genomes recently generated by M. Soto and V. Lam (UBC), in addition to 95 previously sequenced angiosperm genomes retrieved from GenBank and from matrices presented in Barrett et al. (2013; a monocot-focused sampling) and Henriquez et al. (2014; sampling focused on Araceae) (Appendix 1). I included representatives of all currently recognized orders of monocots (APG, 2009), in addition to representatives of ANITA-grade angiosperms, eudicots and magnoliids.   2.2 DNA and library preparation I performed DNA extractions using the method of Doyle and Doyle (1987) or by using a DNeasy plant mini kit (Qiagen, Germantown, Maryland, USA), eluting in water. I used paired-end whole genome shotgun sequencing on an Illumina HiSeq 2000 (Illumina Inc, San Diego, California, USA) to retrieve plastid genomes, making the sequencing libraries with a variety of kits (the choice of kit depended in part on the quantity of genomic DNA). Libraries for most species were  5  generated using a Bioo NEXTflex DNA sequencing kit (Bioo Scientific, Austin, Texas, USA), following the protocols outlined in the kit, which requires 1000 ng of starting DNA. For a subset of samples for this kit I first prepared whole genome amplifications of genomic DNA using a REPLI-g Mini Kit, Qiagen (samples were amplified with five replicates and then pooled, to reduce amplification bias). I used two other kits (Kapa Library Preparation Kit, Kapa Biosystems, Wilmington, Massachusetts, USA; Nugen Ovation Library systems, NuGEN, San Carlos, California, USA) when DNA amounts were lower (I was successful with as little as 4 ng of starting DNA using the Nugen kit). For all three kits, genomic DNA was sheared to 400 bp fragments using a Covaris sonicator (model: S220, Woburn, Massachusetts, USA). I performed gel size-selection for library preparation with the Bioo kit (~400 bp, plus the adapter size, an additional ~130bp), purifying the resulting DNA from a 2% agarose gel using a Zymoclean gel recovery kit (Zymo Research, Irving, California, USA). For the Kapa and Nugen kits I used magnetic bead size selection (Agencourt AMPure XP magnetic beads, Beckman Coulter Genomics, Brea, California, USA), because DNA concentrations for these samples were typically too low for gel extraction. I multiplexed individual libraries on several Illumina lanes (16 to 30 samples per lane) (Cronn et al., 2008) after performing quality-control checks on completed libraries. An initial Qubit reading was taken to ensure library concentrations were at least 0.5 ng/uL (Qubit 2.0 Flurometer, Life Technologies, Thermo Fisher Scientific, Waltham, Massachusetts, USA). For samples that passed the first quality control step, a Bioanalyzer analysis was run to ensure the libraries had the required size distributions (2100 Bioanalyzer, Agilent Technologies, Santa Clara, California, United States), and a qPCR (quantitative polymerase chain reaction) reaction was carried out to measure the concentration of the library precisely, to allow for balance in the amounts used in multiplexed reactions.  6    2.3 Contig assembly and plastid gene annotation The sequence reads resulting from the Ilumina sequencing runs were first processed using CASAVA 1.8.2 (Illumina Inc, San Diego, California, USA) to pull individual sequences from the multiplexed data. I made de novo assemblies for individual plastid genomes based on these filtered FASTQ reads, using CLC Genomics Workbench 6.5.1 (CLC Bio Aarhus, Denmark) with default settings, occasionally using Sequencher v.4.8 (Gene Codes, Inc., Ann Arbor, MI, USA) as a further aid in assembly. Following assembly, I filtered all contigs <500 bp in length and with <10x coverage, and exported the resulting contigs in FASTA format, in a single file per species. I then performed a local BLAST search (Altschul et al., 1990) with this as a query against the published full plastid sequence of Wolffia australiana (GenBank accession: NC015899) to sort plastid contigs from mitochondrial and nuclear ones. Mitochondrial inserts of plastid genes were identified occasionally (BLAST searches of suspect contigs identified contiguous mitochondrial sequences), and were removed. FASTA files containing only plastid contigs were uploaded to the online annotation program DOGMA (Wyman et al., 2004). I retrieved and annotated up to 79 protein coding genes and four plastid rDNA genes per species; in several cases fewer genes were recovered due to poorer assemblies, or because of gene loss. With the exception of the taxa with loss of subunits of plastid genes in the ndh complex, all but four taxa (Limnocharis flava, Astonia australiensis, Zostera mulleri, Pleea tenuifolia) had a minimum of 80 of the 83 genes recovered, with fewer than 70 genes recovered in only one case (Pleea tenuifolia; 62 genes for this taxon, due to a poorer quality run). I used Sequencher and BLAST to confirm gene start and end locations for ycf1, pseudogenized genes and several genes with stretches of Ns, with reference to gene boundaries in Wolffia australiana, Elodea canadensis (GenBank accession: NC018541)  7  and Stratiodes aloides (this study). In cases where DOGMA failed to produce any evidence of a retention of an ndh subunit in one of the taxa showing pseudogenization, I carried out a follow-up search in Sequencher. In many cases this was successful in recovering additional pseudogenized fragments of the subunits. If no remnants of the pseudogene were recovered after extensive Sequencher-based searching, I considered the subunit lost in that species. The final set of genes was output in a single FASTA file for each species.  2.4 Alignment and matrix construction I used a python script to convert the FASTA files from each species into separate gene-based FASTA files. The latter were combined with gene files obtained from GenBank and the published alignments (Barrett et al., 2013; Henriquez et al., 2014), creating gene files with 152 sequences (species) per file (blanks were left when the gene was absent for a taxon). I aligned genes individually using MUSCLE v3.8.31 (Edgar 2004) with default parameters, followed by manual adjustment in Mesquite (Maddison and Maddison, 2007) using alignment criteria laid out in Graham et al. (2000), staggering hard-to-align regions in the alignment (e.g., Saarela and Graham, 2010). I adjusted the ends to make sure that all alignments for protein-coding genes were maintained as open reading frames (i.e., full triplets were set at the start and end of each locus; where necessary including blank columns or cells). Gap cells were coded as missing data. MUSCLE could not align the ycf1 locus (apparently due to its sequence complexity and large size; ~5.5 kb per taxon, unaligned), and so I used Mafft v7.164 (Katoh, 2013) for it. I did not attempt to maintain reading frame for this locus as the alignment was long and complex, but I did otherwise check the alignment carefully in Mesquite, and trimmed nucleotides not present for at least half the taxa (including ends) using GBlocks v0.91b (Talavera and Castresana, 2007). I  8  combined individual alignments into a concatenated set by pasting them into a single interleaved NEXUS-based matrix. The final alignment length was 98,305 bp (derived from 69,444 bp of unaligned plastid sequence data in Wolffia australiana, for reference).   2.5 Error checking To ensure that copy-paste errors or other editing errors were not introduced during alignment and matrix construction I performed cross-checks in Sequencher, by comparing concatenated sequences for each taxon in the final matrix back to original de novo contig files or published sequences for that taxon (after stripping them of gaps). I also checked for potentially anomalous sequences by inferring heuristic parsimony trees for each gene using PAUP* v4.0a134 (Swofford, 2003), and then checking these visually to trees inferred in Les et al. (1997), Henriquez et al. (2014), and Barrett et al. (2013) for core alismatids, Araceae and the rest of the monocots, respectively. I checked them for sequences placed in an obviously spurious position that closely resembled a taxon known to be a distant relative: none were observed. Gene boundaries were set up in the NEXUS file using CHARSETs. The original gene start and stop points were then cross-checked to the original genes using the “Show data matrix” function in PAUP*. The reading frames of the protein-coding genes were also verified by translation of the matrix in Se-Al (Rambaut, 2002).  2.6 Phylogenetic inference I analyzed the concatenated DNA sequence alignment using a heuristic parsimony analysis in PAUP*, with tree-bisection-reconnection branch-swapping, 100 stepwise addition replicates, and otherwise using default settings. I repeated this analysis using the full data set and for a version  9  of the data set in which I entered all ndh subunits as blanks for those species in which one or more subunit was pseudogenized or lost (Table 2), but still retaining the ndh genes for other taxa.  I also performed several different maximum likelihood (ML) searches using RAxMLv7.4.2 (Stamatakis, 2006a), considering 10 multiple independent searches per analysis. One ML search was performed with all data unpartitioned (the ‘unpartitioned’ scheme), but I also partitioned data for ML analysis according to two data partitioning schemes. The first of these, referred to as the ‘codon’ partitioning scheme below, allocated nucleotides in protein-coding genes according to whether they belonged to the first, second or third codon position, with the four rDNA genes combined in a fourth data partition (a short overlapping fragment of psbD and psbC was counted only for psbC), and the ycf1 gene included as a fifth partition (the underlying reading frame of this gene was not constrained following alignment, see above). I used PartitionFinder v1.1.1 (Lanfear et al., 2014) to assess which data partitions have significantly different DNA substitution models or model parameters, using the BIC criterion and the hierarchical clustering algorithm. All five data partitions were retained as distinct data partitions using this test (Table S2).  I also set up a second partitioning scheme, which I refer to below as the ‘GxC’ (gene-by-codon) scheme, by initially allocating nucleotides in protein-coding genes to three different codon positions per gene (excluding ycf1, see below, and with the trans-spliced exons of rps12 treated operationally as distinct genes). The resulting 237 initial partitions were assessed using PartitionFinder to combine those that did not have significantly different DNA substitution models. A total of 93 distinct data partitions remained after this procedure (Table S2), and the subsequent likelihood analysis included an additional four partitions (several genes were not included in the PartitionFinder analysis because of apparent limitations in the maximum number  10  of partitions that can be considered (ycf1 and the 23S and 16S rDNA genes were treated as additional partitions; I combined two small genes, 5S and 4.5S rDNA in a single partition). Finally, I performed a likelihood analysis with the same partitioning scheme as the GxC scheme, but with all ndh subunits entered as blanks for those species in which one or more subunit was pseudogenized or lost (Table 2), which I refer to as the ‘GxC-n’ scheme (‘-n’ = minus ndh genes, for affected taxa).  The PartitionFinder analysis for the codon scheme supported the GTR+I+Γ model or close variants (Table S2a); the analysis for the GxC scheme supported the GTR+ Γ or GTR+I+Γ model for different partitions (Table S2b). I used the GTR+Γ model for the heuristic ML search for the best tree for the full matrix in the unpartitioned analysis, and for individual data partitions in the partitioned analyses. I did not use a separate ‘I’ parameter to account for invariable sites in searches for the best tree or bootstrap analyses, because these may be accommodated by the gamma parameter, Γ (Yang, 2006); the TVM and TIM DNA substitution models found for some partitions in the codon model are not supported in the version of RAxML I used for analysis.  I performed bootstrap analyses (Felsenstein 1985) to assess the strength of branch support, using 1,000 replicates for the parsimony search with 100 random addition replicates per bootstrap replicate, and 200 replicates for the ML searches. I considered well-supported branches to have bootstrap support of at least 95%, and poorly supported branches to have less than 70% of bootstrap support, following Zgurski et al. (2008). For the ML bootstrap analysis I used a rapid search, and employed the GTRCAT approximation of the GTR+Γ model (Stamatakis 2006b).  I also conducted parsimony and likelihood character-state reconstructions of the evolutionary history of ndh loss/pseudogenization using Mesquite. I repeated these analyses  11  across all of the best trees found in the different parsimony and likelihood heuristic searches, pruned to Alismatales. I coded individual species as “1” (for those that retain all plastid-encoded ndh genes as open reading frames) or “0” (for species in which one or more ndh genes is absent or has an interrupted reading frame). For parsimony analysis I used unordered character states for ancestral-state reconstruction. For the likelihood analysis I compared the overall likelihood for one- vs. two-rate (asymmetric) models. The one-rate model (the Markov model of Lewis, 2001) was the better-fitting model according to Mesquite, based on the likelihood ratio test in Pagel (1999), and so this was the model used for likelihood reconstructions of loss of function. In the likelihood reconstructions, I considered an ancestral state of a node to be statistically significant when the difference in log-likelihood scores for the two possible characters states was greater than 2.0 (Edwards, 1972).     12  3. RESULTS  3.1 Phylogenetic position of Alismatales, and resolution of the first split in the order  Outgroup relationships in the two parsimony and four likelihood analyses performed here are generally congruent with each, and well supported by bootstrap analysis (Figs. S1-S6). Alismatales are strongly supported in all analyses as the sister group of all monocots, excluding Acorus (Acorales). The very first phylogenetic split in Alismatales is uncertain in all analyses: five of six analyses recover Tofieldiaceae as the sister group of other Alismatales in the best trees, with weak support for this arrangement from likelihood analyses, and moderately strong support from the parsimony analyses (54-55% bootstrap support from three of the four ML analyses; 75-76% support from the two parsimony analyses; Fig. 1). An alternative weakly supported arrangement is recovered in one analysis, with Araceae depicted as the sister group of the other Alismatales (48% bootstrap support in the GxC ML analysis; fig 1e). Relationships in Araceae agree with those in Henriquez et al. (2014), from which these sequences were taken, with a few exceptions concerning branches that are poorly supported here and in that study; these relationships are not discussed further here (see Henriquez et al., 2014, for further details).  3.2 Relationships in core Alismatales In general, most branches in core Alismatales are strongly supported and congruent across the parsimony and likelihood analyses (Fig. 2 shows the result for the GxC-n ML analysis; Table 1 summarizes bootstrap support across analyses for branches in shortest trees that had less than 100% bootstrap support from at least one analysis). Considering the 50 internal branches in the  13  rooted subtree that corresponds to the core Alismatales, 38 have 100% bootstrap support from parsimony analysis (41 branches have at least 95% bootstrap support), and 40-43 branches have 100% bootstrap support from likelihood analysis (44-46 branches have at least 95% bootstrap support). The implications of these trees for our understanding of relationships between and within the core alismatid families are discussed below.  3.3 Loss of ndh genes Ten taxa lack one or more plastid genes coding for ndh subunits using the search criteria outlined in the Materials and Methods section, and in all cases they also show interruptions in all or most of the remaining ndh genes, likely consistent with the affected loci being pseudogenes (Table 2). These absences and putative pseudogenes are consistent with a loss of functionality of the plastid NADH dehydrogenase complex.  Parsimony and likelihood character-state reconstructions indicate four independent losses of NADH dehydrogenase function across Alismatales. One loss scenario is shown in Fig. 3 (based on Fig. S6); the same four branches are predicted to have losses for all ancestral-state reconstructions on the best parsimony and likelihood trees (see Fig. S1-S5), and considering parsimony and likelihood character-state reconstructions (data not shown). A single loss is predicted in a subclade of Hydrocharitaceae (i.e., the clade comprising the three seagrasses Enhalus, Thalassia and Halophila, and three freshwater submerged species, Najas, Nechamandra and Vallisneria); two losses are indicated in the tepaloid Alismatales (one involving a well-supported clade comprising two species in the seagrass family Cymodoceaceae, Amphibolis and Thalassodendron; the other observed in another seagrass taxon, Posidonia, Posidoniaceae); and a single loss is indicated in Tofieldiaceae (Triantha occidentalis).    14  4. DISCUSSION  4.1 The phylogenetic positions of Alismatales and Acorus in monocot phylogeny My study provides strong support for Alismatales as the sister group of all monocots except Acorus (Acorales). This arrangement is unaffected by the inclusion or exclusion of ndh genes for taxa in which these have become pseudogenized (e.g., compare Figs. 1a and 1b, 1e and 1f). This result is in line with most molecular phylogenetic studies that include or focus on plastid data (e.g., Tamura et al., 2004; Givnish et al., 2006; Soltis et al., 2007; Givnish et al., 2010; Azuma and Tobe, 2011; Iles et al., 2013). However, several studies with extensive mitochondrial data or that exclusively focus on genes from this genome, infer Acorus to be the sister group of core alismatids, or a clade with Araceae and Tofieldiaceae, or more distantly related to them (Qui et al., 2000; Davis et al., 2004; Petersen et al., 2006a), or even distantly related to any members of Alismatales (for the mitochondrial gene cox1 in Duvall et al., 2008). These diverse positions for data sets that have mitochondrial data, and the conflict between plastid and mitochondrial results, may be due to strongly elevated rates of evolution for either genome in some genes or lineages (long-branch attraction, see Felsenstein, 1978, 1983; Hendy and Penny, 1989), or may result from parallel instances of retroprocessed, RNA-edited loci, prevalent in many mitochondrial genes (Bowe and dePamphilis, 1996; Petersen et al., 2006a, b, 2013; Duvall et al., 2008). RNA editing may be ruled out as a possible cause of Acorus placement for mitochondrial genes (G. Petersen, University of Copenhagen, pers. comm), and it seems unlikely to be a problem for plastid loci because of its generally low incidence in angiosperm plastid genomes (e.g., Fryer et al., 1997; Tsudzuki et al., 2001; Inada et al., 2004).   15  It is possible that very long branches in Alismatales and Acorus (e.g., see Fig. 4 in Davis et al., 2004 for mitochondrial atpA) may also influence their placement in monocot phylogeny. Several relevant lineages in my plastome-based phylogeny also show evidence of substantial rate elevation (e.g., Fig. 2, S6). However, these instances appear to be limited to lineages that are well nested in the petaloid and tepaloid clades of core alismatids (in particular, subfamilies Hydrilloideae and Hydrocharitoideae of Hydrocharitaceae, several lineages within Alismataceae, and a sub-clade of tepaloid alismatids that includes seagrasses). They also appear to exclude Acorus, Araceae and Tofieldiaceae, which do not show obvious signs of substantial rate elevation here (Figs. S1-S6). Rate elevation in the plastid genome may therefore be an unlikely explanation for the conflict between organellar genomes, and the source of this incongruence needs further attention. A recent study based on a combined analysis of 18S and 26S rDNA genes in a large-scale sampling of angiosperms and relatives (Maia et al., 2014) also recovered Acorus as the sister group of all monocots (however, Alismatales monophyly was weakly contradicted in their study, as Tofieldiaceae did not form a clade with Araceae and core alismatids)  4.2 Resolving the first split in Alismatales phylogeny The first split in Alismatales phylogeny is uncertain, even with 83 genes included (summarized in Fig. 1 for the six analyses performed here). Most of my analyses identify Tofieldiaceae as the sister group of all other Alismatales, but the support for this arrangement is moderate at best (75-76% in parsimony analysis, weaker support in likelihood analysis). This appears to be largely unaffected here by the inclusion vs. exclusion of all ndh genes (Fig. 1) in taxa for which these genes have become pseudogenized or lost (i.e., for several core alismatid lineages and a member  16  of Tofieldiaceae; Fig. 3). The ‘Tofieldiaceae-sister’ arrangement has been observed in several other large-scale studies (Chase et al., 2000; Duvall et al., 2006; Graham et al., 2006; Iles et al., 2013). One of my analyses supported an alternative arrangement in which Araceae are the sister group of other Alismatales (i.e., the likelihood analysis with GxC partitioning scheme; Figs. 1e, S5), but with only 48% bootstrap support for that arrangement (Fig. 1e) vs. 49% for the arrangement with Tofieldiaceae sister, in the same analysis; note that a similar analysis but with ndh genes excluded supported the contrasting Tofieldiaceae-sister arrangement (with 55% support; Fig. 1f, S6). Other studies have supported either of two alternative arrangements of these three lineages (e.g., Qiu et al., 2000, 2006, 2010; Davis et al., 2004; Tamura et al., 2004; Chase et al., 2006; Givnish et al., 2006; Petersen et al., 2006b; Li and Zhou, 2007; Soltis et al., 2007; Cuenca et al., 2012; von Meering and Kadereit, 2010; Azuma and Tobe, 2011), see also Les and Tippery (2013), with the bulk favouring Araceae as the sister group of other Alismatales. Generally this arrangement lacked strong support in these analyses. My likelihood analyses point to a possible explanation for this uncertainty: the branch supporting the different arrangements is extremely short here (i.e., Fig. 1c-f). This uncertainty may therefore reflect a very rapid evolutionary radiation in the order that may be hard to resolve, although it is possible that large nuclear genome-scale data sets will provide additional illumination on this difficult problem.  4.3 Relationships in core alismatids: Higher-order relationships The division of Alismatales into two major clades that bear either petaloid vs. tepaloid floral forms (Posluszny et al., 2000) has been recognized since the first major molecular phylogenetic study of the order by Les et al. (1997), and has been recovered in other studies (e.g., Chen et al.,  17  2004; von Mering and Kadereit, 2010). Strong support for this relationship was found by Iles et al. (2013) and in my study; see Fig. 4, which summarizes support for relationships in several higher-order phylogenetic studies. I consistently found strong support for family monophyly across all of the analyses performed here for all families where more than one genus is recognized (Fig. 2, S1-S6); note that Aponogetonaceae; Butomaceae; Maundiaceae; Posidoniaceae; Ruppiaceae; Scheuchzeriacae are all monogeneric (Butomaceae, Maundiaceae and Scheuchzeriacae are monotypic; I did not sample more than one species per genus in the other monogeneric families).   4.4 Relationships in core alismatids: The petaloid clade Within the petaloid clade, I recovered Butomaceae (Butomus) as the sister group of Hydrocharitaceae. This relationship is known to be sensitive to limited taxon sampling (Iles et al., 2013), and was only weakly supported in the rbcL-based study of Les et al. (1997). It was recovered with strong support by Iles et al. (2013), and also here (Fig. 2, 4). My taxonomic sampling in Alismataceae and Hydrocharitaceae is nearly complete at the genus level (I am missing only Albidella, Burnatia and Limnophyton in Alismataceae, and Appertiella and Hydrilla in Hydrocharitaceae), a sampling that has been matched only by Les and Tippery (2013) in an expansion and reanalysis of the rbcL-based data set of Les et al. (1997). I recovered nearly the same topology for Alismataceae as Les and Tippery (2013) did with a single gene, generally with strong support for most branches here. Some of these branches were also well supported in the study of Les and Tippery (2013), but multiple branches with weak support there are strongly supported here (summarized in Fig. 4). This includes the nested placement of three taxa sometimes referred to as Limnocharitaceae (Butomopsis, Hydrocleys and Limnocharis)  18  within Alismataceae (Fig. 2). In total, only two branches are inferred to have less than 95% support for any parsimony or likelihood analysis performed here (Fig. 2, Table 1). One of these concerns the sister-group relationship between Echinodorus and Helanthium (which was moderately to strongly supported in all parsimony and likelihood analyses; clade g in Fig. 2 and Table 1); the other concerns a possible relationship between Butomopsis and Limnocharis (weakly to strongly resolved across analyses here as clade h in Fig. 2 and Table 1). The latter is the only conflict with the results of Les and Tippery (2013), who instead recovered a strongly supported clade comprising Butomopsis and Hydrocleys.  My sampling of Hydrocharitaceae is almost identical to that of Les and Tippery (2013) at the genus level, although I did not sample Hydrilla. Multiple branches in the study of Les and Tippery (2013) had only weak to moderate support, including relationships among the four subfamilies (Anacharidoideae, Hydrilloideae, Hydrocharitoideae, Stratiotoideae; Fig. 4). Here I recovered a well-supported clade comprising Hydrilloideae and Hydrocharitoideae in all analyses. I also found Anacharidoideae to be the sister group of this clade in all analyses (Figs. S1-S6), but there was weak to at best moderate bootstrap support for this arrangement across analyses (see clade a in Fig. 2 and Table 1). This arrangement effectively places Stratiotes (subfamily Stratiotoideae) as the sister group of all other taxa in the family. Previous studies recovered various relationships among these four subfamilies with weak support (Les et al., 2006; Petersen et al., 2006a; see the weak disagreements noted with arrows in Fig. 4 here for Hydrocharitaceae). Iles et al. (2013) also recovered Stratiotes in this position in an analysis based on 17 plastid genes. In contrast, Les and Tippery (2013) found Hydrocharitoideae to be the sister group of all remaining Hydrocharitaceae, but with <50% bootstrap support for that arrangement.   19  The only other disagreements with the results of Les and Tippery (2013) concerning Hydrocharitaceae are: (1) the position of Najas within Hydrilloideae, which we recovered as the sister group of a clade comprising the three hydrocharit seagrasses (Enhalus, Halophila and Thalassia) and two fresh-water taxa, Nechamandra and Vallisneria. That relationship was well supported in likelihood analysis, but only weakly supported in parsimony analysis (clade b in Fig. 2 and Table 1). In contrast, Les and Tippery (2013) recovered Najas as the sister group of the hydrocharit seagrasses, but with poor support. (2) The relative arrangement of Apalanthe, Egeria and Elodea. My parsimony analysis weakly supported a clade comprising Apalanthe and Egeria (clade n in Figs. S1-2; Table 1), whereas likelihood analysis consistently supported a sister-group relationship between Egeria and Elodea, with strong support (clade c in Fig. 2; Table 1). The latter relationship is supported by Les et al. (2006) who also found Egeria to be the sister group to Elodea, with Apalanthe then sister to the other two with strong support, using plastid and nuclear internal transcribed spacer (ITS) regions. The monophyly of the genus Egeria is unclear (Fig. 6.10 in Les and Tippery, 2013), but these workers also recovered clade c (see Fig. 2) in a separate analysis of the nuclear ITS region (the same arrangement with respect to the species in common in our studies). This result is also the only instance of a disagreement in the petaloid clade here between my parsimony and likelihood analyses.  4.5 Relationships in core alismatids: The tepaloid clade Almost all higher-order relationships among the eight-nine families currently recognized in the tepaloid clade of core Alismatales (APG, 2009; Les and Tippery, 2013) are strongly supported across all analyses performed here (Figs. 2, 4, S1-S6; Table 1). I recovered Aponogetonaceae and Scheuchzeriaceae as successive sister groups of the rest of this clade, in line with Petersen et  20  al. (2006b) and Iles et al. (2013); this relationship was well supported in likelihood analysis but more poorly supported in parsimony analysis (clade i in Fig. 2, Table 1). Iles et al. (2013) were the first workers to recover this relationship with strong support (Les et al., 1997, recovered a reciprocal arrangement using rbcL, but this was poorly supported). The next successive split is between Juncaginaceae and the remaining taxa, with the monotypic Australian genus Maundia (Maundiaceae) confirmed here with strong support (e.g., Fig. 2) to have a phylogenetic position that is isolated from its original home in Juncaginaceae (von Mering and Kadereit, 2010; Iles et al., 2013; Les and Tippery, 2013; Sokoloff et al., 2013). Maundiaceae are the sister group of a clade comprising two major groups: Zosteraceae (the zosterid seagrasses; Zostera and Phyllospadix) and Potamogetonaceae which comprise a clade that is the sister group of the ‘Cymodoceaceae complex’ (Les et al., 1997). The latter includes three families, Posidoniaceae, Ruppiaceae and Cymodoceaceae. These relationships are all well supported here (e.g., Fig. 2), as is a sister-group relationship between Ruppiaceae and the seagrass family, Cymodoceaceae. In contrast, Ruppia was weakly supported as nested within Cymodoceaceae in Iles et al. (2013), who included only two members of Cymodoceaceae in their analysis. However, a sister-group relationship between Ruppia and Cymodoceaceae is still consistent the view of Les and Tippery (2013) that Ruppia should be recognized under the latter family, a solution that optimizes phylogenetic information when a small taxon is the sister group of a larger taxon with the same rank (Backlund and Bremer, 1998). Using the same principle, the small monogeneric seagrass family Posidoniaceae could be collapsed into Cymodoceaceae (this family name that has taxonomic priority in the complex), and similarly, Butomaceae in Hydrocharitaceae.  Many of the inter-familial relationships in the tepaloid clade were recovered for the first time with strong support in Iles et al. (2013), although not always with strong support in their  21  parsimony analysis (Fig. 2). My expanded taxonomic and genomic sampling has improved the strength of relationships based on parsimony bootstrap analysis, and my data provide the first strongly support resolution of the relationships among the three lineages within Cymodoceaceae s.l. (i.e., Ruppia, Posidonia and Cymodoceaceae s.s.; Fig. 2). Three tepaloid-clade families have more than two genera: Cymodoceae, Juncaginaceae, and Potamogetonaceae. The relationships inferred among the three genera of Juncaginaceae (Fig. 2) are well supported and consistent with other recent studies, including von Mering and Kadereit (2010) and Les and Tippery (2013); note that Triglochin scilloides (not sampled here) is sometimes referred to as Lilaea scilloides (Lilaeaceae), but is clearly a species of Triglochin (Juncaginaceae) (von Mering and Kadereit, 2010; Les and Tippery, 2013). My taxon sampling in Potamogetonaceae excludes Althenia and Pseudalthenia (which remain unsampled in a phylogenetic analysis; Les and Tippery, 2013). Relationships among the genera sampled here have eluded satisfactory resolution to date (Les and Tippery, 2013). The arrangements seen in the best parsimony and likelihood trees disagree within Potamogetonaceae regarding the position of Groenlandia (this is only instance of this kind of conflict in the tepaloid clade). The parsimony results support Groenlandia as the sister group of all other Potamogetonaceae with moderate bootstrap support (labelled as clade o in Fig. S1-S2 and Table 1), consistent with the arrangement of these taxa in Les et al. (1997; weakly supported there). All likelihood analyses here support Groenlandia as the sister group of a clade comprising Lepilaena and Zannichellia, with poor to strong support for this arrangement (support is strongest in the GxC partitioning schemes, see clade l in Fig. 2; Table 1). In all analyses Potamogeton is the sister group of Stuckenia, and Lepilaena is the sister group Zannichellia at current taxon sampling. The latter clade is no longer recognized as  22  Zannichelliaceae, a taxonomic solution that is widely accepted (APG, 1998; Les and Tippery, 2013), and that is supported here regardless of uncertainty over the placement of Groenlandia. The most comprehensive sampling of Cymodoceaceae s.s. in phylogenetic studies are those of Les and Tippery (2013), which was based on rbcL, and Petersen et al. (2014), which was based on two plastid genes and four mitochondrial genes. Systematic inferences in the family are complicated by the non-monophyly of the genus Cymodocea, a result that is supported here (considering the two species we sampled in this genus) and in Les and Tippery (2013) and Petersen et al. (2014). Les and Tippery (2013) suggested that three genera (Amphibolis, Syringodium and Thalassodendron) could be recognized under Cymodocea. However, they deferred nomenclatural changes until taxon sampling in future studies is improved. The relationships that I inferred among the genera of Cymodoceaceae s.s. are completely consistent with those reported in Petersen et al. (2014) (Fig. 4), generally with moderate to strong support here. Parsimony bootstrap support values for these relationships tend to be weaker than likelihood (clades j and k in Fig. 2, Table 1). Halodule was strongly supported as the sister group of all remaining taxa in Cymodoceaceae s.s. here (Fig. 2; 4; Table 1) and in the analysis of Petersen et al. (2014); the remaining relationships conflict with those reported by Les and Tippery (2013) considering their best tree (Fig. 6.18b), apart from the sister-group relationship between Cymodocea serrulata and Syringodium filiforme here (Fig. 2, 4). However, all of these conflicts are poorly supported in the Les and Tippery (2013) study.  4.6 Repeated loss of the plastid NADH dehydrogenase complex in Alismatales The plastid gene sets recovered here support the prediction of Iles et al. (2013) that the functionality of the plastid NADH dehydrogenase complex has been lost in parallel in multiple  23  lineages of Alismatales. Iles et al. (2013) predicted two parallel losses in the Cymodoceaceae complex (one in Posidoniaceae, one in a subset of Cymodoceaceae s.s.), which we confirm here (Fig. 3). The loss of the complex in Cymodoceaceae appears to be limited to Amphibolis and Thalassondendron (Fig. 3), both seagrasses. Since all members of the Cymodoceaceae complex are seagrasses (Ruppia is sometimes counted as one, e.g., Les et al., 1997), living submerged in a marine environment does not necessarily lead to ndh pseudogenization. A third loss was predicted in the common ancestor of Najas, Thalassia and Vallisneria in Hydrocharitaceae by Iles et al. (2013), see also Peredo et al. (2013), who confirmed the loss in Najas. My expanded taxon and gene sampling clarify the taxonomic scope of the loss in Hydrocharitaceae (Fig. 3, Table 2): the complex is also missing in Enhalus, Halophila and Nechamandra, consistent with the predictions of Iles et al. (2013). This loss spans six of the seven genera of subfamily Hydrilloideae. It remains to be seen whether the complex is also absent in Hydrilla, which was not sampled here. However, preliminary evidence seems to indicate ndh reading frame disruption in Hydrilla, consistent with other members of the subfamily (unpublished data). Unlike Cymodoceaceae, all Hydrocharitaceae seagrasses show loss of ndh. However, in this family the loss is not limited to marine taxa as Najas, Nechamandra and Vallisneria live in fresh water.  A fourth loss in Alismatales is indicated in one of two species of Triantha (Tofieldiaceae) sampled here, T. occidentalis, representing the first reported loss in unsubmerged members of Alismatales. Of these four losses, the least degraded set of ndh genes is observed in T. occidentalis (which also has four genes with full open reading frames, the highest retention of uninterrupted ndh genes; Table 2). (Iles et al., 2013, noted apparent pseudogenized remnants of ndhF in Vallisneria that were not recovered using the search strategy employed here, so it is possible that  24  I have overestimated the number of full gene losses). Enhalus, Halophila and Nechamandra are the most degraded taxa in terms of lost genes (where no remnants were apparent); in all three taxa, seven genes are deleted (Table 2) based on the search criteria discussed earlier. Only three of these are common across the three species (ndhA, ndhE and ndhI), and it seems likely that all three of these involve recent parallel deletions of these genes, as close relatives (Thalassia for Enhalus and Halophila, and Vallisneria for Nechamandra) have these genes present as pseudogenes or open reading frames (Figs. 2, 3; Table 2). Reversible gains of gene function are biologically unlikely for large multi-gene complex, and this scenario is not inferred here.  Plastid ndh genes code for most of the subunits of the plastid NADH dehydrogenase enzyme complex (11 of the 15 subunits; Martin and Sabater, 2010). Plastid ndh genes have been retained in most land plants, with most exceptions being plants that are partially or fully heterotrophic through parasitism or mycoheterotrophy (dePamphilis and Palmer, 1990; Stefanovic and Olmstead, 2005; Wicke et al., 2011). Our current understanding of the function of plastid NADH dehydrogenase remains unclear, but it is known to be up-regulated at times of photooxidative stress and nutrient deficiency (Martin and Sabater, 2010; Wicke et al., 2011). There are a limited number of losses of plastid NADH deydrogenase reported in fully autotrophic groups outside Alismatales, with reports in Gnetales, Pinaceae, Erodium (Geraniaceae) and some species of Lentibulariaceae (Wakasugi et al., 1994; Braukmann et al., 2009; Blazier et al., 2011; Wicke et al., 2011). Iles et al. (2013) and Peredo et al. (2013) hypothesized that loss of plastid NADH dehydrogenase function is tenable in submerged habitats, if there is reduced photooxidative stresses in subtidal zones (the enzyme complex appears to be involved in this stress response; Wicke et al., 2011). An alternative explanation would be required for the loss of plastid NADH dehydrogenase function in T. occidentalis, as it  25  is an emergent aquatic/helophyte. Arbuscular mycorrhizae are reported in Tofieldiaceae, but mycoheterotrophy is unknown in the order (Wang and Qiu, 2006). Stems of T. occidentalis are covered in glandular hairs, and it is possible that these could be involved in carnivory (see Chase et al., 2009 for other examples of cryptic carnivory in plants with glandular hairs). Carnivory may be associated with ndh loss in Lentibulariaceae (Wicke et al., 2011). However, other species of Triantha also have glandular hairs, including T. glutinosa, which has retained its ndh genes (Fig. 3), and so this speculative hypothesis needs further study.  4.7 Conclusion I achieved my main thesis goal of producing a well-supported, genus-focused, whole plastid-genome based phylogeny of Alismatales. Earlier studies based on a single or few genes recovered many of the same relationships recovered here, but I was able to satisfactorily resolve multiple areas of uncertainty in those studies, with strong support inferred for most higher-order relationships here, including the monophyly of individual families and most aspects of inter- and intrafamilial relationships. These inferences were not sensitive to the method of phylogenetic inference (i.e., maximum likelihood vs. parsimony; partitioning schemes that allowed different DNA subsitution models for different subsets of the data for likelihood inference; inclusion vs. exclusion of ndh pseudogenes in affected taxa). The remaining areas of uncertainty (e.g., the relationships among Araceae, core alismatids, and Tofieldiaceae) may require substantial additional data to resolve, such as additional nuclear genomic data. The improved understanding of phylogenetic relationships recovered here should aid future comparative and evolutionary studies of Alismatales and their relatives, including studies of the often unusual ecological and morphological diversity in the order. I also mapped out a better understanding of the pattern of  26  loss of the plastid NADH dehydrogenase complex, confirmed predictions based on an earlier study using two ndh genes (Iles et al., 2013).  I also demonstrated the first loss of this complex in the order outside the core alismatids, in an unsubmerged species (Triantha occidentalis, Tofieldiaceae). Future studies could investigate the physiological basis of these losses.     27  FIGURES AND TABLES     28  Table 1. Summary of bootstrap support for core alismatid clades found in shortest trees, that have less than 100% bootstrap in at least one analysis (labelled with letters in Fig. 2 and Suppl. Figs. 1-2). Clades with <50% bootstrap support are indicated with a dash (--): “P”, parsimony analysis of the full data set; “P-n”, parsimony analysis with all ndh genes removed from taxa showing pseudogenization or loss (all genes retained for other taxa); “ML-u”, unpartitioned likelihood analysis; “ML-c”, codon partitioning scheme; “ML-GxC”, gene x codon partitioning scheme; “ML-GxC-n”, gene x codon partitioning scheme with all ndh genes removed from taxa showing evidence of pseudogenization or loss, see text and Table S2 for details of partitioning scheme.                      Analysis: P P-n ML-u ML-c ML-GxC ML-GxC-n  Tree score  (steps or -lnL): 226,111 224,469 1 -1,298,968.262 -1,285,402.105 -1,281,368.706 -1,271,860.239 1           Taxon bipartition  a 66 86 81 80 71 84 b 67 66 100 100 100 100 c -- -- 99 100 98 98 d 97 97 100 100 100 100  e 99 99 100 100 100 100 f 98 97 99 100 98 98 g 100 100 95 100 73 73 h 77 77 100 77 52 55 i 67 60 100 100 98 98 j -- 76 79 80 67 85 k 89 91 94 94 96 94 l -- -- -- -- 97 96 m 85 87 95 96 96 96 n 58 55 -- -- -- -- o 70 76 -- -- -- --           1These values are based on a modified data set (ndh genes removed for some taxa); tree scores not comparable with full data set.    29  Table 2. Species in Alismatales with pseudogenization or loss of at least one plastid-encoded ndh subunit gene. ‘O’ = present as an open reading frame; ‘I’ = interrupted reading frame (putative pseudogene); ‘’—‘ = not recovered (even as a detectable remnant).                     Species ndhA ndhB ndhC ndhD ndhE ndhF ndhG ndhH ndhI ndhJ ndhK                  Hydrocharitaceae   Enhalus acoroides -- I I I -- -- -- -- -- I --  Thalassia testudinum I I I I I -- I I O I I Halophila decipiens  -- I -- -- -- I I I -- -- -- Nechamandra alternifolia -- -- I I -- -- -- -- -- I I  Vallisneria americana I I I I I -- I I I I I  Najas guadalupensis -- I I I I -- -- I -- -- --  Cymodoceaceae Thalassodendron pachyrhizum -- O O I -- I -- -- -- I O  Amphibolis griffithii I I I I -- I I I I I I  Posidoniaceae Posidonia ostenfieldii -- I -- I I I I I I -- I   Tofieldiaceae Triantha occidentalis I O I I O I -- I I O O     30  Figure 1. Overall relationships among Araceae, core alismatids, and Tofieldiaceae inferred in (a-b) parsimony analysis, (c-f) likelihood analysis, illustrating which taxon is the sister group of the remainder of the order. Analyses: (a) “P”, parsimony analysis of the full data set; (b) “P-n”, parsimony analysis with all ndh genes removed from taxa showing pseudogenization or loss (all genes retained for other taxa); (c) “ML-u”, unpartitioned likelihood analysis; (d) “ML-c”, codon-based partitioning scheme; (e) “ML-GxC”, a gene x codon partitioning scheme; (f) “ML-GxC-n”, a gene x codon partitioning scheme with all ndh genes removed from taxa showing evidence of pseudogenization or loss (see text and Table S2 for more details of partitioning schemes). Bootstrap support values are noted for branches around the root of Alismatales; the support for this arrangement is noted in a large circle; support values are also noted within Tofieldiaceae (species names included here). Scale bars represent the number of inferred steps (parsimony) or the estimated substitutions per site (likelihood). Full trees are shown in Figs. S1-S6.    31   32  Figure 2. Core alismatid phylogeny inferred in a likelihood analysis of 83 plastid genes using the “GxC-n” partitioning scheme (see text and Table S2 for details). Bootstrap support values are indicated beside branches: thick lines indicate 100% bootstrap support, and letter labels indicate branches with <100% bootstrap support in other analyses (see Table 1). This 12-family clade is a subset of a larger angiosperm-wide phylogeny (Fig. S6, represented as an inset phylogram here; the shaded portion represents the core alismatids). The scale bar indicates estimated substitutions per site in core alismatids.   33    34  Figure 3. Parsimony reconstructions showing predicted losses of the NADH dehydrogenase complex genes in Alismatales, using the likelihood tree from Fig. S6. Black: ndh genes all intact (open reading frames, ORFs); white: one or more ndh genes pseudogenized or lost.    35    36  Figure 4. Branch support in core alismatids compared to other studies. Likelihood bootstrap support values from the current plastid genome data set are indicated above branches using thick lines (100% bootstrap support) or using large font; letter codes correspond to other studies: (a) Les and Tippery (2013; Alismataceae, their Fig. 6.4; Cymodoceaceae, their Fig. 6.18; Hydrocharitaceae, their Fig. 6.10; Juncaginaceae, their Fig. 6.17; Potamogetonaceae and Zosteraceae, their Fig. 6.15); (b) Petersen et al. (2014, Cymodoceaceae, their Fig. 1); (c) Iles et al. (2013; interfamilial relationships, their Fig. 1.1); (d) Les et al. (1997; interfamilial relationships, their Fig. 2). Support values are based on likelihood bootstrap (a-c); parsimony jackknife (b), parsimony bootstrap (a, c, d); when two values are noted for a reference study, the first one is the likelihood result. Arrowhead = weakly conflicting relationship observed in shortest trees (this study vs. the reference study).    37    38  BIBLIOGRAPHY  APG 1998. An ordinal classification for the families of flowering plants: APG. Annals of the Missouri Botanical Garden, 85: 531-553.  APG 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Botanical Journal of the Linnean Society 161:105-121.  Altschup, S.F., Gish, W., Miller, W., Myers, E.W. Lipman, D.J. 1990. Basic Local Alignment Search Tool. Journal of Molecular Biology 215: 403–410.  Azuma, H., Tobe, H., 2011. Molecular phylogenetic analyses of Tofieldiaceae (Alismatales): family circumscription and intergeneric relationships. Journal of Plant Research 124: 349-357.  Backlund, A. and Bremer, K. 1998. To be or not to be - principles of classification and monotypic plant families. Taxon, 47: 391–400.  Barbrook, A. C., Howe, C. J., and Purton, S. 2006. Why are plastid genomes retained in non-photosynthetic organisms? Trends in Plant Science, 11:101–108.   Barrett, C.F., Davis, J.I., Leebens-Mack, J., Conran, J.G., and Stevenson, D.W. 2013. Plastid genomes and deep relationships among the commelinid monocot angiosperms. Cladistics 29: 65–87.  Barrett, C. F., Specht, C. D., Leebens-Mack, J., Stevenson, D. W., Zomlefer, W. B., and Davis, J. I. 2014. Resolving ancient radiations: can complete plastid gene sets elucidate deep relationships among the tropical gingers (Zingiberales)?. Annals of Botany, 113: 119–33.  Blazier, J. C., Guisinger, M. M. & Jansen, R. K. (2011). Recent loss of plastid-encoded ndh genes within Erodium (Geraniaceae). Plant Molecular Biology, 76: 263-272.  Bowe, L.M., dePamphilis, C.W., 1996. Effects of RBA editing and gene processing on phylogenetic reconstruction. Molecular Biology and Evolution, 13, 1159-1166.  Braukmann, T. W. A., Kuzmina, M. and Stefanovic, S. 2009. Loss of all plastid ndh genes in Gnetales and conifers: Extant and evolutionary significance for the seed plant phylogeny. Current Genetics, 55: 323-337.  Cai, Z., Guisinger, M., Kim, H.-G., Ruck, E., Blazier, J.C., McMurtry, V., Kuehl, J.V., Boor, J., and Jansen, R.K. 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.    39  Campbell, L.M. and Dorr, L.J. 2013. A synopsis of Harperocallis Tofieldiaceae, Alismatales with ten new combinations. PhytoKeys, 21: 37-52.  Chang, C.-C., Lin, H.-C., Lin, I.-P., Chow, T.-Y., Chen, H.-H., Chen, W.-H., Cheng, C.-H., Lin, C.-Y., Liu, S.-M., Chang, C.-C., Chaw, S.-M. 2006. The chloroplast genome of Phalaenopsis aphrodite (Orchidaceae): comparative analysis of evolutionary rate with that of grasses and its phylogenetic implications. Molecular Biology and Evolution, 23: 279–91.   Chase, M. W., Soltis, D. E., Soltis, P. E., Rudall, P. J., Fay, M. F., Hahn, W. H., Sullivan, S., Joseph, J., Molvray, M., Kores, P. J., Givnish, T. J., Sytsma, K. J. and Pires, C. 2000. Higher-levels of the monocotyledons: an assessment of current knowledge and a new classification. In: Monocots: Systematics and Evolution. Eds K.L. Wilson and D.A. Morrison, CSIRO, Melbourne, Australia, pp. 3-16.  Chase, M.W., Fay, M.F., Devey, D.S., Maurin, O., Rønsted, N., Davies, T.J., Pillon, Y., Petersen, G., Seberg, O., Tamura, M.N., Asmussen, C.B., Hilu, K., Borsch, T., Davis, J.I., Stevenson, D.W., Pires, J.C., Givnish, T.J., Sytsma, K.J., McPherson, M.A., Graham, S.W., Rai, H.S., 2006. Multigene analyses of monocot relationships: a summary. Aliso, 22: 63-75.  Chen, J.-M., Chen, D., Gituru, W.R., Wang, Q.-F., Guo, Y.-H., 2004. Evolution of apocarpy in Alismatidae using phylogenetic evidence from chloroplast rbcL gene sequence data. Botanical Bulletin of Academia Sinica. 45: 33-40.  Chen, L.-Y., Chen, J.-M., Gituru, R.W., Temam, T.D., and Wang, Q.-F. 2012a. Generic phylogeny and historical biogeography of Alismataceae, inferred from multiple DNA sequences. Molecular Phylogenetics and Evolution, 63:407–416.  Chen, L.-Y., Chen, J.-M., Gituru, R. W., & Wang, Q.-F. 2012b. Generic phylogeny, historical biogeography and character evolution of the cosmopolitan aquatic plant family Hydrocharitaceae. BMC Evolutionary Biology, 12: 30.  Cuenca, A., Petersen, G., Seberg, O., and Jahren, A.H. 2012. Genes and processed paralogs co-exist in plant mitochondria. Journal of Molecular Evolution, 74: 158–69.   Cronn, R., Liston, A., Parks, M., Gernandt, D.S., Shen, R., and Mockler, T. 2008 Multiplex sequencing of plant chloroplast genomes using Solexa sequencing-by-synthesis technology. Nucleic Acids Research, 36: e122.  Davis, J.I., Stevenson, D.W., Petersen, G., Seberg, O., Campbell, L.M., Freudenstein, J.V.,  Goldman, D.H., Hardy, C.R., Michelangeli, F.A., Simmons, M.P., Specht, C.D., Vergara-Silva, F., Gandolfo, M. 2004. A phylogeny of the monocots, as inferred from rbcL and atpA sequence variation, and a comparison of methods for calculating jackknife and bootstrap values. Systematic Botany, 29: 467-510.   40  dePamphilis, C. W. & Palmer, J. D. 1990. Loss of photosynthetic and chlororespiratory genes from the plastid genome of a parasitic flowering plant. Nature, 384: 337-339.  Doyle, J. J. and Doyle, J. L. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin, 19: 11-15.  Duvall, M.R., Mathews, S., Mohammad, N., Russell, T., 2006. Placing the monocots: conflicting signal from trigenomic analyses. Aliso, 22: 79-90.  Duvall, M.R., Robinson, J.W., Mattson, J.G., Moore, A., 2008. Phylogenetic analyses of two mitochondrial metabolic genes sampled in parallel from angiosperms find fundamental interlocus incongruence. American Journal of Botany, 95: 871-884.  Edgar, R.C. 2004 MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research, 32: 1792-1797.  Edwards A.W.F. 1972. Likelihood. Cambridge University Press, Cambridge.  Felsenstein, J. 1978. Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zoology, 27: 401–410.  Felsenstein, J. 1983. Parsimony in systematics: Biological and statistical issues. Annual Review of Ecology and Systematics, 14: 313–333.  Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 394: 783-791.  Fourqurean, J. W., Duarte, C. M., Kennedy, H., Marbà, N., Holmer, M., Mateo, M. A., Apostolaki, E. T., Kendrick, G.A., Krause-Jensen, D., McGlathery, K.J., Serrano, O. 2012. Seagrass ecosystems as a globally significant carbon stock. Nature Geoscience, 5(6): 1-5.  Freyer, R., Kiefer-Meyer, M.-C., Kössel, H., 1997. Occurrence of plastid RNA editing in all major lineages of land plants. Proceeding of the National Academy of Science, 94: 6285-6290.  Givnish, T. J., Pires, J. C., Graham, S. W., McPherson, M. A., Prince, L. M., Patterson, T. B., Rai, H. S., Hoalston, E. H., Evans, T. M., Hahn, W. J., Millam, K. C., Meerow, A. W., Molvray, M., Kores, P. J., O'Brien, H. E., Hall, J. C., Kress, W. J. and Sytsma, K. J. 2006. Phylogenetic relationships of monocots based on the highly informative plastid gene ndhF: evidence for widespread concerted convergence Aliso 22: 28-51.  Graham, S.W., Reeves, P.A., Burns, A.C.E., and Olmstead, R.G. 2000. Microstructural changes in noncoding chloroplast DNA: Interpretation, evolution, and utility of indels and inversions in basal angiosperm phylogenetic inference. International Journal of Plant Science, 161: S83–S96.    41  Graham, S.W., Zgurski, J. M., McPherson, M. A., Cherniawsky, D. M., Saarela, J. M., Horne, E. S. C., Smith, S. Y., Wong, W. A., O'Brien, H. E., Biron, V. L., Pires, J. C., Olmstead, R. G., Chase, M. W. and Rai, H. S. 2006. Robust inference of monocot deep phylogeny using an expanded multigene plastid data set Aliso 22: 3-21.  Guisinger, M.M., Chumley, T.W., Kuehl, J.V, Boore, J.L., and Jansen, R K. 2010. Implications of the plastid genome sequence of Typha (Typhaceae, Poales) for understanding genome evolution in Poaceae. Journal of Molecular Evolution, 70: 149–66.   Haynes, R.R., Les, D.H. and Holm-Nielsen, L.B. 1998. Zannichelliaceae. In The Families And Genera Of Vascular Plants, Vol. IV, Flowering Plants. Monocotyledons: Alismatanae And Commelinanae (Except Gramineae), ed. K. Kubitzki. Berlin: Springer-Verlag, pp. 470-474.  Hendy, M.D., and Penny, D. 1989. A framework for the quantitative study of evolutionary trees. Systematic Zoology, 38: 297–309.  Henriquez, C.L., Arias, T., Pires, J.C., Croat, T.B., and Schaal, B.A. 2014. Phylogenomics of the plant family Araceae. Molecular Phylogenetics and Evolution, 75: 91–102.   Iles, W.J.D., Smith, S.Y., Graham, S.W. 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: Early Events in Monocot Evolution, Eds. P. Wilkin and S.J. Mayo. Cambridge University Press Systematics Association Special Series United Kingdom.   Inada, M., Sasaki, T., Yukawa, M. Tsudzuki, T., Sugiura, M., 2004. A systematic search for RNA editing sites in pea chloroplasts: an editing event causes diversification from the evolutionarily conserved amino acid sequence. Plant Cell Physiology, 45: 1615-1622.  Ito, Y., Ohi-Toma, T., Murata, J., & Tanaka, N. (2010). Hybridization and polyploidy of an aquatic plant, Ruppia (Ruppiaceae), inferred from plastid and nuclear DNA phylogenies. American Journal of Botany, 97: 1156–67.   Jacobson, A., and Hedrén, M. 2007. Phylogenetic relationships in Alisma (Alismataceae) based on RAPDs, and sequence data from ITS and trnL. Plant Systematics and Evolution, 265, 27–44.  Jansen, R.K., Saski, C., Lee, S.-B., Hansen, A.K., and Daniell, H. 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, S. 2013 MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution, 30: 772-780.   42  Krause, K. 2008. From chloroplasts to “cryptic” plastids: evolution of plastid genomes in parasitic plants. Current Genetics, 54: 111–121.  Lanfear, R., Calcott, B., Kainer, D., Mayer, C., and Stamatakis, A. 2014. Selecting optimal partitioning schemes for phylogenomic datasets. BMC Evolutionary Biology, 141: 82.   Lehtonen, S. 2006. Phylogenetics of Echinodorus (Alismataceae) based on morphological data. Botanical Journal of the Linnean Society, 150: 291-305.  Lehtonen, S., and Myllys, L. 2008. Cladistic analysis of Echinodorus (Alismataceae): simultaneous analysis of molecular and morphological data. Cladistics, 24, 218–239.  Les, D.H., Garvin, D.K., and Wimpee, C.F. 1993. Phylogenetic studies in the monocot subclass Alismatidae: evidence for a reappraisal of the aquatic order Najadales. Molecular Phylogenetics and Evolution, 2: 304-314.  Les, D. H., Cleland, M. A., & Waycott, M. 1997. Phylogenetic studies in Alismatidae, II : evolution of marine angiosperms seagrasses and hydrophily, Systematic Botany, 22: 443-463.  Les, D.H., Moody, M.L., and Jacobs, S.W.L. 2005. Phylogeny and systematics of Aponogeton (Aponogetonaceae): the Australian species. Systematic Botany, 30: 503–519.   Les, D.H., Moody, M.L., and Soros, C.L. 2006. A reappraisal of the phylogenetic relationship in the monocotyledon family Hydrocharitaceae (Alismatidae). Aliso, 22: 211-230.  Les, D.H., Sheldon, S.P., and Tippery, N.P. 2010. Hybridization in hydrophiles: natural interspecific hybrids in Najas (Hydrocharitaceae). Systematic Botany, 35: 736–744.  Les, D.H., and Tippery, N.P. 2013. Evolution of alismatid aquatics. In: Early Events in Monocot Evolution, Eds. P. Wilkin and S.J. Mayo. Cambridge University Press Systematics Association Special Series United Kingdom.   Lewis L.A., Lewis, P.O. 2005 Unearthing the molecular phylodiversity of desert soil green algae (Chlorophyta). Systematic Botany 54: 936–947.  Li, X., Zhou, Z., 2007. The higher-level phylogeny of monocots based on matK, rbcL and 18S rDNA sequences. Journal of Systematics and Evolution, 45: 113-133.  Li, X., and Zhou, Z. 2009. Phylogenetic studies of the core Alismatales inferred from morphology and rbcL sequences. Progress in Natural Science, 19: 931–945.  Lindqvist, C., De Laet, J., Haynes, R.R., Aagesen, L., Keener, B.R., & Albert, V.A. 2006. Molecular phylogenetics of an aquatic plant lineage, Potamogetonaceae. Cladistics, 22: 568–588.    43  Maddison W.P., Maddison D.R. 2007. Mesquite: a modular system for evolutionary analysis. Version 2.0, http://mesquiteproject.org.  Maia, V.H., Gitzendanner, M.A., Soltis, P.S., Wong, G.K.-S., and Soltis, D.E. 2014. Angiosperm phylogeny based on 18S/26S rDNA sequence data: constructing a large data set using next-generation sequence data. International Journal of Plant Sciences, 175: 613–650.   Martin, M., and Sabater, B. 2010. Plastid ndh genes in plant evolution. Plant Physiology and Biochemistry, 48: 636-645.  Merckx, V.S.F.T. 2013. Mycoheterotrophy: An Introduction. In: Mycoheterotrophy, Eds. Merckx V.S.F.T.. Springer, New York, pp 1–17.  Moore, M.J., Soltis, P.S., Bell, C.D., Burleigh, J.G., and Soltis, D.E. 2010. Phylogenetic analysis of 83 plastid genes further resolves the early diversification of eudicots. Proceedings of the National Academy of Sciences, 107: 4623–4628.  Pagel, M. 1999. The maximum likelihood approach to reconstructing ancestral character states of discrete characters on phylogenies. Systematic Botany 48:612.  Parks, M., Cronn, R., and Liston, A. 2009. Increasing phylogenetic resolution at low taxonomic levels using massively parallel sequencing of chloroplast genomes. BMC Biology, 7: 84.  Peredo, E.L., King, U.M., and Les, D.H. 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. Public Library of Science ONE, 8: e68591.   Petersen, G., Seberg, O., Davis, J. I and Stevenson, D.W. 2006a. RNA editing and phylogenetic reconstruction in two monocot mitochondrial genes. Taxon, 55, 871-886.  Petersen, G., Seberg, O., Davis, J.I., Goldman, D.H., Stevenson, D.W., Campbell, L.M., Michelangeli, F.A., Specht, C.D., Chase, M.W., Fay, M.F., Pires, J.C., Freudenstein, J.V., Hardy, C.R., Simmons, M.P., 2006b. Mitochondrial data in monocot phylogenetics. Aliso 22: 52–62.  Petersen, G., Seberg, O., Davis, J.I., 2013. Phylogeny of the Liliales (Monocotyledons) with special emphasis on data partition congruence and RNA editing. Cladistics 29: 274-295.  Petersen, G., Seberg, O., Short, F.T, and Fortes, M.D. 2014. Complete genomic congruence but non-monophyly of Cymodocea (Cymodoceaceae), a small group of seagrasses. Taxon, 63: 3–8.  Qiu, Y.-L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Solitis, P.S., Zanis, M., Zimmer, E.A., Chen, Z., Savolainen, V., Chase, M.W., 2000. Phylogeny of basal angiosperms: analyses of five genes from three genomes. International Journal of plant Science 161: S3-S27.   44  Qiu, Y-L., Li, L., Hendry, T.A., Li, R., Taylor, D.W., Issa, M.J., Ronen, A.J., Vekaria, M.L. and White, A.M. 2006. Reconstructing the basal angiosperm phylogeny: evaluating information content of mitochondrial genes. Taxon, 55: 837–856.  Qiu, Y.-L., Li, L., Wang, B., Xue, J.-Y., Hendry, T.A., Li, R.-Q., Brown, J.W., Liu, Y., Hudson, G.T., Chen, Z.-D. 2010. Angiosperm phylogeny inferred from sequences of four mitochondrial genes. Journal of Systematics and Evolution, 48: 391–425.   Rambaut, A., 2002. Se-Al Sequence Alignment Editor, Ver. 2.0a11. University of Oxford, Oxford.  Raubeson L.A. and Jansen R.K. 2005. Chloroplast genomes in plants. In: Plant Diversity and Evolution: Genotypic and Phenotypic Variation in Higher Plants, Eds Henry R.J., Cambridge, MA, pp 45–68.  Remizowa, M.V., Sokoloff, D.D., Campbell, L.M., Stevenson, D.W., and Rudall, P.J. 2011. Harperocallis is congeneric with Isidrogalvia Tofieldiaceae, Alismatales: Evidence from comparative floral morphology. Taxon, 60: 1076-1094.  Ruhfel, B.R., Gitzendanner, M.A, Soltis, P.S., Soltis, D.E., and Burleigh, J.G. 2014. From algae to angiosperms-inferring the phylogeny of green plants (Viridiplantae) from 360 plastid genomes. BMC Evolutionary Biology, 14: 23.   Saarela, J.M. and Graham, S.W. 2010 Inference of phylogenetic relationships among the subfamilies of grasses Poaceae: Poales using meso-scale exemplar-based sampling of the plastid genome. Botany, 881: 65-84.    Sculthorpe, C.D. 1967. The Biology of Aquatic Vascular Plants. Ed. Arnold, E. London, England.  Sokoloff, D., von Mering, S., Jacobs, S.W.L., and Remizowa, M.V. 2013. Morphology of Maundia supports its isolated phylogenetic position in the early‐divergent monocot order Alismatales. Botanical Journal of the Linnean Society, 173: 12–45.   Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach, D.C., Zanis, M., Savolainen, V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensem, S.M., Prince, L.M., Kress, W.J., Nixon, K.C. and Farris, J.S. 2000, Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381–461.  Soltis, D.E., Gitzendanner, M.A., and Soltis, P.S. 2007. A 567‐taxon data set for angiosperms: the challenges posed by Bayesian analyses of large data sets. International Journal of Plant Sciences 168: 137–157.   Stamatakis, A. 2006a RAxML-VI-HPC: Maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22:2688–2690.   45  Stamatakis, A. 2006b. Phylogenetic models of rate heterogeneity: a high performance computing perspective. Proceedings of 20th IEEE⁄ACM International Parallel and Distributed Processing Symposium IPDPS2006, High Performance Computational Biology Workshop, Rhodos, Greece, April 2006.  Steele, P. R., Hertweck, K.L., Mayfield, D., McKain, M.R., Leebens-Mack, J., and Pires, J.C. 2012. Quality and quantity of data recovered from massively parallel sequencing: Examples in Asparagales and Poaceae. American Journal of Botany, 99: 330–48.  Stefanovic, S. and Olmstead, R.G. 2005. Down the slippery slope: plastid genome evolution in Convolvulaceae. Journal of Molecular Evolution, 61: 292-305.  Stevens, P.F. (2001 onwards). Angiosperm Phylogeny Website. Version 12, July 2012 [and more or less continuously updated since]. http://www.mobot.org/mobot/research/apweb/  Stevenson, D.W., Davis, J.I., Freudenstein, J.V., Hardy, C.R., Simmons, M.P., Specht, C.D. 2000. A Phylogenetic Analysis of the Monocotyledons Based on Morphological and Molecular Character Sets, With Comments on the Placement of Acorus and Hydatellaceae. In: Monocots: Systematics and Evolution. Eds K.L. Wilson and D.A. Morrison, CSIRO, Melbourne, Australia, pp. 17-24.  Swofford, D. L. 2003. PAUP*. Phylogenetic Analysis Using Parsimony *and Other Methods. Version 4. Sinauer Associates, Sunderland, Massachusetts.  Talavera, G., and Castresana, J. 2007. Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Systematic Biology, 56: 564-577.    Tamura, M.N. 1998. Nartheciaceae. In The Families And Genera Of Vascular Plants, Vol. III, Flowering Plants, Monocotyledons: Lilianae (Except Orchidaceae), ed. K. Kubitzki. Berlin: Springer-Verlag, pp. 381-392.  Tamura, M.N., Fuse, S., Azuma, H., and Hasebe, M. 2004. Biosystematic studies on the family Tofieldiaceae I. Phylogeny and circumscription of the family inferred from DNA sequences of matK and rbcL. Plant Biology: 6(5), 562–567.  Tanaka, N., Setoguchi, H., and Murata, J. 1997. Phylogeny of the family hydrocharitaceae inferred from rbcL and matK gene sequence data. Journal of Plant Research, 110: 329–337.   Thiers, B. [continuously updated]. Index Herbariorum: A global directory of public herbaria and associated staff. New York Botanical Garden's Virtual Herbarium. http://sweetgum.nybg.org/ih/  von Mering, S., and Kadereit, J.W., 2010. Phylogeny, systematics, and recircumscription of Juncaginaceae—a cosmopolitan wetland family. In Diversity, Phylogeny, and Evolution in  46  the Monocotyledons, Eds. O. Seberg, G. Petersen, A. S. Barfod & J. I Davis. Aarhus, Denmark: Aarhus University Press, pp. 55-79.  Wakasugi, T., Tsudzuki, J., Ito, S., Nakashima, K., Tsudzuki, T., and Sugiura, M. 1994. Loss of all ndh genes as determined by sequencing the entire chloroplast genome of the black pine Pinus thunbergii. Proceedings of the National Academy of Sciences, 91: 9794–9798.   Wang, B. and Qiu, Y.-L. 2006. Phylogenetic distribution and evolution of mycorrhizas in land plants. Mycorrhiza, 16: 299-363.  Waycott, M., Procaccini, G., Les, D. H. and Reusch, T. B. H. 2006. Seagrass evolution, ecology and conservation: a genetic perspective. In Seagrasses: Biology, Ecology and Conservation, ed. A.W.D. Larkum, R.J. Orth and C.M. Duarte. Dordrecht, Netherlands: Springer, pp. 25–50.  Waycott, M., Duarte, C.M., Carruthers, T.J.B., Orth, R.J., Dennison, W.C., Olyarnik, S., Calladine, A., Fourqurean, J.W., Heck Jr, K.L.., Hughes, A.R., Kendrick, G.A., Kenworthy, W.J., Short, F.T., Williams, S.L. 2009. Acceleration loss of seagrasses across the globe threatens coastal ecosystems. Proceedings of the National Academy of Sciences, 106:12377-12381.  Wicke, S., Schneeweiss, G.M., dePamphilis, C.W., Müller, K.F., and Quandt, D. 2011. The evolution of the plastid chromosome in land plants: gene content, gene order, gene function. Plant Molecular Biology, 76: 273–97.   Wyman, S.K., Jansen, R.K., Boore, J.L. 2004 Automatic annotation of organellar genomes with DOGMA. Bioinformatics, 20: 3252-3255.       47  APPENDICES Supplementary Table S1 - Specimen source information; herbarium abbreviations follow Thiers (continuously updated).        Species1 Family Specimen voucher number     [Collector number (herbarium)]        Alismatales Alisma triviale Pursh Alismataceae Barrett s.n. (TRT) Astonia australiensis (Aston) S.W.L.Jacobs Alismataceae Jacobs 9670 (NSW)  Baldellia ranunculoides (L.) Parl. Alismataceae Charlton s.n. (MANCH) Butomopsis latifolia (D.Don) Kunth Alismataceae Jacobs 9257 (NSW) Caldesia oligococca (F.Muell.) Buchanan Alismataceae F. Rasmussen et al., C246 (C)  Damasonium minus (R.Br.) Buchenau Alismataceae Smith & Hopper RJS281 (PERTH) Echinodorus amazonicus Rataj Alismataceae Les s.n. (CONN) Helanthium bolivianum Alismataceae G. Petersen & O. Seberg C2585 (C)  (Rusby) Lehtonen & Myllys   Hydrocleys nymphoides Alismataceae Chase 16878 (K)  (Humb. & Bonpl. ex Willd.) Buchenau  Limnocharis flava (L.) Buchenau Alismataceae Les s.n. (CONN) Luronium natas (L.) Raf. Alismataceae Chase 14237 (K) Ranalisma humile (Rich. ex Kunth) Hutch. Alismataceae Charlton s.n. (MANCH) Sagittaria latifolia Willd. Alismataceae D.W. Stevens 178/2010 (NY) Wiesneria triandra (Dalzell) Micheli Alismataceae C.D.K. Cook s.n. (Z) Aponogeton distachyos L.f. Aponogetonaceae R.A. Stockey & G.W. Rothwell, (living     collection; Bot. Garten München-Nymphenburg) Butomus umbellatus L. Butomaceae Chase 6414 (K) Amphibolis griffithii (J.M.Black) Hartog Cymodoceaceae J.G. Conran 3085 (AD, ADU) Cymodocea serrulata (R.Br.) Asch. & Magnus Cymodoceaceae O'Donohue 21395 (BRN) Cymodocea nodosa (Ucria) Asch. Cymodoceaceae Procaccini s.n. (CONN) Halodule wrightii Asch. Cymodoceaceae D.A. Kolterman & I. López 1003 (ALTA)   48         Species1 Family Specimen voucher number     [Collector number (herbarium)]        Syringodium filiforme Kütz. Cymodoceaceae Wimpee s.n. (CONN) Thalassodendron pachyrhizum Hartog Cymodoceaceae Waycott 95002 (UWA) Apalanthe granatensis (Humb. & Bonpl.) Planch. Hydrocharitaceae C.D.K. Cook s.n. (Z) Blyxa aubertii Rich. Hydrocharitaceae Charlton s.n. (MANCH) Egeria najas Planch. Hydrocharitaceae G. Petersen & O. Seberg C2594 (C)    Enhalus acoroides (L.f.) Royle Hydrocharitaceae D.T. Dy s.n. (UBC) Halophila decipiens Ostenf. Hydrocharitaceae D. A. Kolterman & I. López 1004 (ALTA) Hydrocharis morsus-ranae L. Hydrocharitaceae S.Y. Smith 51 (ALTA) Lagarosiphon major (Ridl.) Moss Hydrocharitaceae C.D.K. Cook s.n. (Z) Limnobium laevigatum Hydrocharitaceae G. Petersen & O. Seberg C2589 (C)  (Humb. & Bonpl. ex Willd.) Heine   Najas guadalupensis (Spreng.) Magnus Hydrocharitaceae D.W. Stevens 42/2011 (NY) Nechamandra alternifolia  Hydrocharitaceae C.D.K. Cook s.n. (Z)  (Roxb. ex Wight) Thwaites Ottelia ovalifolia (R.Br.) Rich. Hydrocharitaceae J. Bogner s.n. (M) Stratiotes aloides Hydrocharitaceae J. Bogner s.n. (ALTA) Thalassia testudinum Hydrocharitaceae D. A. Kolterman & I. López 1001 (ALTA)  Banks & Sol. ex K.D.Koenig  Vallisneria americana Michx. Hydrocharitaceae D.W. Stevens 39/2011 (NY) Cycnogeton procerum (R.Br.) Buchenau Juncaginaceae J.G. Conran 3082A (AD, ADU) Tetroncium magellanicum Willd. Juncaginaceae Alvarez s.n. (CONN) Triglochin maritima L. Juncaginaceae G. Petersen & O. Seberg C1440 (C) Maundia triglochinoides F.Muell. Maundiaceae L. Stanberg & G. Sainty LS 80 (NSW) Posidonia ostenfeldii Hartog Posidoniaceae J.G. Conran 3101 (AD, ADU) Groenlandia densa (L.) Fourr. Potamogetonaceae J. Bogner s. n. (ALTA) Lepilaena australis J.Drumm. ex Harv. Potamogetonaceae D.W. Stevens 41/2011 (NY) Potamogeton richardsonii (A.Benn.) Rydb. Potamogetonaceae S.Y. Smith 50 (ALTA)  49         Species1 Family Specimen voucher number     [Collector number (herbarium)]        Stuckenia pectinata (L.) Börner Potamogetonaceae J.G. Conran 3103 (AD, ADU) Zannichellia palustris L. Potamogetonaceae D.W. Stevens 40/2011 (NY) Ruppia maritime L. Ruppiaceae J.G. Conran 3086 (AD, ADU) Scheuchzeria palustris L. Scheuchzeriaceae D.W. Stevens 23/2012 (NY) Pleea tenuifolia Michx. Tofieldiaceae W. Zomlefer 798 (GA) Tofieldia coccinea Richardson Tofieldiaceae M.J. Waterway 2006-241 (UBC) Triantha glutinosa (Michx.) Baker Tofieldiaceae Stefanovic s.n. (UBC) Triantha occidentalis (S.Watson) R.R.Gates Tofieldiaceae Manton 300 (UBC) Phyllospadix scouleri Hook. Zosteraceae A.O. Shelton 30.viii.2007 (no voucher)     Location: Shi Shi Beach, WA, USA Zostera muelleri Irmisch ex Asch. Zosteraceae J.G. Conran 3100 (AD, ADU)  Dioscoreales Burmannia bicolor Mart. Burmanniaceae Maas 9649 (U) Lophiola aurea Ker Gawl. Nartheciaceae Whitten 95028 (K)  Liliales Campynema lineare Labill. Campynemataceae M.F. Duretto 1842 (HO)  Pandanales Cyclanthus bipartitus Poit. ex A.Rich Cyclanthaceae M.W. Chase 1237 (K) Freycinetia banksii A. Cunn. Pandanaceae S.W. Graham 02-03-14 (UBC) Stichoneuron caudatum Ridl. Stemonaceae Rothwell & Stockey 45 (ALTA)  Xerophyta retinervis Baker Velloziaceae B.G. Reeves 14 (NBG)    50  1Other sequences: Acorus calamus L. (NC_007407); Amborella trichopoda Baill. (NC_005086); Buxus microphylla Siebold & Zucc. (NC_009599); Drimys granadensis L.f. (NC_008456); Illicium oligandrum Merr. & Chun (NC_009600); Lemna minor L. (NC_010109); Elodea canadensis Michx. (NC_018541); Liriodendron tulipifera L. (NC_008326); Nandina domestica Thunb. (NC_008336); Nuphar advena (Aiton) W.T.Aiton (NC_008788); Piper cenocladum C.DC. (NC_008457); Plantanus occidentalis L. (NC_008335); Spirodela polyrhiza (L.) Schleid. (NC_015891); Veratrum patulum Loes. (NC_022715); Vitis vinifera L. (NC_007957); Wolffia australiana (Benth.) Hartog & Plas (NC_015899); Wolffiella lingulata (Hegelm.) Hegelm. (NC_015894). See Barrett et al. (2013) for details for additional monocots, and Henriquez et al. (2014) for Araceae, specifically.  51  Supplementary Table S2 - DNA substitutions models and partitioning scheme and partition subsets resulting from two different partition analyses conducted in PartitionFinder using the BIC criterion. (a) Matrix partitioned using the codon partition scheme; (b) matrix partitioned using the ‘GxC’ scheme (see text). Genes are indicated before the underscore; the ‘pos’ term after the underscore indicates the codon position (‘all’ includes pooled data for the relevant gene or genes).   Partition no. Best Model Partition subsets              a) 1 GTR+I+Γ pos_1 2 GTR+I+Γ pos_2 3 TVM+I+ Γ pos_3 4 TIM+I+ Γ rDNA_all 5 TVM+I+ Γ ycf1_all   b) 1 GTR+I+Γ accD_pos1, atpI_pos3 2 GTR+I+Γ accD_pos2, cemA_pos3, psbF_pos3, ycf4_pos3 3 GTR+I+Γ accD_pos3, rps15_pos3 4 GTR+I+Γ atpA_pos1, atpI_pos2, ndhG_pos2, ndhK_pos1,     psbC_pos2, psbF_pos2, psbH_pos1, psbN_pos2,     rpoA_pos2, rpoC1_pos1, rps11_pos1, rps8_pos2,     ycf3_pos2, ycf4_pos1 5 GTR+I+Γ atpA_pos2, ndhA_pos2 6 GTR+I+Γ atpA_pos3 7 GTR+I+Γ atpB_pos1, atpE_pos1, ndhC_pos1, ndhD_pos2,     petA_pos1, psaI_pos1, psaJ_pos2, psbE_pos2, psbL_pos3,   rpl23_pos3, rpl2_pos3, rps14_pos1, rps14_pos2 8 GTR+I+Γ atpB_pos2, ndhD_pos1, ndhG_pos1, psbN_pos3,     rps8_pos1 9 GTR+I+Γ atpB_pos3 10 GTR+I+Γ atpE_pos2, atpF_pos1, atpF_pos2, cemA_pos1,     infA_pos1, ndhI_pos1, petL_pos1, psaC_pos1,     psbA_pos2, psbL_pos2, psbT_pos1, rpl33_pos2,     rpl36_pos1, rpoB_pos1, rpoC1_pos2, rps4_pos1,     ycf2_pos1, ycf2_pos2 11 GTR+I+Γ atpE_pos3, psaI_pos3, rpl14_pos3 12 GTR+I+Γ atpF_pos3, ccsA_pos1, ndhF_pos1, psbK_pos3,     rbcL_pos2, rpl33_pos3, rpl36_pos3    52  Partition no. Best Model Partition subsets             13 GTR+I+Γ atpH_pos1, ndhB_pos2, ndhJ_pos1, psaC_pos2,     psbM_pos1, psbN_pos1, rpl14_pos2, rpl23_pos2,     rpl2_pos1, rps12_3end_pos3, rps12_pos1, rps12_pos2 14 GTR+I+Γ atpH_pos2, petG_pos2, psbF_pos1 15 GTR+Γ atpH_pos3, rpl22_pos1 16 GTR+I+Γ atpI_pos1, ndhK_pos2, petB_pos2, psaA_pos1, psbJ_pos3 17 GTR+I+Γ ccsA_pos2 18 GTR+I+Γ ccsA_pos3 19 GTR+Γ cemA_pos2, ndhE_pos1 20 GTR+I+Γ clpP_pos1, matK_pos2, rpl22_pos2, rps11_pos2 21 GTR+I+Γ clpP_pos2, lhbA_pos3, ndhA_pos1, ndhI_pos2,     petD_pos1, psbJ_pos1, psbK_pos1, rps15_pos1,     rps16_pos1 22 GTR+I+Γ clpP_pos3, petA_pos3, psaA_pos3  23 GTR+I+Γ infA_pos2, ndhH_pos2, psaA_pos2, rps19_pos1,     rps2_pos1, rps2_pos2 24 GTR+I+Γ infA_pos3, rps3_pos3 25 GTR+I+Γ lhbA_pos1, ndhH_pos1, psbB_pos1, rps4_pos2 26 GTR+I+Γ lhbA_pos2, ndhB_pos1, ndhB_pos3, psbM_pos2,     rps12_3end_pos2, rps7_pos2, ycf3_pos1 27 GTR+I+Γ matK_pos1, rpoC1_pos3 28 GTR+I+Γ matK_pos3 29 GTR+I+Γ ndhA_pos3 30 GTR+I+Γ ndhC_pos2, ndhE_pos2, rpl23_pos1 31 GTR+Γ ndhC_pos3, rps8_pos3 32 GTR+I+Γ ndhD_pos3, rpl32_pos3, rps16_pos3 33 GTR+I+Γ ndhE_pos3, ndhH_pos3, psaC_pos3 34 GTR+I+Γ ndhF_pos2, psbM_pos3, rpoA_pos3, rps2_pos3 35 GTR+I+Γ ndhF_pos3 36 GTR+I+Γ ndhG_pos3 37 GTR+Γ ndhI_pos3 38 GTR+I+Γ ndhJ_pos2, psbJ_pos2, rpoB_pos2, rps18_pos2 39 GTR+Γ ndhJ_pos3 40 GTR+I+Γ ndhK_pos3 41 GTR+I+Γ petA_pos2 42 GTR+I+Γ petB_pos1 43 GTR+I+Γ petB_pos3 44 GTR+I+Γ petD_pos2 45 GTR+Γ petD_pos3 46 GTR+Γ petG_pos1 47 GTR+I+Γ petG_pos3, psbD_pos2 48 GTR+I+Γ petL_pos2   53  Partition no. Best Model Partition subsets             49 GTR+I+Γ petL_pos3, rps12_pos3, rps18_pos1, rps3_pos1,     rps3_pos2 50 GTR+Γ petN_pos1 51 GTR+Γ petN_pos2 52 GTR+Γ petN_pos3, rpl32_pos1, rps16_pos2 53 GTR+I+Γ psaB_pos1 54 GTR+I+Γ psaB_pos2 55 GTR+I+Γ psaB_pos3, psbC_pos3, psbD_pos3, rpl32_pos2 56 GTR+I+Γ psaI_pos2, rpoC2_pos2 57 GTR+Γ psaJ_pos1 58 GTR+Γ psaJ_pos3 59 GTR+I+Γ psbA_pos1, rpl20_pos1, rpl33_pos1, rps15_pos2 60 GTR+I+Γ psbA_pos3, rpl16_pos3 61 GTR+I+Γ psbB_pos2 62 GTR+I+Γ psbB_pos2 63 GTR+I+Γ psbC_pos1 64 GTR+I+Γ psbD_pos1, rpl2_pos2 65 GTR+I+Γ psbE_pos1 66 GTR+I+Γ psbE_pos3 67 GTR+I+Γ psbH_pos2, rbcL_pos1 68 GTR+I+Γ psbH_pos3, rpl20_pos3 69 GTR+Γ psbI_pos1, rps12_3end_pos1 70 GTR+Γ psbI_pos2 71 GTR+I+Γ psbI_pos3 72 GTR+I+Γ psbK_pos2 73 GTR+I+Γ psbL_pos1 74 GTR+I+Γ psbL_pos1 75 GTR+Γ psbT_pos3 76 GTR+I+Γ rbcL_pos3 77 GTR+I+Γ rpl14_pos1 78 GTR+I+Γ rpl16_pos1, rpoC2_pos1 79 GTR+I+Γ rpl16_pos2, rpl20_pos2, rps14_pos3 80 GTR+Γ rpl22_pos3 81 GTR+Γ rpl36_pos2 82 GTR+I+Γ rpoA_pos1 83 GTR+I+Γ rpoB_pos3 84 GTR+I+Γ rpoC2_pos3 85 GTR+Γ rps4_pos3 86 GTR+Γ rps7_pos1, ycf4_pos2 87 GTR+Γ rps7_pos3 88 GTR+Γ rps11_pos3 89 GTR+Γ rps18_pos3 90 GTR+Γ rps19_pos2  54  Partition no. Best Model Partition subsets             91 GTR+Γ rps19_pos3 92 GTR+Γ ycf2_pos3 93 GTR+Γ ycf3_pos3   55  Supplementary Figure S1. Phylogram of angiosperm phylogeny inferred from a parsimony analysis of coding regions from whole plastid genomes (83 genes total), one of four shortest trees (length = 226,111 steps; CI = 0.327; RI = 0.688). Branches that collapse in the strict consensus are indicated with arrows. Bootstrap support values are indicated beside branches where less than 100% (‘--’ indicates <50% bootstrap support). The scale bar indicates the inferred number of changes. The two branches with letter labels (n, o) were recovered here but not in Fig. 2 (and Fig. S6, see Table 1).   56   57  Supplementary Figure S2. Phylogram of angiosperm phylogeny inferred from a parsimony analysis of coding regions from whole plastid genomes (83 genes total), in which ndh genes were entered as blanks for those species in which one or more of these was pseudogenized or lost. This is one of two shortest trees (length = 224,469 steps; CI = 0.326; RI = 0.688). Branches that collapse in the strict consensus are indicated with arrows. Bootstrap support values are indicated beside branches where less than 100% (‘--’ indicates <50% bootstrap support). The scale bar indicates the inferred number of changes. The two branches with letter labels (n, o) were recovered here but not in Fig. 2 (and Fig. S6, see Table 1).   58   59  Supplementary Figure S3. Phylogram of angiosperm phylogeny inferred from an unpartitioned likelihood analysis of coding regions from whole plastid genomes (83 genes total; -lnL = 1,298,968.262). Bootstrap support values are indicated beside branches where less than 100%  (‘--’ indicates <50% bootstrap support). The scale bar indicates estimated substitutions per site.   60   61  Supplementary Figure S4. Phylogram of angiosperm phylogeny inferred from a codon-partitioned likelihood analysis of coding regions from whole plastid genomes (83 genes total; -lnL = 1,285,402.105); see text and Table S2 for partitioning scheme. Bootstrap support values are indicated beside branches where less than 100% (‘--’ indicates <50% bootstrap support). The scale bar indicates estimated substitutions per site.   62   63  Supplementary Figure S5. Phylogram of angiosperm phylogeny inferred from a gene-by-codon (GxC) partitioned likelihood analysis of coding regions from whole plastid genomes (83 genes total; -lnL = 1,281,368.706); see text and Table S2 for partitioning scheme. Bootstrap support values are indicated beside branches where less than 100% (‘--’ indicates <50% bootstrap support). The scale bar indicates estimated substitutions per site.   64   65  Supplementary Figure S6. Phylogram of angiosperm phylogeny inferred from a gene-by-codon (GxC) partitioned likelihood analysis of coding regions from whole plastid genomes (83 genes total; -lnL = 1,271,860.239) in which ndh genes were entered as blanks for those species in which one or more of these was pseudogenized or lost; see text and Table S2 for partitioning scheme. Bootstrap support values are indicated beside branches where less than 100% (‘--’ indicates <50% bootstrap support). The scale bar indicates estimated substitutions per site. 66   


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