Organisms Diversity and Evolution

Interrelationships of Nemertodermatida --Manuscript Draft--

Manuscript Number: ODAE-D-15-00083R1 Full Title: Interrelationships of Nemertodermatida Article Type: Original Article Keywords: Nemertodermatida; LSU; SSU; molecular phylogeny; cryptic species; Approximately Unbiased test Corresponding Author: Inga Meyer-Wachsmuth, PhD Biology Centre of the Czech Academy of SciencesBiology Centre of the Czech Academy of Sciences České Budĕjovice, CZECH REPUBLIC Corresponding Author Secondary Information: Corresponding Author's Institution: Biology Centre of the Czech Academy of SciencesBiology Centre of the Czech Academy of Sciences Corresponding Author's Secondary Institution: First Author: Inga Meyer-Wachsmuth, PhD First Author Secondary Information: Order of Authors: Inga Meyer-Wachsmuth, PhD Ulf Jondelius, Prof Dr Order of Authors Secondary Information:

Funding Information: Vetenskapsrådet Ulf Jondelius (2009-5147) Vetenskapsrådet Ulf Jondelius (2012-3913) Riksmusei Vänner Dr Inga Meyer-Wachsmuth (stipend 2011) Stiftelsen Lars Hiertas Minne Dr Inga Meyer-Wachsmuth (FO2011-0248) Royal Swedish Academy of Sciences Dr Inga Meyer-Wachsmuth (FOA11H-352)

Abstract: Nemertodermatida is a small taxon of microscopic marine worms, which were originally classified within Platyhelminthes. Today they are hypothesized to be either an early bilaterian lineage or the sister group to Ambulacraria within Deuterostomia. These two hypotheses indicate widely diverging evolutionary histories in this largely neglected group. Here, we analyze the phylogeny of Nemertodermatida using nucleotide sequences from the ribosomal LSU and SSU genes and the protein coding Histone 3 gene. All currently known species except Ascoparia neglecta and A. secunda were included in the study in addition to several yet undescribed species. Ascopariidae and Nemertodermatidae are retrieved as separate clades, although not in all analyses as sister groups. Non-monophyly of Nemertodermatida was rejected by the Approximately Unbiased test. Nemertodermatid nucleotide sequences deposited in Genbank before 2013 were validated against our dataset; some of them are shown to be chimeric implying falsification of prior hypotheses about nemertodermatid phylogeny: other sequences should be assigned new names. We also show that the genus Nemertoderma needs revision. Response to Reviewers: Answers will be uploaded in a separate file.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Response to reviewer Click here to download Authors' Response to Reviewers' Comments: AnswerToReviewers.docx

Reviewer #1: The manuscript is summarising a good piece of work and is addressing an important problem of internal phylogeny of Nemertodermatida. The methods are state of the art. I have no comments or suggestions for changes to this manuscript. I suggest publication as it is. Dear Reviewer #1,

Thank you very much for your favorable view on our work; we are delighted to hear this.

Reviewer #2: The manuscript deals with the phylogeny of Nemertodermatida. The authors have done an important sequencing effort, the phylogenetic analyses are well done, and the trees seem informative enough. I think the ms needs, however, some serious re-writting to better emphasize the findings, specially with regards the higher-level classification and implications. Other than that, I suggest publication, although I have some minor comments and corrections to be addressed.

Dear Reviewer #2,

First of all I would like to thank you for your thorough review of our manuscript. It is good to see such detailed comments. And thank you also for pointing out these nasty small mistakes such as missing italics, double words, wrong spellings etc. One just doesn’t seem to see them after some time anymore. Otherwise we went with nearly all your suggestions but for those on Figure 1, since the sequences do not blast to any specific species with high identity. Table 2 was hard to understand, because we missed to add a note that the new accesion numbers had not been added yet, or placeholders for them. We hope that the addition of those and the changes of the asterisks make this table much better to understand now. Please find point-by-point answers below:

Minor comments

Abstract I suggest to include a bit of background of Nemertodermatida and why it is important to decipher its phylogeny. The authors have done a good job in the introduction but the abstract goes directly to results. We added a few sentences on the importance of Nemertodermatida.

Page 3, lines 70, 71. "not part of Platyhelminthes or Acoelomorpha". Change to something like "not part of Platyhelminthes, and not sister-group to Acoela". Changed as suggested, thanks.

Page 3, line 73. Ruitort, change to "Riutort" Changed in manuscript and in Papers.

Page 3, line 83. "i) is Acoelomorpha monophyletic, i.e., is Nemertodermatida the sister group of Acoela? ". I sugest changing to : i) is Nemertodermatida the sister group of Acoela forming the Acoelomorpha?" Changed to be easier to read and be more precise.

Material and methods Please specify exactly which are the specimens collected. It is not clear at all. Are there the one without asterisk in table 2. If so, I suggest to depict them by asterisk instead. We changed this as you suggested, the asterisk highlighting the new sequences now. Since we did not have accession numbers for new sequences, the table was possibly even more confusing; we should have added placeholders for the review process.

Page 6, lines 150-151. "Meara sp. and species of the genus Ascoparia and could not be collected". Maybe deleted "and"? Done!

Page 6, lines 151-152. "About 45 new specimens of...were sequenced". Which 45 specimens? Please clarify. Maybe with a table? The dataset was built in steps; many of the sequences of the 45 specimens are not being published here. We nevertheless mention the process of how we arrived at the current dataset in order to forestall criticism of the dataset built on such a small number of specimens per species.

Page 6, lines 155-156. "highly similar sequences of Meara, Flagellophora and Nemertoderma were discarded in order to reduce the size of the dataset.". How similar? how many of them were similar? similar for which genes? Please clarify. We specified the similarity (99.9% identity) in the text and that we did this in all three genes. We believe that it is not important for this manuscript, and in contrast will rather introduce more confusion, to distinguish for example in the case of Nemertoderma different lineage and describe how many specimens each had in the preliminary analyses. Flagellophora sequences were more variable then previously expected and sequences should only be published with a proper discussion of population genetics and species delimitation, which is beyond the scope of this manuscript.

Page 6, line 164. "new sequences are published in GenBank (Tab. 2)." which are the accession numbers? The taxa without asterisk does not have GenBank accession numbers. Are those the new sequences? Again, I think table 2 and the taxon sampling is not clear at all. The accession numbers have been added now that they were available. We also changed the use of the asterisk according to your remarks, they now are directly in front of the accession numbers and highlight sequences published new in this manuscript.

Page 8. I do not find the explanation of Nucleotide frequencies that relevant. I suggest to take most of that paragraph out. We agree, it is a lot of detail. We took most of the paragraph (we only left the first sentence) and added it to the table in the online resources 1.

Page 10. Including the nodal support in figure 4 will allow a better understanding of this part. Nodal support has been added.

Page 13, lines 389-392. I do not see the LSU tree (Figure 4) recovering (Nemertoderma 391 (Meara, (Nemertinoides + Sterreria)))). Please clarify. In Figure 4, the addition of chimeric sequences changed the topology of the tree. Figure 4 only exemplifies the effect that these two chimeric sequences had on the phylogenetic inference of this dataset. The tree(s) referred to (lines389-392) are the concatenated tree of Fig. 3 (references added to the manuscript) and the gene trees in the supplementary material (as referred to in the same sentence).

Page 14-15. I think the authors could re-write this part a bit to make the findings and the implications more clear. I also think a note should be done here for the need of additional taxon sampling, specially with regards Ascopariidae. We edited these paragraphs to make them clearer and added a call for colleting more data. We refrained from using the word taxon sampling, since we believe that what is needed is a wider geographical sampling that might very well turn out to be the same as enhanced taxon as shown in Meyer-Wachsmuth et al. 2014, but we do not know that.

Table 1. I suggest including the sequences of the primers. This way the table would be more complete and informative. We added the sequences of the primers as well as the references.

Figures Figure 1 is hard to understand. I would suggest including the sequence of the ones they blast to (the gastrotrich species, the unidentified platyhelminth species, and the copepod). The ID and sometimes query coverage of the blasted sequences is often very bad, as reference in the text. Adding these sequences into the figure would be an arbitrary choice in 3 out of the 4 cases and more pointedly ad further confusion with mismatches. Had clear

Figure legend 1. Italize "Meara stipochi". Done

Figure 3. Some of the lines of the coloured rectangles interfere with a good visual of the nodal support. Consider removing them (change by colour labels on the right--there is enough space if species names are situated at the tip of the branches) or make them without lines. We chose to keep the coloured boxes, which make colours less ambiguous than only lines, which have to have a certain weight for colours to be easily distinguishable. But we removed the lines around the boxes and added arrows where nodal support conflicted with the tree.

Figure 4. Why there are not nodal support values? If possible, include labels "LSU" and "SSU" below each tree. I also suggest including labels on the right for each group. Nodal support values have been added as well as the labels LSU and SSU, but instead of a) and b) above the tree. We also added group labels next to the trees.

We are very happy that this manuscript has been regarded favorably.

With kind regards,

Inga Meyer-Wachsmuth on behalf of all authors Click here to download Manuscript: Manuscript_IMW.UJ_revision.docx Click here to view linked References

1 2 3 4 5 6 7 1 Interrelationships of Nemertodermatida 8 9 2 10 11 3 Inga Meyer-Wachsmuth1,2,3 and Ulf Jondelius1,2 12 13 4 14 5 1 Department of Zoology, Swedish Museum of Natural History, Box 50007, 104 05 15 16 6 Stockholm, Sweden

17 2 18 7 Department of Zoology, Stockholm University, 104 05 Stockholm, Sweden 19 8 3 present address: Institute of Parasitology, Biology Centre of the Czech Academy of 20 21 9 Sciences, Branišovská 31, 370 05 České Budĕjovice, Czech Republic 22 10 23 24 11 Corresponding author: Ulf Jondelius ([email protected]) 25 26 12 27 28 13 Abstract: 29 30 14 Nemertodermatida is a small taxon of microscopic marine worms, which were Formatted: Left, Widow/Orphan control 31 15 originally classified within Platyhelminthes. Today they are hypothesized to be either 32 33 16 an early bilaterian lineage or the sister group to Ambulacraria within Deuterostomia. 34 17 These two hypotheses indicate widely diverging evolutionary histories in this largely 35 18 neglected group. Here, we analyze Tthe phylogeny of Nemertodermatida, a group of 36 37 19 microscopic marine worms, was analysed using nucleotide sequences from the ribosomal 38 20 LSU and SSU genes and the protein coding Histone 3 gene. All currently known species 39 40 21 except Ascoparia neglecta and A. secunda were included in the study in addition to several 41 22 yet undescribed species. Ascopariidae and Nemertodermatidae, are retrieved as separate 42 23 clades, although not in all analyses as sister groups. Non-monophyly of Nemertodermatida 43 44 24 was rejected by the Approximately Unbiased test. Nemertodermatid nNucleotide 45 25 sequences deposited in Genbank before 2013 as nemertodermatid were validated against 46 26 our dataset; some of them are shown to be chimeric implying falsification of prior 47 48 27 hypotheses about nemertodermatid phylogeny: other sequences should be assigned new 49 28 names. We also show that the genus Nemertoderma needs revision. Formatted: Font: (Default) Cambria, Not Italic, (Intl) 50 Arial Unicode MS 51 29 52 53 30 Keywords: Nemertodermatida; LSU; SSU; molecular phylogeny; cryptic species; 54 55 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 31 Approximately Unbiased test 8 9 32 Acknowledgements 10 11 33 We would like to thank Ms Keyvan Mirbakhsh for her work in the molecular lab. We are 12 34 also grateful to the staff at the Sven Lovén Centre for Marine Sciences, Dr Ana Amaral at 13 14 35 CCMAR, Faro, Ms Lisa Del Monte, ISZN, Naples with colleagues and Ms Margret Krüß, 15 36 BAH, Helgoland with colleagues for help with our fieldwork. We would also like to 16 37 express our gratitude to Prof. Mark Martindale then of the University of Hawaii Kewalo 17 18 38 Marine Lab. Furthermore are we indebted to the Professor Philippe Bouchet of the 19 39 Muséum National d'Histoire Naturelle for organizing sampling in Papua New Guinea. 20 40 Collections in Papua Guinea took place during the Our Planet Reviewed Papua Niugini 21 22 41 Expedition in November–December 2012, organized by the Muséum National d’Histoire 23 42 Naturelle (MNHN), Pro Natura International, the Institut de Recherche pour le 24 25 43 Développement (IRD), and the University of Papua New Guinea. The principal 26 44 investigators of this expedition were Philippe Bouchet, Sarah Samadi (MNHN) and Claude 27 45 Payri (IRD) and funding was provided by the Total Foundation, Prince Albert II of 28 29 46 Monaco Foundation, Fondation EDF, Stavros Niarchos Foundation and Entrepose 30 47 Contracting, with support from the Divine Word University and operated under a permit 31 32 48 delivered by the Papua New Guinea Department of Environment and Conservation. 33 49 Financial support from the Swedish Research Council to UJ is gratefully acknowledged 34 50 (grant numbers 2009-5147 and 2012-3913) as are the stipends by Föreningen Riksmusei 35 36 51 Vänner (stipend 2011), Stiftelsen Lars Hiertas Minne grant FO2011-0248 and the Royal 37 52 Swedish Academy of Sciences grant FOA11H-352 to I. Meyer-Wachsmuth. 38 39 53 40 41 54 42 43 44 45 46 47 48 49 50 51 52 53 54 55 2 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 55 Introduction 8 9 56 The phylogenetic position of Nemertodermatida, a group of microscopic marine worms, 10 57 has been a matter of contention ever since the first species was described by Steinböck 11 12 58 (1930), who classified it within the Platyhelminthes and considered it close to the ancestor 13 59 of the flatworms in terms of morphology. Westblad (1937) placed nemertodermatids in the 14 15 60 group Acoela, at that time considered an order within Platyhelminthes. Nemertodermatida 16 61 differ from acoels in a number of morphological characters including the presence of a gut 17 62 with an epithelial lining and the possession of a statocyst with two statoliths; these 18 19 63 differences led Karling (1940) to recognise the order Nemertodermatida as separate from 20 64 Acoela. Ultrastructural studies revealed similarities between acoels and nemertodermatids 21 65 in the epidermal cilia, and the two groups were hypothesised to form the monophylum 22 23 66 Acoelomorpha (Ehlers 1985). Analyses of nucleotide sequence data resulted in conflicting 24 67 phylogenetic hypotheses regarding the position of Nemertodermatida. Initially, SSU rDNA 25 26 68 sequences reported from the species Nemertinoides elongatus Riser, 1987 placed 27 69 Nemertodermatida within rhabditophoran flatworms (Carranza, Baguñà, & Ruitort 1997; 28 70 Littlewood, Rohde, Bray, & Herniou 1999). Subsequent analyses using sequences from 29 30 71 additional species placed Nemertodermatida as a separate early branching clade predating 31 72 the split into Deuterostomia and Protostomia, i.e., not part of Platyhelminthes or sister 32 33 73 group to Acoela, and demonstrated that the original “Nemertinoides” sequences were either Formatted: Font: Italic 34 74 contaminated or originating from a misidentified specimen (Jondelius, Ruiz-Trillo, 35 75 Baguñà, & Riutort 2002; Wallberg, Curini-Galletti, Ahmadzadeh, & Jondelius 2007). The 36 37 76 nemertodermatid position as a separate early bilaterian clade, i.e., rejecting Acoelomorpha, 38 77 was further supported by several protein coding nuclear genes (Paps, Baguñà, & Riutort 39 40 78 2009; Ruiz-Trillo et al. 2002). The first phylogenomic study including nemertodermatid 41 79 sequences (Hejnol et al. 2009) on the other hand supported Acoelomorpha, i.e. 42 80 Nemertodermatida and Acoela, as the sister group of Nephrozoa (the remaining Bilateria). 43 44 81 However, a contradictory result was found in a second phylogenomic study that upheld 45 82 Acoelomorpha, but placed this group nested within Deuterostomia (Philippe et al. 2011). 46 47 83 48 49 84 To summarise: the phylogenetic position of Nemertodermatida is highly controversial on 50 85 two levels: i) is Nemertodermatida the sister group of Acoela, thus forming the 51 86 monophyletic Acoelomorpha? and ii) are nemertodermatids a primarily simple early 52 53 87 bilaterian clade or are they drastically reduced deuterostomes? In either case a strongly 54 55 3 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 88 corroborated hypothesis regarding the interrelationships of nemertodermatids is of great 8 9 89 interest, as it would allow modelling of character evolution and ancestral character states 10 90 within the group and, by extension, either during early bilaterian evolution or within 11 91 deuterostomes. 12 13 92 The current classification of Nemertodermatida is based on Sterrer (1998) and Lundin 14 15 93 (2000) who recognised two families and six genera to accommodate the eight 16 94 nemertodermatid species described between 1930 and 1998 (Faubel & Dörjes 1978; Riedl 17 95 1983; Riser 1987; Steinböck 1930; Sterrer 1998; Westblad 1937; 1949). Recently, nine 18 19 96 new species were named and one junior synonym reinstated using molecular delimitation 20 97 methods, raising the nominal species count to 18, and there is evidence of further, as of yet 21 98 undescribed, species (Meyer-Wachsmuth, Curini-Galletti, & Jondelius 2014). 22 23 99 There are two conflicting hypotheses regarding the phylogenetic relationships within 24 25 100 Nemertodermatida. The first, which is derived from a cladistic analysis of light microscope 26 101 and ultrastructural characters (Lundin 2000), retrieved a basal dichotomy splitting the 27 102 taxon into two groups, Nemertodermatidae and Ascopariidae, in concordance with 28 29 103 Sterrer’s classification (1998). The second hypothesis iswas based on analyses of the large 30 104 and small ribosomal subunit genes with the primary goal of testing the monophyly of 31 32 105 Acoelomorpha (Wallberg et al. 2007), which was found to be non-monophyletic. In the 33 106 latter study Flagellophora apelti Faubel and Dörjes, 1978 and Meara stichopi (Bock) 34 107 Westblad 1949 formed a monophyletic group, thus rendering Nemertodermatidae and 35 36 108 Ascopariidae non-monophyletic. 37 38 109 Here we use nucleotide sequences from three molecular markers, large and small subunit 39 110 rDNA and histone H3 to generate the most comprehensive phylogenetic hypothesis of 40 111 Nemertodermatida to-date, including all but two of the known as well as several yet 41 42 112 undescribed species. The phylogenetic hypothesis provides a framework for the 43 113 interpretation of the evolution of such features as the nemertodermatid nervous system 44 114 (Raikova et al.,dealt with in a separate publication in this issue), and for an updated 45 46 115 classification of Nemertodermatida, including a test of the monophyly of nemertodermatid 47 116 families and genera. Additionally, we also validate nemertodermatid sequences currently 48 49 117 available on GenBank. 50 118 51 52 119 53 54 55 4 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 120 Material and Methods 8 9 121 Collection of Specimens 10 11 122 Marine meiofauna was extracted from sandy sediments using isotonic magnesium chloride 12 123 solution (Sterrer 1968) and from muddy sediments by the siphoning off the uppermost 13 14 124 layer of settled mud through a 125µm sieve. Specimens were then sorted and identified 15 125 under a dissecting and compound microscope, if possible equipped with differential 16 126 interference contrast optics, before fixation in ethanol or RNAlater®. Whenever feasible 17 18 127 microphotographs were taken of live specimens to serve as vouchers, because DNA is 19 128 extracted from whole specimens. 20 21 129 22 23 130 DNA extraction, amplification and sequencing 24 131 DNA was extracted using the Qiagen Micro Tissue Kit. Limited availability of sequence 25 26 132 data and low yield of DNA, owing to the microscopic size of the worms, restricted testing 27 133 of new primers and therefore choice of genetic markers. We were able to amplify and 28 29 134 sequence part of the large and small ribosomal subunit genes (LSU and SSU, respectively), 30 135 as well as a part of the nuclear protein coding Histone 3 gene (H3). All markers were 31 136 amplified and sequenced using several different primer combinations (Tab. 1) and, in case 32 33 137 of SSU, a nested PCR approach. Sequences were edited, aligned and, in case of H3, 34 138 translated into amino acids and checked for open reading frames (standard genetic code), 35 36 139 using the commercial Geneious Pro 7.1.5. software package created by Biomatters 37 140 (http://www.geneious.com). Alignments were conducted using the MAFFT algorithm 38 141 (Katoh, Misawa, Kuma, & Miyata 2002). 39 40 142 41 42 143 Dataset assembly and outgroup choice 43 44 144 Outgroup taxa were chosen from various groups based on the different phylogenetic 45 145 hypotheses for Nemertodermatida. Early branching species of Acoela (Tab. 2) were added 46 47 146 due to their hypothesized sister group relationship to Nemertodermatida (Edgecombe et al. 48 147 2011; Hejnol et al. 2009; Philippe et al. 2011). In order to account for the unknown 49 148 phylogenetic position we also added species of hemichordates and crinoids 50 51 149 (Deuterostomia, Ambulacraria) and early branching molluscs (Protostomia). To test for 52 150 different rates of evolution between the in- and outgroups Tajima’s relative rate test 53 54 55 5 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 151 (Tajima 1993) was conducted using Mega 5.2.2. (Tamura et al. 2011). 8 9 152 The final dataset was assembled step-by-step. First, aA preliminary dataset was built 10 153 covering 24 out of 28 nominal or putative species of Nemertodermatida (Meyer- 11 12 154 Wachsmuth et al. 2014; Sterrer 1998). Meara sp. and species of the genus Ascoparia could 13 155 not be collected. About 45 new specimens of Nemertoderma, Meara and Flagellophora 14 15 156 were sequenced for all three genes (not all data included in final dataset). For all species of 16 157 Nemertinoides and Sterreria one specimen each was chosen for this dataset from a 17 158 previously published dataset based on high gene coverage (Meyer-Wachsmuth et al. 2014). 18 19 159 Secondly, and Aafter preliminary analyses, highly similar sequences of Meara, 20 160 Flagellophora and Nemertoderma that were 99.9% or more identical in any gene were 21 161 discarded and only one exemplary specimen left, based on gene coverage, in order to 22 23 162 reduce the size of the dataset. In case of the genus Nemertoderma similar sequences were 24 163 not discarded where there was conflict between morphological identification and genetic 25 26 164 placement. Our eleven Flagellophora specimens always formed a monophyletic group 27 165 separate from Nemertodermatidae in preliminary analyses but the sequence variation 28 166 between specimens was higher than in other nemertodermatids, indicating potential 29 30 167 presence of more than one Flagellophora species (a matter that cannot be studied until Formatted: Font: Italic 31 168 more specimens have been collected). To avoid confusion we chose to include only the 32 33 169 sequences from a single individual in our analyses. Finally, Aall nemertodermatid 34 170 sequences uploaded to GenBank prior to 2013 were then added to this reduced dataset for 35 171 validation. Four of those appeared chimeric in the alignments and were subsequently 36 37 172 excluded again (s. Tab. 2, Fig. 1). The effects of these sequences on phylogenetic analyses 38 173 were tested in separate analyses. 39 40 174 The final dataset consists of an ingroup of 4129 terminals covering 24 of the 28 nominal or 41 175 putative species of Nemertodermatida including nemertodermatid sequences downloaded 42 43 176 from GenBank for validation and outgroup taxa. with 28 sequences for each LSU and SSU, 44 177 and 21 sequences for H3, Nnew sequences, two of them for the outgroup species 45 178 Paratomella rubra Rieger & Ott, 1971, are published in GenBank (marked with an asterisk 46 47 179 in Tab. 2). Ten, twelve and one sequence(s) for LSU, SSU and H3, respectively, were 48 180 downloaded from GenBank and included for validation (Tab. 2). Additional to the 49 50 181 sequences downloaded from GenBank for the outgroup taxa new sequences were produced 51 182 for the acoel species Paratomella rubra Rieger & Ott, 1971 (Tab. 2). For the concatenated 52 183 dataset, sequences downloaded from GenBank were combined to form “terminals” based 53 54 55 6 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 184 either on submitted specimen collection codes (e.g. Ascoparia sp. with collection code 8 9 185 “AWHel-24” stated in GenBank), similarity of accession number, or taxon, and coverage 10 186 of at least two genes (Tab. 2). 11 12 187 13 14 188 Data properties 15 189 Mega 5.2.2. (Tamura et al. 2011) was used to calculate the nucleotide frequencies for each 16 17 190 specimen, gene and codon position as well as Tajima’s D for H3 (Tajima 1989) and the 18 191 Disparity Index (Kumar & Gadagkar 2001) for the final dataset. The GC-content per 19 192 lineage and gene was calculated in Excel. Alignments were tested for random similarity 20 21 193 with the program Aliscore (Kück et al. 2010; B. Misof & Misof 2009) using the default 22 194 settings. jModeltest v. 2.1.1. (Darriba, Taboada, Doallo, & Posada 2012) was used with a 23 24 195 BioNJ starting tree to test for the importance of the proportion of invariable sites (I, Pinvar) 25 196 and the rate variation across sites (G), and to obtain values to set useful priors, based on the 26 197 Bayes Information Criterion (BIC). 27 28 198 29 30 199 Phylogenetic informativeness profiling and saturation 31 32 200 Phylogenetic informativeness profiles were produced with the PhyDesign webserver 33 201 (López-Giráldez & Townsend 2011) available at http://phydesign.townsend.yale.edu/ with 34 35 202 site rates being calculated by the inbuilt DNAr8es algorithm (unpublished). The ultrametric 36 203 tree was created from the concatenated dataset using the chronos algorithm (lambda=0.1, 37 204 model=correlated) of the splits package in R. Saturation plots were created by plotting the 38 39 205 pairwise uncorrected p-distances against the phylogenetic distance of the best tree inferred 40 206 from the concatenated dataset with RAxML using R (Klopfstein, Vilhelmsen, Heraty, 41 42 207 Sharkey, & Ronquist 2013). 43 208 44 45 209 Phylogenetic analyses 46 47 210 Gene trees and species trees were calculated using two different methods of phylogenetic 48 211 inference. Maximum Likelihood trees were calculated with the RAxMLGUI (Silvestro & 49 50 212 Michalak 2011; Stamatakis 2006) using the GTR+G+I model (Tab. 3), and the ML + rapid 51 213 bootstrapping algorithm (Stamatakis, Hoover, & Rougemont 2008) with 1000 replicates. 52 214 Bayesian estimations were conducted with the program MrBayes v.3.2. (Ronquist et al. 53 54 55 7 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 215 2012) with no model set and the program being allowed to sample the entire model space 8 9 216 of the GTR model. Priors for the proportion of invariable sites and G were set according to 10 217 the results of jModeltest (Tab. 3). To ensure sufficient convergence, analyses were run 11 218 until the standard deviation of split frequencies reached values below 0.02 and the values 12 13 219 for Potential Scale Reduction Factor (PSRF+) were close to 1.0. Trees were visualized 14 220 using FigTree v1.3.1. (Rambaut 2009). 15 16 221 We also tested different topological hypotheses discussed here and indicated in preliminary 17 222 analyses with the Approximately Unbiased (AU) test. The test is implemented in the 18 19 223 program consel (Shimodaira & Hasegawa 2001) and calculates the overall likelihoods of a 20 224 set of topologies based on the per-site log likelihoods (Shimodaira 2002). For this we 21 225 conducted ML analyses (RAxML 7.2.8) of the final dataset with the following threefour 22 23 226 topological constraints: i) non-monophyly of Nemertodermatida (Ascopariidae 24 227 (Nemertodermatidae + Acoela)), ii) Lundin’s (2000) Nemertodermatidae with a luterious 25 26 228 and a psammicolous clade (Ascopariidae, ((Meara + Nemertoderma), (Nemertinoides + Formatted: Font: Italic Font: Italic 27 229 Sterreria))), and iii) Wallberg et al.’s (2007) hypothesis clustering Meara and Formatted: 28 Formatted: Font: Italic 230 Flagellophora and rejecting the two-family hypothesis (Acoela, ((Flagellophora + Meara), 29 Formatted: Font: Italic 30 231 (Nemertoderma, (Nemertinoides + Sterreria)))). Topological hypotheses were only tested Formatted: Font: Italic 31 232 down to genus level. Species relationships have to be addressed by increased Formatted: Font: Italic 32 Formatted: Font: Italic 233 taxon/geographical sampling. 33 Formatted: Font: Italic 34 234 Formatted: Font: Italic 35 Formatted: Font: Italic 36 235 Results Formatted: Font: Italic 37 38 236 The final datasets consists of 44 (LSU), 46 (SSU) and 26 (H3) sequences, including ten, 39 237 twelve and one sequence(s), respectively, downloaded from GenBank for validation 40 41 238 (Tab. 2) and 16 sequences of outgroup taxa, six each for LSU and SSU, and four for H3. 42 43 239 44 240 Data properties 45 46 241 The relative rate test did not show significantly different rates between tested lineages, 47 242 allowing the use of the chosen outgroups. 48 49 243 Nucleotide frequencies within the ingroup are close to equilibrium, only Thymine can 50 51 244 reach frequencies as low as 15.2% in H3, and Guanine frequencies of 34.2% in LSU 52 245 (summary in Tab. 4, complete table as online resource 1). DThe observed small biases are 53 54 55 8 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 246 mostly consistent across the ingroup, for example Thymine has the lowest frequency of all 8 9 247 four nucleotides in LSU across all lineages. There are also lineage specific biases. 10 248 Nemertoderma, for example, has slightly lower frequencies of Cytosine and Guanine and 11 249 subsequently higher levels of Adenine and Thymine than Meara, Nemertinoides and 12 13 250 Sterreria in the LSU data (Tab. 4). These lineage specific biases, however, are overall 14 251 small. The GC-content of LSU and H3 is slightly elevated between 51.8% and 58.7%, 15 16 252 whereas in SSU it is balanced between 48.1% and 52.2%. The disparity indices reveal 17 253 homogenous substitution patterns throughout the whole LSU dataset but not the SSU and 18 254 H3 datasets (online resource 2). The main deviations in the SSU data are shown between 19 20 255 lineages, for example N. bathycola Steinböck, 1930 and N. westbladi Steinböck, 1937 21 256 versus pacific Sterreria species, Flagellophora versus S. psammicola and S. lundini, and 22 257 between two subgroups within Sterreria. Substitutions patterns in the H3 dataset are 23 24 258 overall heterogeneous. 25 26 259 Aliscore identified ambiguously aligned sites in both the LSU and SSU gene datasets but 27 260 not in the coding Histone 3 gene. In the LSU dataset overall 845bp out of 3751bp (22.5%) 28 261 were considered potentially randomly aligned, whereas in SSU it was 184 out of 2399 sites 29 30 262 (7.7%). Most of the ambiguity was detected in the beginning and end of the alignments, 31 263 where coverage is lower. In the LSU dataset several other areas, often only a few base 32 33 264 pairs long, were highlighted as problematic. These were in highly variable areas, probably 34 265 the loops of stem-loop secondary structures, where indels resulted in local low coverage. 35 266 ML tree topologies estimated from the Aliscore reduced datasets did not differ from the 36 37 267 original topologies in LSU and only in one node in SSU (SFonline resource. 31). 38 268 Nemertinoides sp. N1 was retrieved as earliest branch within the genus Nemertinoides in 39 40 269 the original dataset but as sister taxon to a clade (Nemertinoides (Meara + Sterreria) in the 41 270 Aliscore-reduced dataset. In both datasets the majority of nodes stayed stable or changed 42 271 only little (online resource 43a-f). Given the rule of thumb that any node below 75% 43 44 272 bootstrap support (BS) should not be considered reliable, the datasets did not change 45 273 substantially: two nodes in the LSU and none in the SSU dataset passed this threshold due 46 274 to the exclusion of randomly aligned sites, none of the nodes in either dataset decreased 47 48 275 below it. For both genes we decided to use the original datasets due to negligible effects of 49 276 the data exclusion. 50 51 52 53 54 55 9 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 277 With D = 3.100814 Tajima’s D was insignificant. The results of jModeltest strongly 8 9 278 suggested the use of the proportion of invariable sites (Pinvar, I) and rate variation across 10 279 sites (G) (Tab. 3). 11 12 280 13 14 281 Phylogenetic informativeness (PI) profiling and saturation 15 282 Absolute measures of PI are not possible. Due to a lack of nemertodermatid fossils and 16 17 283 widely varying phylogenetic hypotheses, reliable time calibration of the ultrametric tree 18 284 used for PI profiling is not feasiblepossible. Relative comparisons of the genes are, 19 285 however, achievable (Fig. 2). Overall, the LSU gene is the most informative gene across 20 21 286 the depth of the tree, while Histone 3 is the least informative one, partially due to its short 22 287 length. The SSU gene is comparatively little informative especially in the shallow nodes 23 24 288 but reaches its peak informativeness only in deeper nodes and stays informative over the 25 289 depth of most of the ingroup. 26 27 290 Saturation plots indicate that H3 is saturated with the third position being the most variable 28 291 and thus the most saturated, while the second position is the most conserved and least 29 30 292 saturated (online resource 54). LSU and SSU and do not show an increase of uncorrected 31 293 p-distance with increased phylogenetic distance und thus no clear signs of saturation. 32 33 294 34 35 295 Phylogeny 36 296 The gene trees of LSU and SSU support the same clusters corresponding to the 37 38 297 morphologically delimited known genera (online resource 45a-d). Furthermore, both 39 298 datasets fully support Nemertodermatidae and Ascopariidae as monophyletic clades; the 40 41 299 latter is never nested within the former. However, only the LSU and concatenated analyses 42 300 recover Ascopariidae and Nemertodermatidae as sister groups, i.e. with BS values above 43 301 75% and Bayesian posterior probability (BPP) above 0.95 with high support (online 44 45 302 resource 45a, b, Fig. 3) whereas SSU does not resolve the relationships between Acoela, 46 303 Nemertodermatidae (72% BS, 0.84 Bayesian posterior probability (BPP)) and 47 304 Ascopariidae (84% BS,. 1.0 BPP; online resource 45c, d). The Histone 3 gene does not 48 49 305 recover any of the genera but Nemertinoides (online resource 45e, f). Most deeper nodes 50 306 are not resolved and bootstrap values overall very low. 51 52 307 The species relationships within the genera are inconsistent between the gene trees. In the 53 54 55 10 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 308 LSU analyses the cluster S. boucheti Meyer-Wachsmuth et al., 2014 + S. variabilis Meyer- 8 9 309 Wachsmuth et al., 2014 is retrieved as sister group to the clade that includes S. psammicola 10 310 (Faubel, 1976) but in the SSU analyses as sister group to the clade that includes S. ylvae 11 311 Meyer-Wachsmuth et al., 2014 in (online resource 45a-d). Similarly are the relationships 12 13 312 within Nemertinoides inconsistent between the two datasets. These variances are the same 14 313 as observed by Meyer-Wachsmuth et al. (2014). Within the genus Nemertoderma, both 15 16 314 LSU and SSU show two or more clusters, which, however, do not support the two 17 315 morphological species N. bathycola and N. westbladi. 18 19 316 On genus level, there are differences in topology within Nemertodermatidae. Analyses of 20 317 the LSU and the concatenated data recovered Sterreria and Nemertinoides as sister groups, 21 318 a clade also supported by Lundin (2000) and Wallberg (2007). In the SSU analyses, 22 23 319 however, this group is not recovered but Meara included into it. 24 25 320 Addition of the chimeric sequences to the final datasets changed the estimated topologies. 26 321 In the LSU dataset the relationships between Nemertoderma, Meara and the clade 27 322 Nemertinoides + Sterreria was not resolved anymore. The changes estimated in the SSU 28 29 323 dataset were more dramatic with Meara being retrieved as sister group to a clade 30 324 Flagellophora + Ascoparia, the topology suggested by Wallberg et al. (2007) and tested 31 32 325 here as hypothesis 3 with the AU test. 33 326 The AU test rejected all three alternative hypotheses, i.e. the non-monophyly of 34 35 327 Nemertodermatida (hypothesis 1), the monophyly of Meara + Nemertoderma (hypothesis 36 328 2, Lundin 2000), and the monophyly of Meara + Flagellophora (hypothesis 3, Wallberg et 37 38 329 al. 2007) (Tab. 5). 39 330 40 41 331 GenBank sequence validation 42 43 332 Four sequences downloaded from GenBank appear to be chimeric and another four are 44 333 registered under incorrect species names (Tab. 2). 45 46 334 The first about 700bp of the LSU sequence with the accession number AM747472 47 335 (Wallberg et al. 2007), registered as F. apelti, are more similar to the sequence of the 48 49 336 gastrotrich species Chaetonotus neptuni registered in GenBank under the accession number 50 337 JQ798610. Despite a high coverage (99%) and an E-value of 0 the identity is only 90%, i.e. 51 52 338 species identity is not clear (Fig. 1a). The rest of the sequence is most similar to the 53 54 55 11 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 339 sequence of Ascoparia sp. registered as FR837762 in GenBank (E-value 0, query cover 8 9 340 79%, identity 93%). The sequence AM747478 (Wallberg et al. 2007) is registered as N. 10 341 bathycola in GenBank but the first 678bp appear different in the alignment (Fig. 1b). Blast 11 342 returned several sequences with low E-values, query coverage of about 61% and identities 12 13 343 of around 82% all belonging to rhabdocoel species. 14 15 344 The SSU sequence AM747471 (Wallberg et al. 2007) is registered as F. apelti but only the 16 345 first about 970bp align well with newly sequenced Flagellophora sequences. The last 17 346 about 840bp are more similar to M. stichopi, by eye in the alignment (Fig. 1c) as well as in 18 19 347 BLAST, where they show E-values of 0 and 99% coverage and identity to several 20 348 sequences of M. stichopi, including the one registered as AM747473. Of the sequence 21 349 AF051328 (Telford, Lockyer, Cartwright-Finch, & Littlewood 2003), registered as 22 23 350 M. stichopi, only the last about 390bp match other M. stichopi sequences in our 24 351 preliminary dataset and GenBank (e.g. AF119085, E-value 0, coverage 97%, identity 25 26 352 100%). The first about 1410bp of the sequence BLAST as several different species of 27 353 copepods, with e-values of 0, coverage of about 96% and identities around 91% (Fig. 1d). 28 29 354 The LSU sequences AM747476 and AM747480 (Wallberg et al. 2007) are registered in 30 355 GenBank as Nemertinoides elongatus and Sterreria psammicola, respectively. In our 31 32 356 dataset, however, these sequences cluster with Sterreria rubra Meyer-Wachsmuth et al., 33 357 2014. 34 35 358 The SSU sequence AM747475 (Wallberg et al. 2007) is registered as N. elongatus but is 36 359 recovered in all analyses as S. psammicola, whereas sequence AM747479 (Wallberg et al. 37 38 360 2007) is registered as S. psammicola and clusters with S. rubra. 39 361 40 41 362 Discussion 42 43 363 Data properties 44 45 364 In this dataset the LSU was found to be the most informative of the three genes used, in the 46 365 shallower (more recent) as well as the deeper nodes. The often used more conserved SSU 47 366 gene appears less informative according to the PI profiles even in the deeper nodes, where 48 49 367 it would have been expected to show higher resolution power. The low resolution power of 50 368 SSU in the shallower nodes is also confirmed in the topology of the gene trees, where it 51 52 369 shows only two clades in the genus Nemertoderma while LSU recovers three distinct 53 54 55 12 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 370 clades (by eye). This confirms a previously noted lower resolution power and subsequent 8 9 371 underestimation of species diversity in several groups of meiofauna (Tang et al. 2012) and 10 372 shows that SSU is not a suitable marker for biodiversity studies. The comparatively poor 11 373 performance of SSU versus LSU is also evident in the PI profiles per site (data not shown). 12 13 374 Our analyses of the ribosomal and concatenated datasets support the same supraspecific 14 15 375 groups that were previously identified on the basis of morphological characters (Lundin 16 376 2000; Meyer-Wachsmuth et al. 2014; Sterrer 1998). Analysis of the Histone 3 nucleotide 17 377 sequences failed to recover any genus level or deeper clusters. 18 19 378 20 21 379 Species identities 22 23 380 Relationships within the genera Sterreria and Nemertinoides are not fully resolved and in 24 381 some cases they differ between the gene trees and estimation method. The position of 25 382 S. martindalei Meyer-Wachsmuth et al., 2014, for example, is unresolved as it shifts 26 27 383 between being sister group to all other Sterreria species in the SSU tree with very low 28 384 support (48% BS. 0.81 BPP) and sister to the pacific Sterreria subgroup including 29 30 385 S. monolithes Meyer-Wachsmuth et al., 2014 in the LSU trees, again with very poor 31 386 support (53% BS, 0.66 BPP). Genetic distances between S. martindalei species and any 32 387 other species are large (17.1% LSU, 14.6% SSU, online resource 6). This is arguably due 33 34 388 to incomplete taxon sampling in our dataset. Knowledge of nemertodermatid diversity and 35 389 distribution in the Pacific is extremely limited, with the whole Pacific being represented in 36 37 390 our dataset by only three different collection areas, all of them tropical (Hawaii, Papua 38 391 New Guinea and New Caledonia). The Indian Ocean is even less known in terms of 39 392 Nemertodermatida: only one record of Nemertodermatida has ever been published (Todt 40 41 393 2009) and no specimens from the Indian Ocean could be incorporated into this dataset. 42 394 More complete geographical and taxon sampling would surely help to resolve species 43 44 395 relationships. This is true alsoeven for Europe, arguably the best studies area in terms of 45 396 nemertodermatid diversity, where no nemertodermatid sequences are available from e.g. 46 397 the Eastern Mediterranean or the British Isles. 47 48 398 According to current classification, the genus Nemertoderma comprises two species, 49 399 N. bathycola and N. westbladi. The latter was said to be “one of the best defined species” 50 51 400 (Sterrer 1998) of Nemertodermatida. The clusters observed on the LSU and SSU datasets, 52 401 however, are not congruent with the two morphologically delimited species (Fig. 3). Given 53 54 55 13 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 402 the unresolved position of the nominal species and the apparent presence of at least one 8 9 403 additional species level taxa even within the small geographic sampling area on the 10 404 Swedish west coast, it is clear that the genus Nemertoderma requires further study. 11 12 405 13 14 406 Supraspecific groups 15 407 In the trees derived from analyses of the LSU and the concatenated datasets the 16 17 408 relationships within the family Nemertodermatidae are resolved as (Nemertoderma 18 409 (Meara, (Nemertinoides + Sterreria)))), while analyses of the SSU dataset did not resolve 19 410 this group (Fig. 3, online resource 45a-d). Nemertinoides and Sterreria share five 20 21 411 ultrastructural apomorphies (Lundin 2000), with Meara they share the apomorphies “testes 22 412 longitudinal follicles” and “ovaries paired”. In his analyses of morphological data Lundin 23 24 413 (2000) recovered Meara and Nemertoderma as sister groups (hypothesis 2) supported by 25 414 the two apomorphies “intracellular tonofilament bundles present” and “thick terminal 26 415 web”. However, this topology was not recovered in any of our gene trees or the 27 28 416 concatenated analyses (Fig. 3) and it was rejected by the AU test (Tab. 5). We suggest that 29 417 the two apomorphies hypothesised by Lundin (2000) for Meara + Nemertoderma may not 30 31 418 be particularly strong as intracellular monofilament bundles should occur in all epithelial 32 419 cells with desmosomes (Lane 1982) and they may have gone undetected in the 33 420 micrographs used for coding other nemertodermatid species. The terminal web was scored 34 35 421 using the three states “weak”, “distinct, but narrow” or “thick” with all other 36 422 nemertodermatids scored as “distinct, but narrow”. As the appearance of the terminal web 37 38 423 is affected by sample preparation, and as much of the data came from several different 39 424 studies in the literature, the distinction between the character states used by Lundin (2000) 40 425 may be vague. 41 42 426 43 44 427 Higher level groups 45 46 428 Wallberg et al. (2007) retrieved a topology in which ((Meara + Flagellophora) + 47 429 (Sterreria + Nemertinoides)) where the position of Flagellophora and as sister taxon of 48 430 Meara are sister groups. This conflicts with the classification of Sterrer (1998) and. The 49 50 431 we neverFlagellophora + Meara group was not recovered it in any of our analyses. 51 432 Furthermore, nor do, these speciesFlagellophora and Meara do not share any derived Formatted: Font: Italic 52 Formatted: Font: Italic 53 433 character states in the cladistic analysis by (Lundin, (2000). This grouptopology was also 54 55 14 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 434 rejected by our AU test (hypothesis 3). Both the LSU and SSU “Flagellophora” sequences 8 9 435 as well as one of the Meara sequences used by Wallberg et al. (2007) can be shown to be 10 436 chimeric (Fig. 1). The hypothesis of Wallberg et al. (2007) regarding the phylogenetic 11 437 position of Flagellophora apelti and Meara stichopi can therefore be rejected as caused by 12 13 438 chimeric input sequences (Fig 4). 14 15 439 In Sterrer’s (1998) classification The two family level groups, Ascopariidae and 16 440 Nemertodermatidae, have been introduced based on several light microscopic differences 17 441 and the same groups were retrieved by Lundin’s cladistic analysis (2000) of morphological 18 19 442 and ultrastructural characters. The divergence between these two clades is supported in all 20 443 our analyses, but the H3 gene analyses. were recovered as separate clades in our 21 444 concatenated analysis. However, they Ascopariidae and Nemertodermatidae are only 22 23 445 retrieved as sister groups in the LSU dataset (91% BS, 0.99 BPP). In the analyses of Tthe 24 446 SSU, H3 and concatenated datasets do not resolve the relationships between 25 26 447 Nemertodermatidae, Ascopariidae and Acoela are not resolved. This prompted us to test 27 448 for non-monophyly of Nemertodermatida, which, however, was rejected by the AU test. 28 449 The monophyly of Nemertodermatida has not previously been tested with a comprehensive 29 30 450 dataset including representatives of Nemertodermatidae and Ascopariidae as well as 31 451 Acoela and an appropriate outgroup. Most morphological and molecular studies of 32 33 452 Nemertodermatida used only species of Nemertodermatidae, e.g. (e.g. Hooge 2001; 34 453 Jimenez-Guri, Paps, Garcia-Fernandez, & Salo 2006; Jondelius et al. 2002; Lundin 2001; 35 454 Lundin & Hendelberg 1998; Raikova, Reuter, Jondelius, & Gustafsson 2000; Ruiz-Trillo et 36 37 455 al. 2002; Todt 2009). Only three studies also included Ascopariidae (Boone, Bert, Claeys, 38 456 Houthoofd, & Artois 2011; Rieger & Ladurner 2003; Wallberg et al. 2007). A striking 39 40 457 synapomorphy of Nemertodermatida is the statocyst, which differs from the acoel 41 458 statocysts notMorphological synapomorphies of the Nemertodermatida to the exclusion of 42 459 Acoela is e.g. the built of the statocyst. Not only in the number of statoliths in the statoliths 43 44 460 differ between Acoela and Nemertodermatida but also in its ultrastructure. In Acoela the 45 461 lithocyte, the statolith-forming cell, contains the statolith and dorsal to that a nucleus, 46 462 which can sometimes be seen under the microscope as round structure on the statocyst 47 48 463 (Ehlers 1985; Ferrero 1973; Ferrero & Bedini 1991). In Nemertodermatidae the lithocyte 49 464 dissolves and merges with the extracellular matrix of the vacuole containing the statolith 50 51 465 and the lithocyte nucleus is attached to the statolith, and visible as a “blister cap” (Sterrer 52 53 54 55 15 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 466 1998). In Ascopariidae the statoliths appear smooth, the ultrastructure of the statocyst, 8 9 467 however, has not been studied yet. 10 468 The present study is the first to focus on the interrelationships of nemertodermatids with 11 12 469 molecular means. The classification of Nemertodermatida in the two main groups 13 470 Ascopariidae and Nemertodermatida (Sterrer 1998) is corroborated by our results. 14 15 471 However, both previous phylogenetic hypotheses (Lundin 2001, Wallberg et al. 2007) are 16 472 falsified. There are still unresolved issues in nemertodermatid evolution: species identities 17 473 in Nemertoderma and Flagellophora are unclear and there are indications that these taxa Formatted: Font: Italic 18 Formatted: Font: Italic 19 474 are more diverse than current classification suggests. Overall geographic sampling of 20 475 Nemertodermatida is very limited (see taxon map at acoela.myspecies.info), and the 21 476 diversity is likely to be drastically underestimated (Meyer-Wachsmuth et al., 2014). Our 22 23 477 phylogenetic hypothesis enables analyses of character evolution and ancestral features 24 478 within Nemertodermatida, e.g. of nervous system, musculature and reproductive organs. 25 26 479 This is of great interest in the study of bilaterian evolution. Formatted: Font: 27 480 28 29 481 30 31 482 References 32 33 483 Boone, M., Bert, W., Claeys, M., Houthoofd, W., & Artois, T. J. (2011). Spermatogenesis 34 484 and the structure of the testes in Nemertodermatida. Zoomorphology, 130(4), 273–282. 35 485 Carranza, S., Baguñà, J., & Ruitort, M. (1997). Are the Platyhelminthes a monophyletic 36 486 primitive group? An assessment using 18S rDNA sequences. Molecular Biology and 487 Evolution, 14(5), 485–497. 37 488 Darriba, D., Taboada, G. L., Doallo, R., & Posada, D. (2012). jModelTest 2: more models, 38 489 new heuristics and parallel computing. Nature Methods, 9(8), 772–772. 39 490 Edgecombe, G. D., Giribet, G., Dunn, C. W., Hejnol, A., Kristensen, R. M., Neves, R. C., 40 491 et al. (2011). Higher-level metazoan relationships: recent progress and remaining 41 492 questions. Organisms Diversity & Evolution, 11(2), 151–172. 42 493 Ehlers, U. (1985). Das Phylogenetische System der Platyhelminthes (pp. 1–15). New York: 43 494 Gustav Fischer Verlag. 44 495 Faubel, A. (1976). Interstitielle Acoela (Turbellaria) aus dem Litoral der nordfriesischen 45 496 Inseln Sylt und Amrum (Nordsee). Mitteilungen aus dem Hamburgischen 46 497 Zoologischen Museum und Institut, 73, 17-56. 47 498 Faubel, A., & Dörjes, J. (1978). Flagellophora apelti gen. n. sp. n.: A remarcable 48 499 representative of the order Nemertodermatida (Turbellaria: Archoophora). 49 500 Senckenbergiana maritima, 10(1), 1–13. 501 Ferrero, E. A. (1973). A fine structural analysis of the statocyst in Turbellaria Acoela. 50 502 Zoologica Scripta, 2(5), 5–16. 51 503 Ferrero, E. A., & Bedini, C. (1991). Ultrastructural aspects of nervous-system and statocyst 52 504 morphogenesis during embryonic development of Convoluta psammophila 53 54 55 16 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 505 (Turbellaria, Acoela). Hydrobiologia, 227, 131–137. 8 506 Hejnol, A., Obst, M., Stamatakis, A., Ott, M., Rouse, G. W., Edgecombe, G. D., et al. 9 507 (2009). Assessing the root of bilaterian animals with scalable phylogenomic methods. 10 508 Proceedings of the Royal Society B: Biological Sciences, 276(1677), 4261–4270. 11 509 Hooge, M. D. (2001). Evolution of the body-wall musculature in the Platyhelminthes 12 510 (Acoelomorpha, Catenulida, Rhabditophora). Journal of Morphology, 249, 171–194. 13 511 Jimenez-Guri, E., Paps, J., Garcia-Fernandez, J., & Salo, E. (2006). Hox and ParaHox 14 512 genes in Nemertodermatida, a basal bilaterian clade. The International Journal of 15 513 Developmental Biology, 50(8), 675–679. 16 514 Jondelius, U., Ruiz-Trillo, I., Baguñà, J., & Ruitort, M. (2002). The Nemertodermatida are 515 basal bilaterians and not members of the Platyhelminthes. Zoologica Scripta, 31, 201– 17 516 215. 18 517 Karling, E. (1940). Zur Morphologie und Systematik der Alloeocoela cumulata and 19 518 Rhabdocoela lecitophora Turbellaria. Acta Societatis pro Fauna et Flora Fennica, 26. 20 519 Katoh, K., Misawa, K., Kuma, K.-I., & Miyata, T. (2002). Mafft: a novel method for rapid 21 520 multiple sequence alignment based on the fast Fourier transform. Nucleic Acids 22 521 Research, 30(14), 3059–3066. 23 522 Klopfstein, S., Vilhelmsen, L., Heraty, J. M., Sharkey, M., & Ronquist, F. (2013). The 24 523 hymenopteran Tree of Life: Evidence from protein-coding genes and objectively 25 524 aligned ribosomal data. PLoS ONE, 8(8), e69344. 26 525 Kumar, S., & Gadagkar, S. R. (2001). Disparity Index: A simple statistic to measure and 27 526 test the homogeneity of substitution patterns between molecular sequences. Genetics, 28 527 158, 1321–1327. 528 Kück, P., Meusemann, K., Dambach, J., Thormann, B., Reumont, von, B. M., Wägele, J. 29 529 W., & Misof, B. (2010). Parametric and non-parametric masking of randomness in 30 530 sequence alignments can be improved and leads to better resolved trees. Frontiers in 31 531 Zoology, 7(1), 10. 32 532 Lane, E. B. (1982). Monoclonal antibodies provide specific intramolecular markers for the 33 533 study of epithelial tonofilament organization. The Journal of Cell Biology, 92, 665– 34 534 673. 35 535 Littlewood, D. T. J., Rohde, K., Bray, R. A., & Herniou, E. A. (1999). Phylogeny of the 36 536 Platyhelminthes and the evolution of parasitism. Biological Journal of the Linnean 37 537 Society, 68, 257–287. 38 538 López-Giráldez, F., & Townsend, J. P. (2011). PhyDesign: an online application for 39 539 profiling phylogenetic informativeness. BMC Evolutionary Biology, 11(1), 152. 40 540 Lundin, K. (2000). Phylogeny of the Nemertodermatida (Acoelomorpha, Platyhelminthes). 41 541 A cladistic analysis. Zoologica Scripta, 29, 65–74. 542 Lundin, K. (2001). Degenerating epidermal cells in Xenoturbella bocki (phylum uncertain), 42 543 Nemertodermatida and Acoela (Platyhelminthes). Belgian Journal of Zoology, 131, 43 544 153–157. 44 545 Lundin, K., & Hendelberg, J. (1998). Is the sperm type of the Nemertodermatida close to 45 546 that of the ancestral Platyhelminthes? Hydrobiologia, 383, 197–205. 46 547 Meyer-Wachsmuth, I., Curini-Galletti, M., & Jondelius, U. (2014). Hyper-Cryptic Marine 47 548 Meiofauna: Species Complexes in Nemertodermatida. PLoS ONE, 9(9), e107688. 48 549 Misof, B., & Misof, K. (2009). A Monte Carlo approach successfully identifies 49 550 randomness in multiple sequence alignments : A more objective means of data 50 551 exclusion. Systematic Biology, 58(1), 21–34. 51 552 Paps, J., Baguñà, J., & Riutort, M. (2009). Bilaterian phylogeny: a broad sampling of 13 52 553 nuclear genes provides a new Lophotrochozoa phylogeny and supports a paraphyletic 53 554 basal Acoelomorpha. Molecular Biology and Evolution, 26(10), 2397–2406. 54 55 17 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 555 Philippe, H., Brinkmann, H., Copley, R. R., Moroz, L. L., Nakano, H., Poustka, A. J., et al. 8 556 (2011). Acoelomorph flatworms are deuterostomes related to Xenoturbella. Nature, 9 557 470(7333), 255–258. 10 558 Raikova, O. I., Reuter, M., Jondelius, U., & Gustafsson, M. K. S. (2000). An 11 559 immunocytochemical and ultrastructural study of the nervous and muscular system of 12 560 Xenoturbella westbladi (Bilateria inc. sed.). Zoomorphology, 120, 107–118. 13 561 Rambaut, A. (2009, July 15). Tree figure drawing tool V 1.3.1. tree.bio.ed.ac.uk. Retrieved 14 562 July 15, 2014, from http://tree.bio.ed.ac.uk/ 15 563 Riedl, R. (1983). Fauna und Flora des Mittelmeeres (3rd ed., pp. 210–211). Paul Parey. 16 564 Rieger, R. M., & Ladurner, P. (2003). The significance of muscle cells for the origin of 565 mesoderm in Bilateria. Integrative Comparative Biology, 43, 47–54. 17 566 Rieger, R. M. & Ott, J. (1971). Gezeitenbedingte Wanderungen von Turbellarien und 18 567 Nematoden eines nordadriatischen Sandstrandes. Troisième Symposium Europeén de 19 568 Biologie Marine, Supplément 21, 425-447. 20 569 Riser, N. W. (1987). Nemertinoides elongatus gen.n., sp.n. (Turbellaria: 21 570 Nemertodermatida) from coarse sand beaches of the Western North Atlantic. 22 571 Proceedings of the Helminthological Society in Washington, 54(1), 60–67. 23 572 Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D. L., Darling, A., Hohna, S., et al. 24 573 (2012). MrBayes 3.2: efficient bayesian phylogenetic inference and model choice 25 574 across a large model space. Systematic Biology, 61(3), 539–542. 26 575 Ruiz-Trillo, I., Paps, J., Loukota, M., Ribera, C., Jondelius, U., Baguñà, J., & Riutort, M. 27 576 (2002). A phylogenetic analysis of myosin heavy chain type II sequences corroborates 28 577 that Acoela and Nemertodermatida are basal bilaterians. Proceedings of the National 578 Academy of Sciences of the USA, 99(17), 11246–11251. 29 579 Shimodaira, H. (2002). An Approximately Unbiased test of phylogenetic tree selection. 30 580 Systematic Biology, 51(3), 492–508. 31 581 Shimodaira, H., & Hasegawa, M. (2001). Consel: for assessing the confidence of 32 582 phylogenetic tree selection. Bioinformatics, 17, 1246–1247. 33 583 Silvestro, D., & Michalak, I. (2011). raxmlGUI: a graphical front-end for RAxML. 34 584 Organisms Diversity & Evolution, 12(4), 335–337. 35 585 Stamatakis, A. (2006). Phylogenetic models of rate heterogeneity: a high performance 36 586 computing perspective, 1–8. 37 587 Stamatakis, A., Hoover, P., & Rougemont, J. (2008). A Rapid Bootstrap Algorithm for the 38 588 RAxML Web Servers. Systematic Biology, 57(5), 758–771. 39 589 Steinböck, O. (1930). Ergebnisse einer von E. Reisinger & O. Steinböck mit Hilfe des 40 590 Rask-Örsted Fonds durchgeführten Reise in Grönland 1926. 2. Nemertoderma 41 591 bathycola nov. gen. nov. spec. Videnskabelige Meddelelser Dansk Naturhistorisk 592 Forening, 90, 1–42. 42 593 Sterrer, W. (1968). Beiträge zur Kenntnis der Gnathostomulida. Arkiv för Zoologi, 1–1. 43 594 Sterrer, W. (1998). New and known Nemertodermatida-a revision. Belgian Journal of 44 595 Zoology, 128, 55–92. 45 596 Tajima, F. (1989). Statistical methods for testing the neutral Mutation hypothesis by DNA 46 597 polymorphism. Genetics, 123, 585–595. 47 598 Tajima, F. (1993). Unbiased estimation of evolutionary distance between nucleotide 48 599 sequences. Molecular Biology and Evolution, 10(3), 677–688. 49 600 Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., & Kumar, S. (2011). 50 601 MEGA5: Molecular Evolutionary Genetics Analysis using maximum likelihood, 51 602 evolutionary distance, and maximum parsimony methods. Molecular Biology and 52 603 Evolution, 28(10), 2731–2739. 53 604 Tang, C. Q., Leasi, F., Obertegger, U., Kieneke, A., Barraclough, T. G., & Fontaneto, D. 54 55 18 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 605 (2012). The widely used small subunit 18S rDNA molecule greatly underestimates true 8 606 diversity in biodiversity surveys of the meiofauna. Proceedings of the National 9 607 Academy of Sciences of the USA, 1–11. 10 608 Telford, M. J., Lockyer, A. E., Cartwright-Finch, C., & Littlewood, D. T. J. (2003). 11 609 Combined large and small subunit ribosomal RNA phylogenies support a basal 12 610 position of the acoelomorph flatworms. Proceedings of the Royal Society B: Biological 13 611 Sciences, 270(1519), 1077–1083. 14 612 Todt, C. (2009). Structure and evolution of the pharynx simplex in acoel flatworms 15 613 (Acoela). Journal of Morphology, 270(3), 271–290. 16 614 Wallberg, A., Curini-Galletti, M., Ahmadzadeh, A., & Jondelius, U. (2007). Dismissal of 615 Acoelomorpha: Acoela and Nemertodermatida are separate early bilaterian clades. 17 616 Zoologica Scripta, 36, 509–523. 18 617 Westblad, E. (1937). Die Turbellariengattung Nemertoderma Steinböck. Acta Societatis 19 618 pro Fauna et Flora Fennica, 60, 45–89. 20 619 Westblad, E. (1949). On Meara stichopi (Bock) Westblad, a new representative of 21 620 Turbellaria Archoophora. Arkiv för Zoologi, 1(5), 1–19. 22 621 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 19 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 622 Fig 1 Parts of alignments showing the transition in the chimeric sequences (shown in 8 623 grey). a LSU sequence AM747472 is registered as Flagellophora apelti in GenBank. The 9 10 624 5’-part is similar to gastrotrich species, the 3’-part can be identified as Meara stichopi. b 11 625 LSU sequence AM747478 registered as Nemertoderma bathycola in GenBank but the 3’- 12 626 part blasts as an unidentified platyhelminth species. c SSU sequence AM747471 is 13 627 registered as Flagellophora apelti in GenBank the 3’-part of the sequence shows the same 14 628 pattern as Meara stichopi. d SSU sequence AF051328 is registered as M. stichopi but the 15 16 629 5’-part BLASTs as copepod species 17 630 18 19 631 Fig 2 Phylogenetic informativeness profiles (net) of all three genes used in this study. The 20 632 X-axis indicates time, but is in absence of calibration points relative to node depths. The 21 633 profile is drawn over the ultrametric tree from which it was calculated 22 23 634 24 635 Fig 3 Majority rule consensus tree estimated using RAxML from the concatenated dataset. 25 26 636 Bootstrap support values and Bayesian posterior probabilities are plotted on those nodes 27 637 retrieved in both analyses. Coloured boxes correspond to genera of the ingroup, asterisks 28 638 indicate species downloaded from GenBank (ingroup only) 29 30 639 31 640 Fig 4 LSU (a) and SSU (b) gene trees of the final datasets plus the chimeric sequences. 32 641 The colours correspond to those in Fig. 3,, chimeric sequences are highlighted in greylight 33 34 642 blue. In both the LSU and SSU datasets the genera Flagellophora and Meara are pulled 35 643 artificially closer together by the chimeric sequences (cf. gene trees) 36 644 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 20 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 645 Table 1 Primers used in this study for sequencing of LSU, SSU and H3 markers. TimA 8 646 and TimB are outer primers spanning the length of the whole fragment, the primer pairs 9 10 647 S30/5FK and 4FB/1806R are internal primer for the 5’- and 3’-part, respectively 11 Gene Name dir. taxon sequence reference 12 Nemertoderma, 28S U178 for GCACCCGCTGAAYTTAAG Telford et al. 2003 Formatted: Font: Italic 13 Meara Nemertoderma, Formatted: Font: Italic, Font color: Black, Border: : L1642 rev CCAGCGCCATCCATTTTCA Telford et al. 2003 14 Meara (No border) Nemertoderma, 15 1200F for CCCGAAAGATGGTGAACTATGC Telford et al. 2003 Formatted: Font: Italic Meara 16 Nemertoderma, Formatted: Font: Italic, Font color: Black, Border: : R2450 rev Telford et al. 2003 17 Meara (No border) Nemertoderma, TAAGGGAAGTCGGCAAATTAGATCC 18 UJ2176 for Wallberg et al 2007 Formatted: Font: Italic Meara G 19 Nemertoderma, Formatted: Font: Italic, Font color: Black, Border: : L3449 rev ATTCTGACTTAGAGGCGTTCA Telford et al. 2003 20 Meara (No border) Nemertoderma, U1846 for AGGCCGAAGTGGAGAAGG Telford et al. 2003 Formatted: Font: Italic 21 Meara 22 Nemertoderma, Formatted: Font: Italic, Font color: Black, Border: : L2984 rev (No border) 23 Meara 28SP1F5 Nemertinoides, Meyer-Wachsmuth et al. for CTGAGAAGGGTGTGAGACCCGTAC Formatted: Font: Italic 24 Ster Sterreria 2014 28SP1R1 Nemertinoides, Meyer-Wachsmuth et al. Formatted: Font: Italic, Font color: Black, Border: : 25 rev TCCCGTAGATCCGATGAGCGTC Ster Sterreria 2014 (No border) 26 Norén & Jondelius 18S TimA for all AMCTGGTTGATCCTGCCAG Formatted: Font: Italic 27 1999 Norén & Jondelius Formatted: Font: Italic, Font color: Black, Border: : TimB rev all TGATCCATCTGCAGGTTCACCT 28 1999 (No border) 29 all except S30 for GCTTGTCTCAAAGATTAAGCC Norén & Jondelius 1999 Formatted: Font: Italic 30 Flagellophora 18SF1_ Formatted: Font: Italic, Font color: Black, Border: : for Flagellophora this publication 31 apelti (No border) Norén & Jondelius 32 5FK rev all TTCTTGGCAAATGCTTTCGC Formatted: Font: Italic 1999 33 all except Norén & Jondelius Formatted: Font: Italic, Font color: Black, Border: : 4FB for CCAGCAGCCGCGGTAATTCCAG 34 Flagellophora 1999 (No border) 4FB 35 for Flagellophora CGCTCGTAGTTGGATCTCTCG this publication Formatted: Font: Italic apelti 36 Norén & Jondelius Formatted: Font: Italic, Font color: Black, Border: : 1806R rev all CCTTGTTACGACTTTTACTTCCTC 37 1999 (No border) 38 H3 H3 AF for all ATGGCTCGTACCAAGCAGACVGC Colgan et al. 1998 Formatted: Font: Italic 39 H3 AR rev all ATATCCTTRGGCATRATRGTGAC Colgan et al. 1998 Formatted: Font: Italic, Font color: Black, Border: : Meyer-Wachsmuth et al. (No border) H3FNem for all ATGGCTCGTACCAAGCAGACG 40 2014 Formatted: Font: Italic Meyer-Wachsmuth et al. 41 H3RNem rev all TCCTTGGGCATGATGGTGAC 42 2014 Formatted: Font: Italic 648 Formatted: Font: Italic, Font color: Black, Border: : 43 (No border) 649 44 Formatted: Font: Italic 45 Formatted: Font: Italic 46 Formatted: Font: Italic, Font color: Black, Border: : 47 (No border) 48 49 50 51 52 53 54 55 21 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 650 Table 2 Sequences used in the phylogenetic analyses, rows indicate how GenBank 8 651 sequences were combined for concatenated analyses. Asterisks indicate new previously 9 10 652 published sequences published first in this studydownloaded from GenBank, † chimeric 11 653 sequences and # sequences which that cluster with a different species than specified in 12 654 GenBank 13 14 species accession number

15 LSU SSU H3 16 outgroup 17 Diopisthoporus longitubus (06006) FR837775.1 FR837692.1 18 Paraphanostoma submaculatum AJ849506.1 Paratomella rubra (11085) *KT698949 *KT698957 19 Chaetoderma nitidulum AY377658.1 AY377763.1 20 Chaetoderma sp. AY145397.1 21 Dumetrocrinus antarcticus KC616753.1 22 Balanoglossus clavigerus 1/FN908667.1 23 Ptychodera flava 2/AF212176.1 24 Ptychodera bahamensis 2/AF236802.1 25 Saccoglossus kowalevskii 1/AF212175.1 26 Lepidopleurus cajetanus FJ445776.1 AF120502.1 KF527282.1 27 ingroup 28 Ascoparia sp. (AWHel-24) FR837762.1 FR837678.1 29 Flagellophora apelti †AM747472.1 †AM747471.1 30 Meara stichopi (03052) FR837797.1 FR837714 31 Meara stichopi (1) AY157605 AF100191.2 32 Meara stichopi (2) AY218125 †AF051328 AY218147.1 33 Meara stichopi (3) AM747474.1 AM747473.1 34 Nemertinoides elongatus (1) AF021326 AY078381.1 35 Nemertinoides elongatus (2) #AM747476.1 #AM747475.1 36 (S. rubra) (S. psammicola) Nemertoderma bathycola (1) AF327725.1 37 Nemertoderma bathycola (2) †AM747478.1 AM747477.1 38 Nemertoderma sp. (06016) FR837800.1 FR837717.1 39 Nemertoderma westbladi (1) AM747482.1 AM747481.1 40 Nemertoderma westbladi (2) AF327726.1 41 Sterreria psammicola #AM747480.1 #AM747479.1 42 (S. rubra) (S. rubra) 43 Sterreria psammicola (07006) KM062708 KM062542 44 Nemertinoides wolfgangi (08096) KM062733 KM062568 KM194628 45 Nemertinoides sp. N3 (08098) KM062734 KM062569 KM194629 46 Nemertinoides sp. N1 (08102) KM062736 KM062571 47 Nemertinoides sp. N2 (08123) KM062748 KM062583 48 Sterreria ylvae (10058) KM062773 KM062607 KM194650 49 Sterreria martindalei (10060) KM062774 KM062608 KM194651 50 Sterreria lundini (10076) KM062779 KM062613 51 Nemertoderma bathycola (12116) *KT698950 *KT698958 52 Nemertoderma westbladi (12122) *KT698951 *KT698959 *KT698952 *KT698960 53 Nemertoderma bathycola (12220) 54 55 22 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 Nemertoderma westbladi (12222) *KT698953 *KT698961 *KT698965 8 Nemertoderma sp. (12231) *KT698954 *KT698962 *KT698966 9 Flagellophora apelti (13110) *KT698955 *KT698963 *KT698967 10 Sterreria rubra (13148) KM062809 KM062643 KM194672 11 Sterreria psammicola (13155) KM194673 12 Sterreria variabilis (13156) KM062811 KM062645 KM194674 13 Sterreria sp. S2 (13157) KM062812 KM062646 KM194675 14 Nemertinoides elongatus (13180) KM062816 KM062649 KM194677 15 Nemertinoides glandulosum (13181) KM062817 KM062650 KM194678 16 Nemertoderma sp. (13330) *KT698956 *KT698964 *KT698968 17 Meara stichopi (A) KM062846 KM062678 KM194693 18 Nemertinoides sp. N4 (MCG09) KM062839 KM062672 KM194689 19 Sterreria sp. S3 (MCG12) KM062842 KM062675 20 Sterreria papuensis (PNG52) KM062853 KM062685 KM194696 KM062858 KM062690 KM194699 21 Sterreria sp. P3 (PNG60) KM062861 KM062693 KM194703 22 Sterreria sp. S7 (PNG67) Sterreria monoliths (PNG84) KM062870 KM062702 KM194710 23 Sterreria boucheti (PNG87) KM062872 KM062704 KM194712 24 655 25 26 656 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 23 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 657 Table 3 Values of parameters estimated with jModeltest based on BIC for the proportion 8 658 of invariable sites (I) and gamma distribution of rates across sites (G) as well as their 9 10 659 importance. The priors used for MrBayes analyses based on the given estimates are shown 11 I importance G importance MrBayes 12 (I/I+G) (G/I+G) 13 LSU 0.2250 0/1.0 0.5269 0/1.0 shapepr=Uniform(0.10,0.70) 14 pinvarpr=Uniform(0.01,0.5) 15 SSU 0.2957 0/1.0 0.4802 0/1.0 shapepr=Uniform(0.10,0.70) 16 pinvarpr=Uniform(0.01,0.5) 17 H3 0.5181 0/1.0 1.0902 0/1.0 shapepr=Uniform(0.40,1.30) 18 pinvarpr=Uniform(0.01,0.8) 19 660 20 661 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 24 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 662 Table 4 Observed nucleotide frequencies and GC-content averaged per lineage for all three 8 663 genes used in this study 9 10 T C A G GC-content 11 LSU 12 Sterreria 21.58 23.56 23.08 31.77 55.34 Formatted: Font: Italic 13 Nemertinoides 21.26 24.36 22.51 31.87 56.23 Formatted: Font: Italic 14 Meara 20.60 24.12 22.63 32.65 56.77 Formatted: Font: Italic 15 Nemertoderma 23.55 22.49 23.93 30.03 52.52 Formatted: Font: Italic 16 Flagellophora 20.7 25.3 21.6 32.4 57.8 Formatted: Font: Italic 17 Ascoparia 20.5 26.3 20.8 32.4 58.7 Formatted: Font: Italic 18 SSU Formatted: Font: Italic 19 Sterreria 24.44 22.46 25.25 27.86 50.32 Formatted: Font: Italic 20 Nemertinoides 21.8 24.1 22.8 31.3 50.23 Formatted: Font: Italic 21 Meara 22.3 24.3 23.0 30.5 51.6 Font: Italic 22 Nemertoderma 25.38 21.53 25.92 27.17 48.7 Formatted: 23 Flagellophora 24.1 24.3 23.7 27.9 52.2 Formatted: Font: Italic 24 Ascoparia 24.5 23.6 25.3 26.6 50.2 Formatted: Font: Italic 25 H3 Formatted: Font: Italic 26 Sterreria 19.0 28.7 26.9 25.4 54.1 Formatted: Font: Italic 27 Nemertinoides 17.5 28.8 27.1 26.5 55.3 Formatted: Font: Italic 28 Meara 17.1 28.4 29.5 25.0 53.4 Formatted: Font: Italic 29 Nemertoderma 21.9 28.9 24.3 24.9 53.8 Formatted: Font: Italic 30 Flagellophora 26.8 26.8 20.4 26.0 52.8 31 Formatted: Font: Italic Ascoparia X X X X X Formatted: Font: Italic 32 664 33 Formatted: Font: Italic 34 665 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 25 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 666 Table 5 Results of the Approximately Unbiased test for the comparison of topologies 8 667 performed by the program Consel, including other statistical tests of tree comparison. The 9 10 668 best tree inferred from the concatenated dataset without constraint was compared to the 11 669 best trees calculated from the same dataset with alternative topological constraints. (AU: 12 670 Approximately Unbiased test; BP: Bootstrap probability of the selection; PP: Bayesian 13 671 Posterior Probability calculated with the Bayesian Information Criterion) 14 15 topological constraint AU BP PP 16 no constraint 0.942 0.905 1.0 17 non-monophyly of Nemertodermatida 0.008 0.006 0 18 Meara + Nemertoderma (Lundin 2000) 0 0 0 Formatted: Font: Italic 19 Meara + Flagellophora (Wallberg et al. 2007) 0 0 0 Formatted: Font: Italic 672 20 Formatted: Font: Italic 21 673 Formatted: Font: Italic 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 26 56 57 58 59 60 61 62 63 64 65 1 2 3 4 5 6 7 674 Online Resource 1 Observed nucleotide frequencies and GC-content per specimen for all 8 675 three genes used in this study and the codon positions of the protein-coding gene Histone 9 10 676 3. Species names highlighted in red have greatly varying sequence lengths, which may 11 677 have affected the observed nucleotide bias 12 678 13 14 679 Online Resource 2 The estimates of the disparity index for all three genes. The Disparity 15 680 Indices per-site are shown for each sequence pair above the diagonal. Below, the p-values 16 681 of the Disparity Indices are shown, those smaller than 0.05 are considered significant. 17 18 682 Evolutionary analyses were conducted in MEGA5.2.2. 19 683 20 21 684 Online Resource 3 Comparison of the topologies of trees estimated from the original (left) 22 685 and Aliscore-reduced (right) SSU datasets. The differing nodes are marked 23 24 686 25 687 Online Resource 4 Saturation plots for all three genes used in this study and for the codon 26 688 positions of H3. Histone 3, and especially its 3rd codon position, show patterns indicating 27 28 689 saturation 29 690 30 31 691 Online Resource 5 Gene trees for LSU, SSU and H3. Maximum Likelihood (ML) trees 32 692 were estimated with RAxML and Bayesian inferences (MB) were performed with 33 693 MrBayes. The colours correspond to genera, and in case of Ascopariidae to the family. a) 34 694 ML LSU. b) MB LSU. c) ML SSU. d) MB SSU. e) ML H3. f) MB H3 35 36 695 37 696 Online Resource 6 Pairwise distances across the a) LSU, b) SSU and c) Histone 3 gene 38 39 697 datasets 40 698 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 27 56 57 58 59 60 61 62 63 64 65 Click here to download Figure: Fig1_rev_3.jpg Click here to download Figure: Fig2_rev.jpg Click here to download Figure: Fig3_rev.tiff Click here to download Figure: Fig4_rev.jpg Click here to download Supplementary Material: OR1.xls