bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Harald Gruber-Vodicka 2 Max Planck Institute for Marine Microbiology, Celsiusstrasse 1; 28359 Bremen, 3 Germany, +49 421 2028 825, [email protected] 4 5 Morphology of obligate ectosymbionts reveals Paralaxus gen. nov., a

6 new circumtropical genus of marine stilbonematine nematodes 7 8 Florian Scharhauser1*, Judith Zimmermann2*, Jörg A. Ott1, Nikolaus Leisch2 and Harald 9 Gruber-Vodicka2 10 11 1Department of Limnology and Bio-Oceanography, University of Vienna, Althanstrasse 12 14, A-1090 Vienna, Austria 2 13 Max Planck Institute for Marine Microbiology, Celsiusstrasse 1, D-28359 Bremen, 14 Germany 15

16 *Contributed equally 17 18 19 20 Keywords: Paralaxus, thiotrophic symbiosis, systematics, ectosymbionts, molecular 21 phylogeny, cytochrome oxidase subunit I, 18S rRNA, 16S rRNA, 22 23 24 Running title: “Ectosymbiont morphology reveals new nematode genus“ 25 Scharhauser et al.

26

27 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 2

28 Abstract

29 Stilbonematinae are a subfamily of conspicuous marine nematodes, distinguished by a 30 coat of sulphur-oxidizing bacterial ectosymbionts on their cuticle. As most nematodes, 31 the worm hosts have a simple anatomy and few taxonomically informative characters, 32 and this has resulted in numerous taxonomic reassignments and synonymizations. 33 Recent studies using a combination of morphological and molecular traits have helped 34 to improve the taxonomy of Stilbonematinae but also raised questions on the validity 35 of several genera. Here we describe a new circumtropically distributed genus 36 Paralaxus (Stilbonematinae) with three species: Paralaxus cocos sp. nov., 37 P. bermudensis sp. nov. and P. columbae sp. nov.. We used single worm metagenomes 38 to generate host 18S rRNA and cytochrome oxidase I (COI) as well as symbiont 39 16S rRNA gene sequences. Intriguingly, COI alignments and primer matching analyses 40 suggest that the COI is not suitable for PCR-based barcoding approaches in 41 Stilbonematinae as the genera have a highly diverse base composition and no 42 conserved primer sites. The phylogenetic analyses of all three gene sets however 43 confirm the morphological assignments and support the erection of the new genus 44 Paralaxus as well as corroborate the status of the other stilbonematine genera. 45 Paralaxus most closely resembles the stilbonematine genus Laxus in overlapping sets 46 of diagnostic features but can be distinguished from Laxus by the morphology of the 47 genus-specific symbiont coat. Our re-analyses of key parameters of the symbiont coat 48 morphology as character for all Stilbonematinae genera show that with amended 49 descriptions, including the coat, highly reliable genus assignments can be obtained. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 3

50 Introduction

51 The identification of many nematode genera and species is difficult based on 52 morphological characters alone (Sudhaus & Kiontke 2007, de-León & Nadler 2010, 53 Derycke et al. 2008, Palomares-Rius et al. 2014). A prime example for this problem are 54 the Stilbonematinae, a subfamily of the Desmodoridae that have experienced several 55 changes in the classification of species and even genera in the past (Tchesunov 2013; 56 Ott et al. 2014). The exclusively marine Stilbonematinae are common members of the 57 interstitial meiofauna in sheltered intertidal and subtidal porous sediments and have 58 been found worldwide (Ott et al. 2004, Tchesunov et al. 2013). Their diversity and 59 abundances are highest in subtropical and tropical shallow-water sands, but some 60 species were also described from higher latitudes (e.g. Gerlach 1950, Wieser 1960, 61 Platt & Zhang 1982, Riemann et al. 2003, Tchesunov 2012), deep-sea sediments (Van 62 Gaever et al. 2004, Tchesunov et al. 2012, Leduc 2013), near shallow hydrothermal 63 vents (Kamenev et al. 1993, Thiermann et al. 1997) and methane seeps (Dando et al. 64 1995). 65 To date, a total of 10 different stilbonematine nematode genera and more than 50 66 different species are described (reviewed by Tchesunov 2013; Leduc 2013, Armenteros 67 et al. 2014b, Ott et al. 2014a,b, Leduc and Sinniger 2018). Most species and genera still 68 lack molecular data, but recent taxonomic work has started to integrate molecular 69 data using the 18S rRNA gene (18S) (Ott et al. 2014a, b; Armenteros et al. 2014ab; 70 Leduc and Zhao 2016, Leduc and Sinniger 2018) and the mitochondrial cytochrome 71 oxidase subunit I (COI) gene (Armenteros et al. 2014a, b). The phylogenetic signal of 72 the two marker genes was however not consistent and called morphological genus and 73 species identifications into question (Armenteros et al. 2014b). While the monophyly 74 of the subfamily Stilbonematinae has strong support based on 18S data (Kampfer et al. 75 1998, Bayer et al. 2009, van Megen et al. 2009, Ott et al 2014a and b; Leduc & Zhao 76 2016), it was questioned in a recent study by Armenteros et al. (2014b) which used 77 both 18S and COI. 78 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 4

79 A trait that unites all stilbonematine nematodes is the conspicuous ectosymbiosis with 80 sulphur-oxidizing Gammaproteobacteria that cover major parts of the cuticle 81 (reviewed by Ott et al. 2004). Stilbonematine symbionts have diverse morphologies, 82 including rod, coccus, coccobacillus, filament, corn-kernel or crescent shapes, and each 83 host species harbours a single morphotype (Ott & Novak 1989, Polz et al. 1992; Polz et 84 al. 1994; Ott et al., 2004; Bayer et al. 2009, Pende et al 2014, Ott et al. 2014a, b). In 85 phylogenetic analyses of the symbiont 16S rRNA gene (16S) sequences, all 86 stilbonematine symbionts fall into a monophyletic clade of Gammaproteobacteria 87 related to Chromatiaceae, which has been described as Candidatus Thiosymbion (from 88 here on ‘Thiosymbion’) (Bayer et al. 2009, Bulgheresi et al. 2011, Heindl et al. 2011, 89 Zimmermann et al. 2016). Based on phylogenetic analyses of host 18S and symbiont 90 16S sequences closely related stilbonematine nematode species of the same genus 91 consistently harbour closely related and genus specific ‘Thiosymbion’ symbionts (Ott et 92 al. 2004, Zimmermann et al. 2016). ‘Thiosymbion’ encompasses not only the 93 ectosymbionts of stilbonematine nematodes, but also the primary symbionts of gutless 94 phallodriline oligochaetes and siphonolaimid nematodes (Zimmermann et al. 2016). 95 Comparative analyses of stilbonematine nematodes and their symbionts showed a 96 high degree of phylogenetic congruence that indicates a long-term co-diversification 97 between ‘Thiosymbion’ and the Stilbonematinae (Zimmermann et al. 2016). 98 Recent 18S-based phylogenetic analyses suggested a novel, morphologically 99 uncharacterized stilbonematine genus from the Caribbean Sea (Belize) and the 100 Australian Great Barrier Reef (Heron Island) (Zimmermann et al. 2016). The genus level 101 of this clade was supported by the distinct 16S sequences of their symbionts that also 102 formed a new phylogenetic clade in the study. Here we combine morphological and 103 molecular data to describe this recently detected host clade as ‘Paralaxus’ gen. nov.. 104 We use single worm metagenomes to provide host 18S and COI gene sequences as 105 well as symbiont 16S sequences from nine Paralaxus specimens, and 20 additional 106 specimens covering six stilbonematine genera. We place Paralaxus in a unified 107 taxonomy of Stilbonematinae and formally describe the new genus Paralaxus with a 108 Caribbean type species P. cocos sp. nov. In addition, we describe two new species, bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 5

109 P. bermudensis sp. nov. and P. columbae sp. nov. from Bermuda and Florida, and 110 report on members of this new genus from localities in the Western and Central Pacific 111 Ocean. While characterizing the key anatomical features of Paralaxus, we identified 112 the high taxonomic value of its symbiont coat. We extend this concept and evaluate 113 coat and symbiont morphology as additional traits for all stilbonematine genera. 114 In addition, we substantially improve the molecular dataset of Stilbonematinae full- 115 length 18S and COI marker gene sequences to answer the following questions: (1) Are 116 Stilbonematinae monophyletic? (2) At which taxonomic level are 18S and COI valuable 117 phylogenetic markers for Stilbonematinae? (3) Can we find suitable COI PCR priming 118 sites that are conserved across the subfamily? 119 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 6

120 Material and Methods

121 Meiofauna collection, preparation and microscopy

122 Sediment samples were collected in the Caribbean Sea (Belize), the Western Atlantic 123 Ocean (Florida and Bermuda), the Western and Central Pacific Ocean (Australia and 124 Hawaii) as well as the North and Mediterranean Sea (Fig. 1 and S1) using cores or 125 buckets. Nematodes were extracted by decantation through a 64 µm mesh sieve and 126 sorted live under a dissecting microscope. Specimens were fixed in 4% formaldehyde 127 for taxonomic preparations or in 2.5% glutaraldehyde in 0.1 M sodium cacodylate

128 buffer and were post fixed in 2% OsO4 (for scanning electron microscopy, SEM) and 129 stored at 4 °C until further analyses. Before fixation, live specimens for sequencing 130 were photographically vouchered in the field i.e. micrographs were taken of the full 131 specimen mounted in seawater, capturing the head region with mouth, amphids and 132 pharynx, the anal and tail region, the symbiont coat, and male and female 133 reproductive organs. After imaging, specimens were washed in 0.2 µm filtered 134 seawater and immediately fixed in either RNAlater® stabilization solution (Ambion, 135 Foster City, US), 70% ethanol or pure methanol and stored at 4° C until the molecular 136 analyses. 137 For detailed light microscopy, the formaldehyde or glutaraldehyde-fixed specimens 138 were transferred into glycerol: water 1:9, slowly evaporated and finally mounted in 139 pure glycerol. Drawings were made using a camera lucida on a Diavar microscope 140 (Reichert, Vienna, Austria). Nomarsky interference contrast photos were taken on a 141 Polyvar (Reichert, Vienna, Austria). Specimens for SEM were critical point dried and 142 coated with palladium using a JEOL JFC-2300HR sputter coater (JEOL Ltd., Akishima, 143 Japan) and examined on a JEOL IT 300 (JEOL Ltd., Akishima, Japan) at high vacuum 144 mode. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 7

145 DNA extraction, marker gene amplification and metagenome sequencing

146 DNA from single nematodes including their attached ectosymbionts was extracted 147 with the DNAeasy Blood and Tissue Micro Kit (Qiagen, Hilden, Germany) following 148 manufacturer’s instructions with two amendments to the protocol – the proteinase K 149 digestion was extended to at least 24 h and the elution step was performed twice with 150 20 µl of elution buffer. DNA concentrations were quantified using a Qubit 2.0 and the 151 high sensitivity assay (Life Technologies, Carlsbad, US). 152 To characterize nematode hosts prior to metagenome sequencing we used both the 153 ribosomal 18S and mitochondrial COI marker genes for PCR-based screenings. To 154 amplify an approximately 600 bp long fragment of the nematode 18S we designed a 155 nematode-specific primer pair: 3FNem 5'-GTTCGACTCCGGAGAGGGA-3' and 5R 5'- 156 CTTGGCAAATGCTTTCGC-3'. Amplification of the mitochondrial COI gene of several 157 stilbonematine nematode genera including Paralaxus was conducted using the 158 published primer pair 1490F 5’-GGTCAACAAATCATAAAGATATTGG-3’ and 2198R 5’- 159 TAAACTTCAGGGTGACCAAAAATCA-3’ (Folmer et al. 1994). For 18S PCRs we used the 160 Phusion® DNA polymerase, HF buffer (Finnzymes, Finland) and the following cycling 161 conditions: 95° C for 5 min, followed by 38 cycles of 95° C for 1 min, 55° C for 1.5 min, 162 72° C for 2 min and followed by 72° C for 10 min. COI PCRs were conducted using the

163 TaKaRa Taq™ DNA polymerase, 10x reaction buffer (Takara Bio Inc., Japan) and the 164 following cycling conditions: 5 min at 95° C, then 36 cycles of 1 min at 95°C, 1.5 min at 165 42° C, and 2 min at 72° C, followed by 72° C for 10 min. While the 18S amplifications 166 were successful for all nematode specimens, we could not obtain a single PCR product 167 for the COI amplifications of any stilbonematine nematode specimen. In contrast, DNA 168 extracts from Lamellibrachia tubeworms (Siboglinidae, Annelida) that were used as

169 positive control always resulted in COI PCR products of the expected size. 170 We performed single specimen shotgun metagenome sequencing to generate data 171 from the host nuclear genome, host mitochondria and symbionts using Illumina short- 172 read technology. For Illumina sequencing 1 - 5 ng DNA per specimen were used for an 173 Ovation Ultralow Library Systems kit (NuGEN Technologies, San Carlos, US). Illumina 174 library preparation including size-selection on an agarose gel was performed at the bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 8

175 Max Planck Genome Centre in Cologne, Germany. 2×100 bp paired-end reads were 176 sequenced on an Illumina HiSeq3000 (San Diego, US) for each individual (8 - 25 million 177 reads each). We intended to extract the triple gene sets (symbiont 16S, host 18S and 178 COI) for all specimens but differing assembly qualitites due to the very low DNA input 179 did not allow us to retrieve all marker genes for each worm (Table S1). Host 18S and 180 symbiont 16S gene sequences were generated from the raw reads with phyloFlash 181 (Gruber-Vodicka et al. 2017, https://github.com/HRGV/phyloFlash). The COI was 182 assembled using the following approach: In brief, after removing adapters and low- 183 quality bases (Q < 2) with bbduk (Bushnell 2014), we retained all reads with a minimal 184 length of 36 bp after quality trimming and conducted combined host and symbiont 185 draft genome assemblies for all samples with SPAdes 3.1 – 3.9 (Bankevich et al. 2012. 186 The contig containing the COI gene was then extracted from each assembly using 187 BLAST 2.26+ searches with the available Stilbonematinae COI sequences as 188 implemented in the Geneious software v. 11.1 (54) (Biomatters, New Zealand). Full- 189 length COI genes were then predicted from the contigs using the Geneious gene 190 prediction tool.

191 COI compositional analyses and primer matching

192 All compositional and primer matching analyses were conducted with the 27 full- 193 length stilbonematine nematode COI sequences we derived from the metagenomics 194 data. Nucleotide composition was analyzed in Geneious. A primer mismatch analysis of 195 the two COI primer sets: 1490F: 5'-GGT CAA CAA ATC ATA AGA TAT TGG-3' and 2198R: 196 5'-TAA ACT TCA GGG TGA CCA AAA AAT CA-3' (Folmer et al. 1994) and JB3F: 5’-TTT TTT 197 GGG CAT CCT GAG GTT TAT-3’ and JB5 R: 5’-AGC ACC TAA ACT TAA AAC ATA ATG AAA 198 ATG-3’ used by Armenteros et al. (2014 a,b), was performed using the MOTIF search 199 option integrated in Geneious, with a maximum mismatch rate range from 1 - 7 for 200 every primer. The mismatches were mapped on a MAFFT v.7 (Katoh & Standley 2013) 201 alignment. The design of a new stilbonematine-specific COI primer set was conducted 202 with Primer 3 (Untergasser et al. 2012). In brief, the consensus sequence of our full- 203 length COI MAFFT alignment (1 bp to 1577 bp) was screened for a stilbonematine bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 9

204 nematodes-specific primer set by applying Tm calculation after SantaLucia (1998) and 205 the following parameters: primer size range 18 to 27 bp, melting temperature (Tm) 206 range 57° C to 63° C and a GC percentage range 20 to 80. In addition, separate 207 searches were conducted for the regions 1 bp to 750 bp and 750 bp to 1577 bp to 208 facilitate the detection of primer pairs that cover similar ranges as the existing primer 209 sets that only amplify parts of the gene.

210 Phylogenetic analyses of host and symbiont genes

211 For the host 18S-based phylogenetic reconstruction we used 23 full-length sequences 212 extracted from the single worm metagenomes together with 55 previously published 213 and non-chimeric stilbonematine nematode 18S sequences longer than 1300 bp 214 available in GenBank. All published non-stilbonematine nematode sequences of the 215 order Desmodorida and Chromadorida sequences longer than 1300 bp available in 216 10/2018 served as outgroup. 217 In an extended 18S dataset with a total of 250 sequences, we included eight partial 218 18S sequences of five stilbonematine species from Armenteros et al. (2014,b) and all 219 non-chimeric 18S sequences of Desmodorida and Chromadorida with a minimum 220 length of 700 bp available in GenBank. 221 For the host COI phylogeny, we constructed a COI matrix using the 27 full-length 222 stilbonematine COI sequences we extracted from the metagenomics data of nine 223 Paralaxus specimens, together with sequences of 18 specimens from six different 224 stilbonematine genera (Catanema, Eubostrichus, Laxus, Leptonemellla, Robbea and 225 Stilbonema). We added the full-length COI sequence from Baylisascaris procyonis 226 (JF951366) as outgroup. All 28 nucleotide COI sequences were translated into protein 227 sequences in Geneious using the invertebrate mitochondrial translation Table 5. In an 228 extended COI matrix, we included partial COI sequences of the five species from 229 Armenteros et al. (2014b) that were also included in the extended 18S dataset. Due to 230 the high morphological similarity between Paralaxus cocos sp. nov. and Leptonemella 231 brevipharynx (Armenteros et al. 2014b) the partial COI sequence of L. brevipharynx 232 was also included in the analysis. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 10

233 For the symbiont phylogenetic reconstruction, we used the 21 new 16S sequences that 234 we extracted from the metagenomes as well as the 54 ‘Thiosymbion’ 16S genes 235 sequences available in GenBank. We also included sequences of closely related 236 uncultured Gammaproteobacteria (JF344100, JF344607, JF344324), and used five 237 cultured free-living gammaproteobacterial sulfur-oxidizing bacteria from the 238 Chromatiaceae (Allochromatium vinosum strain DSM180, Marichromatium 239 purpuratum 984, Thiorhodococcus drewsii AZ1, Thiocapsa marina 5811, 240 Thiorhodovibrio sp. 970) as an outgroup. 241 The host and symbiont ribosomal rRNA sequences were independently aligned using 242 MAFFT v7 (Katoh & Standley 2013) with the Q-INS-I mode (Katoh & Toh 2008) as it 243 considers the predicted secondary structure of the RNA. All COI sequences were 244 translated into AA sequences and aligned in MAFFT v7 with the E-INS-i mode. The 245 optimal substitution model for each alignment was assessed using ModelFinder 246 (Kalyaanamoorthy et al., 2017). The mtZOA+F+G4 model was the best fit for the COI 247 alignment, TIM3e+I+G4 for the 18S alignment and TIM+F+I+G4 for the 16S alignment. 248 Phylogenetic trees were reconstructed using the maximum likelihood-based software 249 IQTREE (Nguyen et al. 2015) utilizing the Ultrafast Bootstrap Approximation UFBoot 250 (Minh et al., 2013) to assess node stability (10000 bootstrap runs). In addition to 251 maximum likelihood, support values were generated using approximate Bayes 252 (aBayes) (Anisimova et al. 2011) and SH-aLRT analyses (Guindon et al., 2010).

253 Data accessibility

254 Sequences from this study were submitted to the European Nucleotide Archive (ENA) 255 via GFBio data submission (Diepenbroek at el. 2014; https://www.gfbio.org) and are 256 available under the accession number PRJEB27096. 257 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 11

258 Results

259 Descriptions

260 We analyzed 33 Paralaxus gen. nov. specimens from seven locations in the Pacific, the 261 Atlantic and the Caribbean using light- and electron microscopy as well as molecular 262 analyses based on marker genes extracted from single specimen metagenomes. 263 264 Paralaxus gen. nov. 265 ZooBank registration (http://zoobank.org): urn:lsid:zoobank.org:act:417691DB- 266 5EE3-49B7-B838-1BBF1569235F 267 GenBank entries for 18S rRNA and COI sequences: see Table S1. 268 We erect a new genus for this clade with the following characters: cephalic capsule 269 without block layer (Fig. 2b, e, i, l, m); amphidial fovea spiral, no sexual dimorphism in 270 shape of fovea (Fig. 2b, g, i, l, m); male tail with velum (Fig. 2c, j, n); symbiotic bacteria 271 coccobacilli, arranged as multi-layered coat covering body except for anterior and 272 posterior end (Fig. 1a, b). All Paralaxus sequences form statistically supported cIades in 273 phylogenetic analyses of the 18S as well as the COI datasets (Fig. 3a and b). The 274 specimens for each Paralaxus species clustered together in both gene sets and clearly 275 separated each Paralaxus species from each other (Fig. 3a and b). 276 277 We describe three new species, and report on two additional ones. 278 279 Paralaxus cocos sp. nov. 280 ZooBank registration (http://zoobank.org): urn:lsid:zoobank.org:act A659918E- 281 FD88-4584-8115-CD4C396AE400 282 GenBank entries for 18S rRNA and COI sequences: see Table S1. 283 284 Species from Belize, Central America can be identified by an amphidial fovea with one 285 and a half turns and the bulbus occupies 32 - 35% of the pharynx length. The velum of bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 12

286 male specimens makes up 43 - 47% of the tail length and the spicula are moderately 287 arcuate (Fig. 2a-g). 288 289 Paralaxus bermudensis sp. nov. 290 ZooBank registration (http://zoobank.org): urn:lsid:zoobank.org:act: 8697C6A0- 291 3508-4833-81DC-D2A86B0B00AC 292 GenBank entries for 18S rRNA and COI sequences: see Table S1. 293 294 Species from Bermuda can be identified by an amphidial fovea with two turns and the 295 bulbus occupies 30 - 32% of the pharynx length. In male specimens the velum makes 296 up 30 – 40% of the tail length and the spicula are strongly arcuate (Fig. 2h-j). 297 298 Paralaxus columbae sp. nov. 299 ZooBank registration (http://zoobank.org): urn:lsid:zoobank.org:act:4BABABB3- 300 1DD4-46F1-A69C-BC06C1271D01 301 302 Species from Florida Keys, USA can be identified by an amphidial fovea with 1– 1.2 303 turns and the bulbus occupies 29 - 33% of the pharynx length. In male specimens the 304 velum makes up 35 - 37% of tail length and the spicula are weakly arcuate (Fig. 2k-0). 305 306 Paralaxus species from Australia and Hawaii 307 GenBank entries for 18S rRNA and COI sequences see Table S1. 308 309 Representatives of two additional species, Paralaxus sp. “heron 1” and P. sp. “oahu 1” 310 have been found off Heron Island, Great Barrier Reef, Australia and Oahu, Hawaii 311 Islands. Due to the lack of properly preserved material, a formal description is 312 currently not possible. Morphological details and measurements taken from 313 microphotographs of live animals prior to fixation are presented in the Supplementary 314 Information (Table S2). 315 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 13

316 Detailed Paralaxus species descriptions are presented in the Supplementary 317 Information (Supplementary Results and Figs. S1 – S8). 318

319 Key to the species: 320 Amphidial fovea with 2 turns; bulbus 30 - 32% of pharynx length, velum 30 – 40% of 321 tail length, spicula strongly arcuate Paralaxus bermudensis sp. nov. 322 Amphidial fovea with 1.5 turns, bulbus 32 - 35% of pharynx length, velum 43 - 47% of 323 tail length, spicula moderately arcuate P. cocos sp. nov. 324 Amphidial fovea with 1 – 1.2 turns, bulbus 29 - 33% of pharynx length, velum 35 - 37% 325 of tail length, spicula weakly arcuate P. columbae sp. nov.

326 Morphological delineation of the genera Paralaxus, Laxus and

327 Leptonemella

328 Most characters that identify the genus Paralaxus are shared with one or more genera 329 of the Stilbonematinae, but their combination is unique and characteristic. Paralaxus is 330 similar to the genus Laxus (as reflected in the genus name), but also to the genus 331 Leptonemella. The three genera share the extreme forward position of the spiral 332 amphidial fovea, a well-developed cephalic capsule, and a gubernaculum without 333 apophysis. Paralaxus can be differentiated from Laxus by the lack of a block-layer in 334 the cephalic capsule, the presence of a velum on the male tail and in addition by the 335 multi-layered symbiont coat. Paralaxus shares the lack of the block-layer and the 336 multilayered symbiont coat, with Leptonemella, but is distinguished by the greater 337 relative pharynx length b’ (pharynx length/body diameter at end of pharynx), the 338 presence of a velum on the male tail, and the lack of sexual dimorphism in the shape 339 of the amphidial fovea. 340 A special case is L. brevipharynx Armenteros et al. (2014 b). According to phylogenetic 341 analyses of its 18S and partial COI data (Fig. 3a and b) the species belongs to the genus 342 Leptonemella. It is morphologically similar to P. cocos sp. nov., and despite being 343 molecularly a bona-fide Leptonemella, has a short pharynx and no sexual dimorphism 344 in the shape of the amphidial fovea. There are, however, two morphological features, bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 14

345 which clearly separate L. brevipharynx from Paralaxus: the lack of a velum on the tip of 346 the tail in males and the lack of the long forward directed cephalic setae characteristic 347 for all Paralaxus species.

348 Molecular data corroborates the monophyly of the subfamily

349 Stilbonematinae and all its genera, including Paralaxus gen. nov.

350 In addition to the nine Paralaxus specimens, we generated and analysed single worm 351 metagenomes from 20 specimens covering six additional stilbonematine genera and 352 extracted both host 18S and COI gene sequences for all specimens (Supplementary 353 Table S1). To analyse the monophyly of the Stilbonematinae, we reconstructed their 354 phylogenetic relationships based on the 18S gene, as this is the only gene with 355 sufficient taxon sampling. The 18S gene matrix, with a minimum sequence length of 356 1300 bp, consisted of sequences from 78 Stilbonematinae as well as from 52 closely 357 related Desmodorida and Microlaimida, and an outgroup of 24 Chromadorida 358 sequences. The Stilbonematinae formed a highly supported clade, with other non 359 stilbonematine Desmodoridae and Draconematidae as closest relatives (Fig. 3a). 360 For the analysis of the Stilbonematinae genera, we created a mitochondrial COI tree in 361 addition to the 18S tree, using 27 metagenome-derived stilbonematine COI sequences 362 and an ascarid as outgroup (Baylisascaris procyonis mitochondrion, JF951366) (Fig. 3b). 363 Based on both datasets, all Stilbonematinae genera including Paralaxus gen. nov. 364 formed well-supported clades (Fig. 3a and b), thus corroborating our morphological 365 assignments. In both the 18S and the COI-based analysis, the genus Leptonemella was 366 phylogenetically most closely related to Paralaxus, with high statistical support in the 367 18S dataset and acceptable support in the COI-based phylogenies (Fig. 3a and b). 368 Other than that, the branching patterns between genera were largely inconsistent 369 between the COI and 18S datasets. This could be linked to the remarkably low support 370 values for many internal nodes in the COI dataset compared to the much higher 371 spectrum of support values for the 18S-based tree topologies. 372 In an extended 18S and COI data-matrix, we also included eight short-length, PCR- 373 amplified sequences of five different Stilbonematinae species with a record of bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 15

374 problematic phylogenetic placement – Laxus parvum, Robbea porosum, Stilbonema 375 brevicolle, Catanema exile and Leptonemella brevipharynx (Armenteros et al. 2014b) 376 (Supplementary Figs. S9 and S10). With our new 18S and COI-based datasets the 377 placements of Leptonemella brevipharynx and Catanema exile sequences were 378 resolved and both clustered with sequences of their respective genera. For the 379 sequences of the other three species we observed inconsistent grouping with 380 Stilbonematinae genera between the datasets (Supplementary Figs S9 and S10). In the 381 extended 18S tree, two Stilbonema brevicolle sequences appeared in two different 382 clades, one closely related to Leptonemella, while the other one clustered with a yet 383 undescribed genus-level clade designated as genus ‘B’ (Zimmermann et al. 2016). In 384 contrast to the 18S-based placement, the Stilbonema brevicolle sequences grouped 385 together in the extended COI tree and formed a novel sister clade to several genera 386 including Stilbonema, Robbea, Leptonemella and Paralaxus (Supplementary Fig. S10). 387 In the extended 18S data-matrix, the Laxus parvum sequences clustered within the 388 genus Robbea (Supplementary Fig. S9). Similarly, in the COI dataset the L. parvum 389 sequence grouped with the only available sequence of the genus Robbea 390 (Supplementary Fig. S10). The sequences designated as Robbea porosum formed a 391 divergent genus-level clade but the placement of this clade was not consistent for the 392 two marker genes (Supplementary Figs S9 and S10). In the 18S-based analyses they 393 were the sister clade to the genus Laxus, while in the COI dataset they formed the 394 sister clade to all other Stilbonematinae. 395

396 Symbiont 16S phylogeny shows genus-level specificity

397 We constructed a matrix from all full-length 16S ‘Thiosymbion’ sequences, available in 398 the databases and 21 metagenome-derived symbiont sequences - using five free-living 399 Chromatiaceae as an outgroup. All symbiont sequences from the same host genus 400 clustered in well-supported clades, irrespective of the number of gutless oligochaete 401 symbionts included within those clades (Fig. 4). In congruence to the host data, the 402 symbionts of all newly sequenced Paralaxus individuals as well as the two previously bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 16

403 published sequences from the Caribbean and Australia (Zimmermann et al. 2016) 404 formed a supported clade in the 16S phylogeny (Fig. 4).

405 COI primers cannot cover the compositional diversity in Stilbonematinae

406 Despite a highly sensitive PCR setup, we could not amplify the COI of any Paralaxus 407 specimen using the widely used primer set by Folmer et al. (1994) (data not shown). To 408 elucidate these PCR-based amplification problems, we analysed all our metagenome- 409 derived full-length stilbonematine nematode COI sequences for compositional 410 patterns. The COI sequences of the 19 stilbonematine nematode species that belonged 411 to seven genera had a highly variable guanine and cytosine (GC) content, ranging from 412 27.7% in the genus Robbea to 53.1 % in the genus Leptonemella. We identified 413 substantial priming problems with the Folmer et al. primer set as well with as a second 414 set (JB3F and JB5R) that is commonly used for COI amplification (Bowles et al. 1992, 415 Derycke et al. 2010, Armenteros et al. 2014a,b). Across the Stilbonematinae diversity 416 the Folmer et al. primers had a range of 3 - 11 mismatches for the 1490F primer and 1 417 - 5 mismatches for the 2198R primer. Similarly, the primers JB3F and JB5R used in a 418 recent study by Armenteros et al. (2014b) showed up to 11 mismatches for each 419 primer (Fig. 5). Of all four primers tested only the JB5R perfectly matched, and only to 420 three sequences from Robbea hypermnestra (Fig. 5). We performed an in-silico 421 analysis that mimics PCR conditions of low stringency, which would overcome up to 11 422 primer mismatches and would cover the whole diversity of Stilbonematinae. Our 423 analysis showed that such permissive conditions would lead to unspecific binding at 424 multiple alternate sites of the COI gene. We tried to design alternative primer sets, but 425 no primer pair, neither for the full-length nor for partial regions, would reliably target 426 the COI gene and would cover all genera of Stilbonematinae. 427 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 17

428 The symbiont coat as a new diagnostic feature for Stilbonematinae

429 genera

430 In addition to several host-based morphological differences, the presence of a multi- 431 layered coat distinguishes the new genus Paralaxus from Laxus, which carries a 432 monolayer of symbionts. Extending from this observation we analysed the structure 433 and morphology of the bacterial coat of all Stilbonematinae genera and found 434 substantial and highly informative differences. We therefore propose to include the 435 description of the bacterial coat given below into the genus trait sets as a new 436 character (see also iconographic overview of the different genera in Fig. 6). 437 438 Adelphos Ott 1997 439 Crescent-shaped bacteria, all cells in contact with host cuticle, double-attached, 440 forming complex monolayer, covering whole body except anterior and 441 posterior tip 442 443 Catanema Cobb 1920 444 Cocci or coccobacilli, all cells in contact with host cuticle, forming simple 445 monolayer, covering whole body except anterior and posterior tip or starting at 446 distance from anterior end 447 448 Centonema Leduc 2013 449 Rods (bacilli), lying parallel to cuticle, all cells in contact with host cuticle, 450 forming simple monolayer, probably covering whole body 451 452 Eubostrichus Greef 1869 453 E. parasitiferus-Type 454 Crescent-shaped bacteria, all cells in contact with host cuticle, double-attached, 455 forming complex monolayer, covering whole body except anterior and 456 posterior tip 457 E. dianeae-Type bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 18

458 Filamentous bacteria, all cells in contact with host cuticle, singly attached, 459 forming complex monolayer, covering whole body except anterior and 460 posterior tip 461 462 Laxus Cobb 1894 463 Cocci, coccobacilli or rods (bacilli), all cells in contact with host cuticle, rods 464 standing perpendicular to cuticle surface, forming simple monolayer, covering 465 whole body except anterior and posterior tip or starting at distance from 466 anterior end 467 468 Leptonemella Cobb 1920 469 Cocci, only a fraction of cells in contact with host cuticle, forming multilayer 470 embedded in gelatinous matrix, covering whole body except anterior and 471 posterior tip 472 473 Parabostrichus Tchesunov, Ingels & Popova 2012 474 Crescent shaped bacteria, all cells in contact with host cuticle, double-attached, 475 forming complex monolayer, covering whole body except anterior and 476 posterior tip 477 478 Paralaxus 479 Cocci or coccobacilli, only a fraction of cells in contact with host cuticle, forming 480 multilayer embedded in gelatinous matrix, covering whole body except anterior 481 and posterior tip 482 483 Robbea Gerlach 1956 484 Cocci, coccobacilli or rods (bacilli), all cells in contact with host cuticle, rods 485 standing perpendicular to cuticle surface, forming simple monolayer, covering 486 whole body except anterior and posterior tip or starting at distance from 487 anterior end bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 19

488 489 Squanema Gerlach 1963 490 Crescent-shaped bacteria, all cells in contact with host cuticle, single or double 491 attached, forming monolayer, starting at distance from anterior end 492 493 Stilbonema Cobb 1920 494 S. brevicolle-Type 495 Cocci or rods, all cells in contact with host cuticle, forming simple monolayer 496 covering whole body except for anterior and posterior tip 497 S. majum-Type 498 Cocci, only a fraction of cells in contact with host cuticle, forming multilayer 499 embedded in gelatinous matrix, covering whole body except anterior and 500 posterior tip 501

502 Discussion

503 COI-based barcoding is not suitable for Stilbonematinae

504 The COI gene is the most common marker gene for PCR-based animal (meta) 505 barcoding, despite the observation that highly diverse base composition have led to 506 unreliable amplification results in many animal phyla including nematodes (Bhadury et 507 al. 2006, Derycke et al. 2010, Deagle et al. 2014). By utilizing single nematode 508 metagenomics we could overcome such amplification problems and show that the COI 509 is a valuable phylogenetic marker gene to distinguish stilbonematine genera and 510 species. However, analyses of the metagenome-derived full-length COI sequences 511 revealed that within Stilbonematinae the COI gene has a highly diverse base 512 composition, prohibiting the design of specific primers on any region of the COI gene. 513 Our mismatch analyses for commonly used primer sets showed multiple mismatches 514 as well as possible alternate binding sites. One may overcome priming mismatches by 515 using very low stringency PCR conditions, at the cost of specificity. Armenteros et al. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 20

516 (2014b) were able to amplify COI fragments from stilbonematine DNA extracts but 517 morphological species descriptions and 18S-based results did not conform to the 518 phylogenetic placement based on the COI. One possible explanation could be that the 519 PCR picked up minimal contaminations of other Stilbonematinae with lower mismatch 520 counts, instead of the gene of the target organism with high numbers of mismatches 521 (Fig. 5). Together these factors show that the COI gene is not well suited for PCR-based 522 barcoding in Stilbonematinae.

523 Metagenomics-based molecular data leads to robust phylogenies of

524 Stilbonematinae

525 Up to this study, 18S and COI genes for Stilbonematinae were PCR-amplified from DNA 526 extracts – a technique known to be highly susceptible to contamination and bias. This 527 is reflected in two studies by Armenteros et al. (2014a,b), which contain phylogenies 528 based on short PCR-amplified stilbonematine nematode sequences. These studies 529 casted doubt on the monophyly of the subfamily Stilbonematinae, as well as the 530 validity of several stilbonematine genera. As discussed by the authors themselves 531 (Armenteros et al. 2014a) and Leduc and Zhao (2016), these conflicting results could 532 have been caused by misidentification of species, PCR bias or artefacts of the 533 phylogenetic reconstruction. To resolve these questions, we used single worm 534 metagenomics from photo-documented specimens to generate high quality and full- 535 length 18S and COI sequences that cover the majority of the described stilbonematine 536 genera. Our dataset corroborated the monophyly of the subfamily Stilbonematinae 537 and verified the taxonomic status of all analysed stilbonematine genera. We caution 538 against the use of short and PCR-based stilbonematine nematode 18S and COI 539 sequences (see Supplementary Figs. S9 and S10) for phylogenetic analyses and 540 barcoding. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 21

541 Low coverage next generation sequencing should replace PCR-based

542 barcoding approaches

543 Instead of using PCR-based barcoding, a straightforward solution to capture marker 544 genes of a target organism are single individual genome or metagenome sequencing 545 approaches employing next generation sequencing (NGS). Our data shows that an 546 NGS-based approach has the advantage that not only single genes but a wide range of 547 marker genes, such as the full rRNA operon as well as the full mitochondrial genome 548 can be recovered from shallow sequencing depths. Recent studies have shown that 549 successful library preparation can be done from as little as 1 picograms of DNA (Rinke 550 et al. 2016) and has been used on a range of microscopic eukaryotic taxa (Jäckle et al. 551 2018, Gruber-Vodicka et al. 2019, Seah et al. 2019). Commercially available kits allow 552 input amounts as low as 1 ng of DNA, and have been used for low cost library 553 construction protocols (Therkildsen & Palumbi 2017). Low coverage NGS-based 554 genomic sequencing could even allow for population genetic studies when combined 555 with appropriate replication and probabilistic allele calling techniques (Therkildsen & 556 Palumbi 2017). Additionally, the associated microbiome is also sequenced, which 557 opens many new avenues for host-microbe research.

558 The symbiont coat is an additional character of host taxonomy

559 In microscopic organisms such as marine meiofauna, taxonomic characters are often 560 inconspicuous and not easily recognized. With their relatively simple body plan only 561 few morphological features are the defining characters for nematode genera. For 562 Stilbonematinae this leads to the problematic situation that characters overlap and 563 only combinations of multiple characters can delineate genera. This was especially 564 evident in the case of Paralaxus gen. nov., as outlined above. The most conspicuous 565 feature of Stilbonematinae is their coat, but initial descriptions of both genera and 566 species often interpreted the obligate symbiotic coat as part of the host (Greeff, 1867), 567 as parasites (Chitwood, 1936) or disregarded the symbiotic coat completely (Cobb, 568 1920). In his taxonomic review Tchesunov (2013) gives brief descriptions of their bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 22

569 cellular morphologies but does not mention the more informative arrangement of the 570 coat itself. 571 The fact that Paralaxus, but also all other stilbonematine host genera are associated 572 with a specific clade of symbiotic bacteria (based on 16S sequences) (Fig. 4 and 573 Zimmermann et al. 2016) allows the use of the symbionts’ cellular morphology and 574 coat structure as taxonomic characters for the holobiont. Both cell morphology and 575 coat structure proved to be characteristic for host taxa at the genus level (Fig. 6). 576 Interestingly, in the stilbonematine genus Eubostrichus, species can host symbionts of 577 either crescent or filamentous shapes (Fig. 6). The 18S and COI sequences of hosts 578 harbouring the two morphotypes are phylogenetically distinct (Fig. 3a and b), and this 579 clustering is also reflected in the symbiont phylogeny (Fig. 4). This suggests that the 580 genus Eubostrichus could require a taxonomical reinvestigation. A similar case can be 581 made for the genus Stilbonema were at least two coat arrangements (Fig.6) are 582 present but comprehensive molecular data does not exist yet. These two examples 583 emphasize the taxonomic value of the symbiotic coat to reassess the assignment of 584 specimens to genera. Furthermore, coat-based characters are easily recognized in live 585 specimens as well as in formaldehyde-preserved material where the coat is often 586 intact. The coat thus facilitates stilbonematine genus level differentiation during 587 stereomicroscopic investigations, e.g. directly in the field. We emphasize that the use 588 of the symbiont coats as additional character will both simplify and enhance 589 Stilbonematinae identification and taxonomy. 590

591 Conclusions

592 The last decade has revealed the ubiquitous and intimate link between animals and 593 the microbial world. Conspicuous holobiont features such as the stilbonematine 594 nematode symbiont coats not only support taxonomic identification, but also 595 emphasize the co-evolution of the symbioses. The highly variable base composition of 596 the Stilbonematinae mitochondrial COI hints at the possible coupling between the 597 obligate and long-term symbiont and host mitochondrial genomes. These findings also 598 reiterate the importance to analyse more than a single marker gene. We are entering a bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 23

599 new age in invertebrate zoology – it is becoming a viable option to employ data- 600 intensive NGS sequencing for multi-locus analyses and, when performed in a shotgun 601 metagenomic approach, this allows to consider the permanently associated 602 microbiome at the same time. Exploring these options, our results show how 603 metagenomics and the integration of holobiont features can shape taxonomic studies 604 and likely many more fields of zoology in the future. 605

606 Acknowledgements

607 We are grateful to the excellent assistance of the staff and for laboratory facilities 608 provided by the Carrie Bow Cay Marine Laboratory (Smithsonian Institution) in Belize, 609 the Heron Island Research Station in Australia, the Bermuda Institute of Ocean 610 Sciences in St. George’s, the HYDRA Institute on Elba, Italy and the Wadden Sea Station 611 Sylt of the Alfred Wegener Institute in Bremerhaven, Germany. We thank Ramon 612 Rosello-Mora and the IMEDEA staff for field work support on Mallorca, Spain. We 613 further gratefully acknowledge the collaboration of the Bermuda Aquarium in Flatts 614 Village for granting us access to the sampling site and their facilities. We thank Nicole 615 Dubilier for funding HGV, JZ and NL through the Max Planck Society and the UNI:DOCs 616 program at the University of Vienna for funding FS. 617 We thank CIUS (Core Facility Imaging and Ultrastructural Research) at the Faculty of 618 Life Sciences, University of Vienna, Austria and especially Daniela Gruber for assistance 619 with electron microscopy. Many thanks to Silke Wetzel, Manuel Kleiner and Cécilia 620 Wentrup for their contributions during the Bermuda and Belize sampling expeditions. 621 This is contribution XXX from the Carrie Bow Cay Laboratory (CCRE program, 622 Smithsonian Institution). Completion of the final version of the manuscript was 623 supported by grant P 31594 (J. Ott, PI) of the Austrian Science Fund (FWF). 624 625 626 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 24

627 References

628 Anisimova, M., Gil, M., Dufayard, J. F., Dessimoz, C., & Gascuel, O. (2011). Survey of 629 branch support methods demonstrates accuracy, power, and robustness of fast 630 likelihood-based approximation schemes. Systematic Biology, 60(5), 685-699. 631 https://doi.org/10.1093/sysbio/syr041 632 Armenteros, M., Rojas-Corzo, A., Ruiz-Abierno, A., Derycke, S., Backeljau, T., & 633 Decraemer, W.(2014a). Systematics and DNA barcoding of free-living marine 634 nematodes with emphasis on tropical desmodorids using nuclear SSU rDNA and 635 mitochondrial COI sequences. Nematology, 16(8),979-989. 636 http://doi.org/10.1163/15685411-00002824 637 Armenteros, M., Ruiz-Abierno, A., & Decraemer, W. (2014b). Taxonomy of 638 Stilbonematinae (Nematoda: Desmodoridae): description of two new and three known 639 species and phylogenetic relationships within the family. Zoological Journal of the 640 Linnean Society, 171(1),1-21. https://doi.org/10.1111/zoj.12126 641 Bayer, C., Heindl, N. R., Rinke, C., Lücker, S., Ott, J. A., & Bulgheresi, S. (2009). 642 Molecular characterization of the symbionts associated with marine nematodes of the 643 genus Robbea. Environmental Microbiology Reports, 1(2), 136-144. 644 https://doi.org/10.1111/j.1758-2229.2009.00019.x 645 Bankevich, A., Nurk, S., Antipov, D., Gurevich, A. A., Dvorkin, M., Kulikov, A. S., ... & 646 Pyshkin, A. V.(2012). SPAdes: a new genome assembly algorithm and its applications to 647 single-cell sequencing. Journal of Computational Biology, 19(5), 455- 648 477.https://doi.org/10.1089/cmb.2012.0021 649 Bhadury, P., Austen, M. C., Bilton, D. T., Lambshead, P. J. D., Rogers, A. D., & Smerdon, 650 G. R. (2006). Development and evaluation of a DNA-barcoding approach for the rapid 651 identification of nematodes. Marine Ecology Progress Series, 320, 1-9. 652 http://doi.org/10.3354/meps320001 653 Bowles J, Blair D, McManus DP (1992) Genetic variants within the genus Echinococcus 654 identified by mitochondrial DNA sequencing. Molecular and Biochemical Parasitology 655 54: 165–174 .https://doi.org/10.1016/0166-6851(92)90109-W bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 25

656 Bulgheresi, S., Gruber-Vodicka, H. R., Heindl, N. R., Dirks, U., Kostadinova, M., 657 Breiteneder, H., & Ott, J. A. (2011). Sequence variability of the pattern recognition 658 receptor Mermaid mediates specificity of marine nematode symbioses. The ISME 659 Journal, 5(6), 986. http://doi.org/10.1038/ismej.2010.198 660 Bushnell, B. (2014). BBMap. Retrieved from https://sourceforge.net/projects/bbmap/ 661 Chitwood, B. G. (1936). Some marine nematodes from North Carolina. Proceedings of 662 the Helminthological Society of Washington 3: 1-16. 663 Cobb, NA (1920). One hundred new nemas. Contr. Sci. Nematol. 9: 217-343 664 Dando, P. R., Hughes, J. A., & Thiermann, F. (1995). Preliminary observations on 665 biological communities at shallow hydrothermal vents in the Aegean Sea. Geological 666 Society, London, Special Publications, 87(1), 303 317. 667 http://dx.doi.org/10.1144/GSL.SP.1995.087.01.23 668 Deagle, B. E., Jarman, S. N., Coissac, E., Pompanon, F., & Taberlet, P. (2014). DNA 669 metabarcoding and the cytochrome c oxidase subunit I marker: not a perfect match. 670 Biology letters, 10(9),20140562 https://doi.org/10.1098/rsbl.2014.0562 671 de León, G. P. & Nadler, S. A. (2010). What we don't recognize can hurt us: a plea for 672 awareness about cryptic species. Journal of Parasitology, 96(2), 453-464. 673 https://doi.org/10.1645/GE-2260.1 674 Derycke, S., Fonseca, G., Vierstraete, A., Vanfleteren, J., Vincx, M., & Moens, T. (2008). 675 Disentangling taxonomy within the Rhabditis (Pellioditis) marina (Nematoda, 676 Rhabditidae) species complex using molecular and morphological tools. Zoological 677 Journal of the Linnean Society, 152(1), 1-15. 678 https://doi.org/10.1111/j.1096-3642.2007.00365.x 679 Derycke, S., Vanaverbeke, J., Rigaux, A., Backeljau, T., & Moens, T. (2010). Exploring 680 the use of cytochrome oxidase c subunit 1 (COI) for DNA barcoding of free-living 681 marine nematodes. PLoSOne, 5(10), 682 e13716.https://doi.org/10.1371/journal.pone.0013716 683 Diepenbroek, M., Glöckner, F. O., Grobe, P., Güntsch, A., Huber, R., König-Ries, B., & 684 Triebel, D. (2014). Towards an integrated biodiversity and ecological research data bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 26

685 management and archiving platform: the German federation for the curation of 686 biological data (GFBio). Informatik 2014: 687 Folmer, O., Black, M., Hoeh, W., Lutz, R., & Vrijenhoek, R. (1994). DNA primers for 688 amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan 689 invertebrates. Molecular Marine Biology and Biotechnology, 3(5), 294-299. 690 Guindon, S., Dufayard, J. F., Lefort, V., Anisimova, M., Hordijk, W., & Gascuel, O. 691 (2010). New algorithms and methods to estimate maximum-likelihood phylogenies: 692 assessing the performance of PhyML 3.0. Systematic Biology, 59(3), 307-321. 693 https://doi.org/10.1093/sysbio/syq010 694 Gerlach, S. A. (1950). Uber einige Nematoden aus der Familie der Desmodoriden. S. 695 178-198 in: Neue Ergebnisse und Probleme der Zoologie (Klatt-Festschrift), Leipzig. 696 Greeff, R., (1869). Untersuchungen ueber einige merkwuerdige Thiergruppen des 697 Arthropoden- und Wurm-Typus. Archiv fuer Naturgeschichte 35:71-121 698 Gruber-Vodicka HR, Pruesse E, Seah B. 2017. phyloFlash. Online: 699 https://github.com/HRGV/phyloFlash 700 Gruber-Vodicka, H. R., Leisch, N., Kleiner, M., Hinzke, T., Liebeke, M., McFall-Ngai, M., 701 Hadfield, M.G. & Dubilier, N. (2019). Two intracellular and cell type-specific bacterial 702 symbionts in the placozoan Trichoplax H2. Nature Microbiology, 1. 703 https://doi.org/10.1038/s41564-019-0475-9 704 Heindl, N. R., Gruber-Vodicka, H. R., Bayer, C., Lücker, S., Ott, J. A., & Bulgheresi, S. 705 (2011). First detection of thiotrophic symbiont phylotypes in the pelagic marine 706 environment. FEMS Microbiology Ecology, 77(1), 223-227. 707 https://doi.org/10.1111/j.1574-6941.2011.01096.x 708 Jäckle, O., Seah, B. K. B., Tietjen, M., Leisch, N., Liebeke, M., Kleiner, M., Berg, J.S. & 709 Gruber-Vodicka, H. R. (2019). Chemosynthetic symbiont with a drastically reduced 710 genome serves as primary energy storage in the marine flatworm Paracatenula. 711 Proceedings of the National Academy of Sciences 116 (17), 8505-8514. 712 https://doi.org/10.1073/pnas.1818995116 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 27

713 Kalyaanamoorthy, S., Minh, B. Q., Wong, T. K., von Haeseler, A., & Jermiin, L. S. (2017). 714 ModelFinder: fast model selection for accurate phylogenetic estimates. Nature 715 Methods, 14(6), 587. http://doi.org/10.1038/nmeth.4285 716 Kamenev, G. M., Fadeev, V. I., Selin, N. I., Tarasov, V. G., & Malakhov, V. V. (1993). 717 Composition and distribution of macro-and meiobenthos around sublittoral 718 hydrothermal vents in the Bay of Plenty, New Zealand. New Zealand Journal of 719 Marine and Freshwater Research, 27(4), 720 407418.https://doi.org/10.1080/00288330.1993.9516582 721 Kampfer, S., Sturmbauer, C., & Ott, J. (1998). Phylogenetic analysis of rDNA sequences 722 from adenophorean nematodes and implications for the Adenophorea-Secernentea 723 controversy. Invertebrate Biology, 29-36. http://doi.org/10.2307/3226849 724 Katoh, K., & Toh, H. (2008). Recent developments in the MAFFT multiple sequence 725 alignment program. Briefings in Bioinformatics, 9(4), 286-298. 726 https://doi.org/10.1093/bib/bbn013 727 Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence alignment software 728 version 7: improvements in performance and usability. Molecular Biology and 729 Evolution, 30(4), 772-780.https://doi.org/10.1093/molbev/mst010 730 Leduc, D. (2013). One new genus and two new deep-sea nematode species 731 (Desmodoridae, Stilbonematinae) from phosphorite nodule deposits on Chatham Rise, 732 Southwest Pacific Ocean. Marine Biodiversity, 43(4), 421-428. 733 https://doi.org/10.1007/s12526-013-0171-6 734 Leduc, D., & Sinniger, F. (2018). Combining morphological and molecular data to 735 classify Laxus sakihariiae sp. n., a new stilbonematine nematode (Nematoda: 736 Desmodoridae) from the coast of Sesoko Island, Japan. Nematology, 20(1), 81-95. 737 https://doi.org/10.1163/15685411-00003126 738 Leduc, D., & Zhao, Z. (2016). Phylogenetic relationships within the superfamily 739 Desmodoroidea (Nematoda: Desmodorida), with descriptions of two new and one 740 known species. Zoological Journal of the Linnean Society, 176(3), 511-536. 741 https://doi.org/10.1111/zoj.12324 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 28

742 Minh, B. Q., Nguyen, M. A. T., & von Haeseler, A. (2013). Ultrafast approximation for 743 phylogenetic bootstrap. Molecular Biology and Evolution, 30(5), 1188-1195. 744 https://doi.org/10.1093/molbev/mst024 745 Nguyen, L. T., Schmidt, H. A., von Haeseler, A., & Minh, B. Q. (2014). IQ-TREE: a fast 746 and effective stochastic algorithm for estimating maximum-likelihood phylogenies. 747 Molecular Biology and Evolution, 32(1), 268-274. 748 https://doi.org/10.1093/molbev/msu300 749 Ott, J, Bright, M & Bulgheresi, S. (2004). Symbiosis between Marine Nematodes and 750 Sulfur oxidizing Chemoautotrophic Bacteria. Symbiosis 36(2), 103-126 751 Ott, J. A., Gruber-Vodicka, H. R., Leisch, N., & Zimmermann, J. (2014a). Phylogenetic 752 confirmation of the genus Robbea (Nematoda: Desmodoridae, Stilbonematinae) with 753 the description of three new species. Systematics and Biodiversity, 12(4), 434-455. 754 https://doi.org/10.1080/14772000.2014.941038 755 Ott, J. A., Leisch, N., & Gruber-Vodicka, H. R. (2014b). Eubostrichus fertilis sp. n., a new 756 marine nematode (Desmodoridae: Stilbonematinae) with an extraordinary 757 reproductive potential from Belize, Central America. Nematology, 16(7), 777-787. 758 https://doi.org/10.1163/15685411-00002807 759 Ott, J. A. & Novak, R. (1989). Living at an Interface: Meiofauna at the Oxygen/Sulfide 760 Boundary of Marine Sediments, In: J. S. Ryland and P. A. Tyler, Eds., Reproduction, 761 Genetics and Distribution of Marine Organisms, Olsen & Olsen, Fredensborg, pp. 415- 762 422

763 Palomares-Rius, J. E., Cantalapiedra-Navarrete, C., & Castillo, P. (2014). Cryptic species 764 in plant parasitic nematodes. Nematology, 16(10), 1105-1118. 765 https://doi.org/10.1163/15685411-00002831 766 Platt, H. M., & Zhang, Z. N. (1982). New species of marine nematodes from Loch Ewe, 767 Scotland. British Museum (Natural History). 768 Polz, M. F., Felbeck, H., Novak, R., Nebelsick, M., & Ott, J. A. (1992). 769 Chemoautotrophic, sulfur-oxidizing symbiotic bacteria on marine nematodes: 770 morphological and biochemical characterization. Microbial Ecology, 24 (3), 313-329. 771 https://doi.org/10.1007/BF00167789 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 29

772 Polz, M., Distel, D., Zarda, B., Amann, R., Felbeck, H., Ott, J., & Cavanaugh, C (1994). 773 Phylogenetic analysis of a highly specific association between ectosymbiotic, sulfur- 774 oxidizing bacteria and a marine nematode. Applied Environmental Microbiology, 60 775 (12), 4461-4467. 776 Riemann, F., Thiermann, F., & Bock, L. (2003). Leptonemella species (Desmodoridae, 777 Stilbonematinae), benthic marine nematodes with ectosymbiotic bacteria, from littoral 778 sand of the North Sea island of Sylt: taxonomy and ecological aspects. Helgoland 779 Marine Research, 57(2), 118. https://doi.org/10.1007/s10152-003-0149-z 780 Rinke, C., Low, S., Woodcroft, B. J., Raina, J-B., Skarshewski A., Le, X.H., Butler, M.K., 781 Stocker, R., Seymour, J., Tyson, G.W. & Hugenholtz, P. (2016) Validation of picogram- 782 and femtogram-input DNA libraries for microscale metagenomics. PeerJ 4:e2486. 783 https://doi.org/10.7717/peerj.2486 784 SantaLucia, J. (1998). A unified view of polymer, dumbbell, and oligonucleotide DNA 785 nearest-neighbor thermodynamics. Proceedings of the National Academy of Sciences, 786 95(4), 1460-1465. https://doi.org/10.1073/pnas.95.4.1460 787 Seah BKB, Antony CP, Huettel B, Zarzycki J, Schada von Borzyskowski L, Erb TJ, Kouris A, 788 Kleiner M, Liebeke M, Dubilier N, Gruber-Vodicka HR (2019) Sulfur-oxidizing symbionts

789 without canonical genes for autotrophic CO2 fixation. mBio 10(3):e01112-19. 790 https://doi.org/10.1128/mBio.01112-19 791 Sudhaus, W., & Kiontke, K. (2007). Comparison of the cryptic nematode species 792 Caenorhabditis brenneri sp. n. and C. remanei (Nematoda: Rhabditidae) with the stem 793 species pattern of the Caenorhabditis Elegans group. Zootaxa, 1456(1), 45-62. 794 Tchesunov, A. V., Ingels, J., & Popova, E. V. (2012). Marine free-living nematodes 795 associated with symbiotic bacteria in deep-sea canyons of north-east Atlantic Ocean. 796 Journal of the Marine Biological Association of the United Kingdom, 92(6), 1257-1271. 797 https://doi.org/10.1017/S0025315411002116 798 Tchesunov, A. V. (2013). Marine free-living nematodes of the subfamily 799 Stilbonematinae (Nematoda, Desmodoridae): taxonomic review with descriptions of a 800 few species from the Nha Trang Bay, Central Vietnam. Meiofauna Marina, 20, 71-94. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 30

801 Therkildsen, N. O. & Palumbi, S. R. (2017). Practical low-coverage genome wide 802 sequencing of hundreds of individually barcoded samples for population and 803 evolutionary genomics in non-model species. Molecular Ecology Resources, 17(2), 194- 804 208. https://doi.org/10.1111/1755-0998.12593 805 Thiermann, F., Koumanidis, I., Hughes, J. A. & Giere, O. (1997). Benthic fauna of a 806 shallow-water gas hydrothermal vent area in the Aegean Sea (Milos, Greece). Marine 807 Biology, 128(1),149159. https://doi.org/10.1007/s002270050078 808 Untergasser, A., Cutcutache, I., Koressaar, T., Ye, J., Faircloth, B. C., Remm, M., & 809 Rozen, S. G. (2012). Primer3—new capabilities and interfaces. Nucleic Acids Research, 810 40(15), e115-e115. https://doi.org/10.1093/nar/gks596 811 Van Gaever, S., Vanreusel, A., Hughes, J. A., Bett, B. J. & Kiriakoulakis, K. (2004). The 812 macro-and microscale patchiness of meiobenthos associated with the Darwin Mounds 813 (north-east Atlantic). Journal of the Marine Biological Association of the United 814 Kingdom, 84(3), 547-556. https://doi.org/10.1017/S0025315404009555h 815 van Megen, H., van den Elsen, S., Holterman, M., Karssen, G., Mooyman, P., Bongers, 816 T., & Helder, J. (2009). A phylogenetic tree of nematodes based on about 1200 full- 817 length small subunit ribosomal DNA sequences. Nematology, 11(6), 927-950. 818 Http://doi.org/10.1163/156854109X456862 819 Wieser, W. (1960). Benthic studies in Buzzards Bay. II. The meiofauna. Limnological 820 Oceanography, 5: 121137 https://doi.org/10.4319/lo.1960.5.2.0121 821 Zimmermann, J., Wentrup, C., Sadowski, M., Blazejak, A., Gruber-Vodicka, H. R., 822 Kleiner, M., Ott, J.A.,Cronholm, B., De Wit, P., Erséus, C. & Dubilier, N. (2016). Closely 823 coupled evolutionary history of ecto-and endosymbionts from two distantly related 824 animal phyla. Molecular Ecology, 25(13), 3203-3223. 825 https://doi.org/10.1111/mec.13554 826 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 31

827 Tables and Figures

828 Figures

829 830 Fig. 1. Paralaxus gen. nov. habitus and sampling sites of stilbonematine nematodes. (a) 831 Micrograph of Paralaxus bermudensis sp. nov. The symbiont free anterior region is clearly 832 visible compared to the densely coated white body. (b) Sampling locations of Paralaxus species 833 included in this study (c) Juvenile specimen of Paralaxus columbae sp. nov. with thick 834 symbiotic coat and bacteria free head region. (d) Sampling locations of other stilbonematine 835 nematode species included in this study. Specimen for this study were sampled from shallow- 836 water sandy sediments in the Caribbean Sea, the Atlantic and the Pacific Ocean as well as the 837 North and Mediterranean Sea. For details see Table S1 (Supporting information). 838 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 32

839 840 Fig. 2. Drawings of three different Paralaxus species comparing their defining characters. (a - 841 c. Paralaxus cocos nov. spec. male, (a) total view, (b) anterior end (c) posterior end (d – g) P. 842 cocos female, (d) total view (e) anterior end, optical section (f) anterior end, surface view (h – bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 33

843 j) P. bermudensis nov. spec. male (h) total view (i) anterior end (j) posterior end (k – o) P. 844 columbae nov. spec. male and female. (k) total view of male paratype, (l) anterior end of 845 male,(m) anterior end of female (n) posterior end of male (o) posterior end of female. Camera 846 lucida drawings. Arrows point to amphidial fovea and velum on male tail. 847 848 849 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 34

(a) 18S rRNA

850 851 (b) Cytochrome oxidase I

852 853 Fig. 3. Phylogenetic relationship of Paralaxus species with other stilbonematine nematodes 854 based on the 18S rRNA (a) and cytochrome oxidase subunit I (COI) genes (b). The trees were 855 calculated using IQ-Tree. Support values are given in the following order: SH-aLRT support (%) 856 / aBayes support / ultrafast bootstrap support (%). Species described in this study are bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 35

857 highlighted in bold. Provisional working names for undescribed genera or species are given in 858 quotes. The scale bar represents average nucleotide substitutions per site. 859 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 36

860 861 862

863 864 Fig. 4. Candidatus Thiosymbion phylogeny, including symbionts of Paralaxus species and 865 other stilbonematine nematodes, based on the 16S rRNA gene. The tree was calculated using 866 IQTree and node support is given in the following order: SH-aLRT support (%) / aBayes support 867 / ultrafast bootstrap support (%). Symbionts of species described in this study are marked bold 868 or highlighted in bold numbers next to collapsed branches. Provisional working names for 869 undescribed genera or species are given in quotes. The scale bar represents average 870 nucleotide substitutions per site. 871 872 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Scharhauser et al. 37

873 874 Fig. 5. Primer mismatch analysis of the two commonly used COI primer sets 875 Potential priming sites (forward and reverse) for the primer sets 1490F/ 2198R (Folmer et al. 876 1994) and J3BF/ J5BR (Armenteros et al. 2014 a,b) along the COI gene are shown in red and 877 blue on the top. The number of mismatches for each primer, when matched to metagenome- 878 extracted full-length COI sequences from seven representative stilbonematine nematode 879 genera, is shown below on the left. Average GC content for each full-length COI sequence is 880 represented by the scale bar on the right. certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint Scharhauser et al. 38 doi: https://doi.org/10.1101/728105 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019. The copyrightholderforthispreprint(whichwasnot . 881 882 Fig. 6. Cellular morphology and coat structure of symbiotic bacteria on different stilbonematine nematode genera. All described genera of 883 Stilbonematinae including Paralaxus nov. gen. are depicted in alphabetical order. Each pictogram shows the anterior end of a male in optical section 884 (left), the head region of a male in surface view (top center), male posterior region with spicular apparatus (lower right) and shape and arrangement 885 of symbionts (upper right). Special features such as arrangement of complex symbiont coat (Adelphos, Eubostrichus), male supplementary structures certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint Scharhauser et al. 39

886 (Catanema, Robbea, Stilbonema), sexual dimorphism in the shape of the amphidial fovea (Leptonemella) or change in cuticular structure (Laxus, doi:

887 Squanema) are shown in the center of the pictogram. Drawings do not represent a defined species but include the general characters of each genus. https://doi.org/10.1101/728105 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019. The copyrightholderforthispreprint(whichwasnot . bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

888 Supplementary Information

889 SI Results

890 Descriptions

891 Class Chromadorea Inglis, 1983 892 Subclass Chromadoria Pearse, 1942 893 Order Desmodorida De Coninck, 1965 894 Suborder Desmodorina De Coninck, 1965 895 Superfamily Desmodoroidea Filipjev, 1922 896 Family Desmodoridae Filipjev, 1922 897 Subfamily Stilbonematinae Chitwood, 1936 898 Paralaxus gen. nov. 899 900 http://zoobank.org: 417691DB-5EE3-49B7-B838-1BBF1569235F 901 Diagnosis. Stilbonematinae. Cuticle transversely striated, except for the head region 902 and the tip of the tail. The striation is coarser in the anterior region, becoming finer at 903 3 - 4 pharynx lengths. Cephalic capsule distinct, without a block-layer, finely punctated; 904 cephalic setae as long or longer than the sub-cephalic setae, usually directed straight 905 forward. Amphidial fovea an open spiral, ventrally wound, in extreme forward 906 position. No sexual dimorphism in the shape of the amphidial fovea. Pharynx bulbus 907 large, muscular. Gubernaculum dorsally directed with terminal hook, without 908 apophysis. Males have a distinct velum at the tip of the tail. Glandular sense organs 909 (gso) enlarged ventrally in post-pharyngeal region in males; anterior and posterior to 910 vulva in females. Body covered by a multi-layered coat of symbiotic coccobacilli. 911 Etymology: Superficially resembling species of the genus Laxus Cobb. 912 Type species: Paralaxus cocos sp. nov. 913 914 Symbiont Diagnosis. The bacterial coat consists of coccobacilli shaped cells between 915 0.8 µm to 1.7 µm in size, which are distributed over the whole body, leaving only the bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

916 tip of the tail and the cephalic capsule uncovered. They are arranged in a thick 917 multilayer from 6 up to 12 layers of bacteria, which are embedded in a gelatinous 918 matrix showing no layer separation or structure. Only bacterial cells at the bottom of 919 the coat are in contact with the cuticle of the nematode host. Most cells show pili-like 920 structures that are connected to the worm surface and/or to neighboring bacteria. The 921 bacterial cells divide transversely. 922 923 Paralaxus cocos sp. nov. 924 (Figs. 2 – 23) 925 ZooBank registration: http://zoobank.org A659918E-FD88-4584-8115-CD4C396AE400 926 Type material: Holotype (male), 3 paratypes (male), 3 paratypes (female), 4 paratypes 927 (juvenile), deposited at the Natural History Museum Vienna (accession numbers 928 NHMWZooEvMikro 5689 - 5699) 929 930 Measurements. See Table 1 931 932 Additional material: Several specimens in the collection of JAO and those used for 933 SEM. 934 935 Type locality: Subtidal fine sand among Red Mangrove (Rhizophora mangle) roots and 936 turtle grass (Thalassia testudinum), 10 – 50 cm depth, W-side of Twin Cayes, BELIZE 937 (16°49'45.6"N 88°06'28.9"W) 938 939 Distribution: Regularly in fine to medium subtidal sands around Carrie Bow Cay, 940 Curlew Cay and Twin Cayes (Belize). 941 942 Etymology: From the Latin name of the Coconut Tree, Cocos nucifera, one of which 943 played an important role in the research trip that led to the discovery of the new 944 genus. 945 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

946 Description: 947 Body cylindrical, head diameter at level of cephalic setae 20 - 25 µm, diameter at level 948 of posterior margin of amphidial fovea 23 -28 µm in males, 20 - 25 µm in females, at 949 end of pharynx 46 - 57 µm, maximum body diameter 46 - 66 µm, anal diameter 50 - 55 950 µm in males, 40 - 42 µm in females. Tail conical, 115 - 120 µm long in males and 98 - 951 125 µm in females. 952 Cuticle transversely striated except for the first 25 - 30 µm of the head and the last 40 - 953 50 µm of the tail (Fig. 14), striae 0.6 to 0.7 µm wide (14 - 16 striae/10 µm) in anterior 954 body part, 0.4 -0.5 µm (20 - 23 striae/10µm) after 3 – 3.7 pharynx length from anterior 955 end. The anterior most circle of head sensillae (inner labial sensillae) is represented by 956 6 spoon-shaped papillae within the mouth opening, in lateral, subventral and 957 subdorsal position. The second circle consists of 6 short outer labial sensillae, 1 µm 958 long, on the margin of the membranous buccal field in lateral, subventral and 959 subdorsal position; 4 cephalic setae near the anterior margin of the amphidial fovea, 960 20 - 28 µm long; 3 circles of 8 subcephalic setae each, 15 - 20 µm long, on the non- 961 striated part of the head region, the submedian rows with additional short setae, 10 962 µm long; 8 rows of somatic setae along the whole length of the body, 12 µm long, 963 spaced 20 µm apart. Non-striated part of the tail in females with 2 pairs of setae, in 964 males with a velum occupying 43 – 47% of tail length. Three caudal glands extending 965 to the gubernaculum. Amphidial foveas situated at the anterior end bordering the 966 buccal field, an open spiral, ventrally wound, with 1.5 turns, 10 – 15 µm long and 14 - 967 15µm wide in males, smaller in females, 6 - 8 µm long and 11 – 12 µm wide. In one 968 male specimen the amphidial fovea was dorsally wound (Fig. 3). 969 Pharynx 89 - 130 µm long, with a minute buccal cavity, 5 µm long, leading into a 970 slightly dilated corpus, which is only indistinctly separated from the isthmus. Posterior 971 bulbus spherical, muscular, large, 28 - 43 µm in diameter. No cardia. The bulbus to 972 pharynx ratio is 32 – 35% in males and 30 - 34% in females. 973 Nerve ring 45 - 55 µm from anterior end; no secretory-excretory pore or ventral gland 974 seen; 8 rows of glandular sense organs (two in each lateral and each median line). bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

975 Males monorchic, testis on the right side of the intestine, beginning at about 36-37% 976 of body length; spicula strong, arcuate, cephalate proximally, 50 - 51 µm (chord) or 60 977 - 65 µm (arc) long; gubernaculum straight, corpus embracing spicules, 31 - 35 µm long, 978 end hook-shaped. Sperm globular. Row of enlarged gso along the mid-ventral line 400 979 µm long, beginning at 250 µm from the anterior end. 980 Females didelphic, ovaries reflexed; vulva at 44 - 45% of body length. Enlarged ventral 981 gso anterior and posterior to vulva, extending 250 µm in either direction. 982 983 Paralaxus bermudensis sp. nov. 984 (Figs. 24 – 30) 985 ZooBank registration: http://zoobank.org 8697C6A0-3508-4833-81DC-D2A86B0B00AC 986 Type material: Holotype (male), 1 paratype (male), 2 paratypes (juvenile), deposited at 987 the Natural History Museum Vienna (accession numbers NHMWZooEvMikro 5700 - 988 5703) 989 990 Measurements. See Table 1 991 992 Type locality: Subtidal sand, 3 m depth, NW of Nonsuch Island, BERMUDA (32°20’ 54’’ 993 N, 64° 39’ 54’’ W) 994 995 Distribution: Also found in subtidal sand in Harrington Sound and Baylis Bay, Bermuda. 996 997 Etymology: From the Bermuda Islands 998 999 Description: 1000 Body cylindrical, head diameter at level of cephalic setae 25 - 35 µm, diameter at level 1001 of posterior margin of amphidial fovea 45 - 50 µm, at end of pharynx 60 - 65 µm, 1002 maximum body diameter 60 - 65 µm, anal diameter 60 - 65 µm. Tail conical, 105 - 125 1003 µm long. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1004 Cuticle transversely striated except for the first 30 - 35 µm of the head and the last 40 - 1005 45 µm of the tail, striae 1 – 1.1 µm wide (9 - 10 striae/ 10 µm) in anterior body part, 1006 appr. 0.5 µm (20/ 10 µm) after 3.2 – 3.4 pharynx lengths from anterior end. The 1007 anterior most circle of head sensillae (inner labial sensillae) not seen, the second circle 1008 as in P. cocos; 4 cephalic setae near the anterior margin of the amphidial fovea, 20 - 25 1009 µm long; 3 circles of 8 subcephalic setae each, 12 - 20 µm long, with additional short 1010 setae, 10 µm long on the non-striated part of the head region; 8 rows of somatic setae 1011 along the whole length of the body, 8 - 10 µm long spaced 20 µm apart. Tip of tail in 1012 males with a velum occupying 36 – 40% of tail length. Three caudal glands extending 1013 to the gubernaculums. Amphidial foveas with 2 turns, 10 – 12 µm long and 12 – 15 µm 1014 wide. 1015 Pharynx 116 – 125 µm long, with a minute buccal cavity, corpus slightly dilated. Bulbus 1016 spherical, muscular, large, 35 – 40 µm in diameter, no cardia. The bulbus to pharynx 1017 ratio is 31 – 33%. 1018 Nerve ring 55 – 65 µm from anterior end; no secretory-excretory pore or ventral gland 1019 seen; 8 rows of glandular sense organs (two in each lateral and each median line). 1020 Males monorchic, testis on the right side of the intestine, beginning at about 36-37% 1021 of body length; spicula strongly arcuate, cephalate proximally, 43 - 45 µm (chord) or 57 1022 - 63 µm (arc) long; gubernaculum straight, 30 – 33 µm long, end hook-shaped. Row of 1023 enlarged GSO beginning at end of pharynx and extending 250 µm backwards. 1024 No female found. 1025 1026 Paralaxus columbae sp. nov. 1027 (Figs. 31 – 39) 1028 ZooBank registration: http://zoobank.org 4BABABB3-1DD4-46F1-A69C-BC06C1271D01 1029 Type material: Holotype (male), 1 paratype (male), 3 paratypes (female), 2 paratypes 1030 (juvenile), deposited at the Natural History Museum Vienna (accession numbers 1031 NHMWZooEvMikro 5704 - 5710) 1032 1033 Measurements. See Table 1 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1034 1035 Additional material: Several specimens in the collection of JAO. 1036 1037 Type locality: Subtidal sand, 2 m depth, N-side of Pigeon Key, Florida, USA (24° 42’ 19’’ 1038 N, 81° 09’ 20’’ W) 1039 1040 Distribution: Only known from type locality. 1041 1042 Etymology: From the Latin columba (pigeon), referring to the type locality. 1043 1044 Description: 1045 Body cylindrical, head diameter at level of cephalic setae 32 - 40, diameter at level of 1046 posterior margin of amphidial fovea 40 - 50 µm, at end of pharynx 54 - 55 µm in males, 1047 55 – 60 µm in females, maximum body diameter 55 µm in males, 55 – 60 µm in 1048 females, anal diameter 50 - 55 µm. Tail conical, 108 - 115 µm long in males and 100 - 1049 120 µm in females. 1050 Cuticle transversely striated except for the first 30 - 35 µm of the head and the last 40 - 1051 50 µm of the tail, striae 0.6 to 0.7 µm wide (14 - 16 striae/10 µm) in anterior body 1052 part, 0.4 -0.5 µm (20 - 23 striae/10µm) after 3 – 3.2 pharynx length from anterior end. 1053 The anterior most circle of head sensillae (inner labial sensillae) not seen. The second 1054 circle as in P. cocos; 4 cephalic setae at the level of the amphidial fovea, 25 - 30 µm 1055 long; 3 circles of 8 subcephalic setae each, 15 - 20 µm long, on the non-striated part of 1056 the head region, the submedian rows with additional short setae, 10 µm long; 8 rows 1057 of somatic setae along the whole length of the body, 7 - 10 µm long, spaced 15 - 20 µm 1058 apart. Males with a velum occupying 35 – 37% of tail length. Three caudal glands 1059 extending to the gubernaculum. Amphidial foveas situated at the anterior end, an 1060 open spiral with 1 – 1.2 turns, ventrally wound, 12 – 13 µm long and wide in both 1061 sexes. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1062 Pharynx 100 - 110 µm long, buccal cavity minute, 5 µm long, corpus not dilated. 1063 Posterior bulbus subspherical, muscular, large, 30 - 35 µm long and up to 45 µm wide. 1064 No cardia. The bulbus to pharynx ratio is 29 - 33% in both sexes. 1065 Nerve ring 45 - 55 µm from anterior end; no secretory-excretory pore or ventral gland 1066 seen; 8 rows of glandular sense organs (two in each lateral and each median line). 1067 Males monorchic, testis on the right side of the intestine, beginning at about 36-37% 1068 of the body length; spicula weakly arcuate, cephalate proximally, 50 - 54 µm (chord) or 1069 57 - 60 µm (arc) long; gubernaculum straight, corpus embracing spicules, 31 - 33 µm 1070 long, end hook-shaped. Enlarged GSO along the mid-ventral line beginning at end of 1071 the pharynx and extending 300 - 350 µm backwards. 1072 Females didelphic, ovaries reflexed; vulva at 44 - 48% of body length. Enlarged ventral 1073 gso anterior and posterior to vulva extending approximately 200 µm in either 1074 direction. 1075 1076 Key to the species: 1077 Amphidial fovea with 2 turns; bulbus 30 - 32% of pharynx length, velum 30 – 40% of 1078 tail length, spicula strongly arcuate 1079 Paralaxus bermudensis sp. nov. 1080 1081 Amphidial fovea with 1.5 turns, bulbus 32 - 35% of pharynx length, velum 43 - 47% of 1082 tail length, spicula moderately arcuate 1083 P. cocos sp. nov. 1084 1085 Amphidial fovea with 1 – 1.2 turns, bulbus 29 - 33% of pharynx length, velum 35 - 37% 1086 of tail length, spicula weakly arcuate 1087 P. columbae sp. nov. 1088 1089 1090 1091 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1092 Paralaxus species from Australia and Hawaii 1093 1094 Two additional Paralaxus species, one from Australia and one from Hawaii have been 1095 discovered by sequencing their 18S rRNA and COI genes alone. Due to the lack of 1096 properly preserved material, a formal description is therefore currently not possible. 1097 However, we were able to make measurements and morphological descriptions from 1098 microphotographs that were taken from the live animals prior to sequencing, which 1099 are listed below: 1100 1101 Paralaxus sp. ”heron 1”” 1102 (Figs. 40 – 42) 1103 1104 Sample location: Heron Island, Great Barrier Reef, AUSTRALIA (24° 26’ 36’’ S, 151°54’ 1105 47’’ W) 1106 Material: 1 female 1107 1108 Description (measurements taken from LM micrographs of sequenced specimen) 1109 Since the total length could not be obtained from the micrographs, the parameters a, b 1110 and c could not be calculated. Maximum diameter at the level of a fully developed egg 1111 was 86 µm, in other regions of the body 78 µm, which was also the diameter at the 1112 end of the pharynx. Cuticle striated, striae 1.2 to 1.3 µm wide. Pharynx 122 µm long, 1113 bulbus 41 µm long (33.6% of pharynx length) and 45 µm wide. Cephalic capsule 28 µm 1114 long, bacterial coat thick, consisting of coccobacilli approximately 1 µm long. 1115 1116 Paralaxus sp. ”oahu 1” 1117 (Figs. 43 – 45) 1118 1119 Sample locations: Oahu, Hawaii Islands, USA (21° 28’ 11,3’’ N, 157° 49’ 4,8’’ W and 1120 21° 16’ 50,4’’ N, 157° 43’ 40,8 W) 1121 Material: 2 individuals, at least 1 male. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1122 1123 Description (measurements taken from LM micrographs of sequenced specimens) 1124 Length 4927 µm 1125 a= 71.4; b= 37.3; c= 35.7 1126 Length of pharynx 132 µm, length of tail 138 µm; maximum diameter 69 µm, head 1127 diameter 41 µm, diameter at end of pharynx 58 µm, cloacal diameter 62 µm, c’= 2.2. 1128 Cuticle transversely striated, except for the cephalic capsule (34 µm long) and the tip 1129 of the tail, striae 1 µm wide, head and body setation not discernible in micrographs. 1130 Amphids in extreme forward position, open spiral with one turn, 13.5 µm wide, 9 µm 1131 long. Pharynx bulbus 36.4 µm wide and 38.8 µm long (29.6% of pharynx length). 1132 Spicula cephalate, curved, 64 µm (arc), 58 µm (chord) long, gubernaculum straight 1133 with terminal hook, 34.5 µm long. Several layers of coccobacilli approximately 1 µm 1134 long covering body. certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

1135 doi: https://doi.org/10.1101/728105

1136 SI Tables

1137 1138 Table S1: Stilbonematine nematode specimens sampled for this study

Accession Accession Habitat Habitat Accession no. Sample ID / Location Sampling no. no. a CC-BY-NC-ND 4.0Internationallicense Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S ;

this versionpostedAugust7,2019. Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

KP943969 KP943985 SW Pacific, -23.44339, W91-3HI / 2028D Paralaxus "heron 1" Heron Island, Australia Aug, 2012 intertidal coarse tropical (Zimmermann (Zimmermann et Coral Sea 151.91307 et al., 2016) al., 2016)

SW Pacific, -23.44339, W91-4HI / 2028E Paralaxus "heron 1" Heron Island, Australia Aug, 2012 intertidal coarse tropical Coral Sea 151.91307 The copyrightholderforthispreprint(whichwasnot .

W Atlantic, 16.82861, W94CBC / 2028G Paralaxus cocos Twin Cays, Belize Apr, 2013 subtidal very fine tropical Caribbean Sea -88.10955

KP943968 KP943984 W Atlantic, 16.82861, W82CBC / 2028F Paralaxus cocos Twin Cays, Belize Apr, 2013 subtidal very fine tropical (Zimmermann (Zimmermann et Caribbean Sea -88.10955 et al., 2016) al., 2016) certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

Paralaxus Harrington Sound, 32.324038, medium BER669 / 2028A N Atlantic May, 2015 subtidal subtropical bermudensis Bermuda -64.738267 coarse a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

Paralaxus Harrington Sound, 32.324038, medium BER658 / 2028B N Atlantic May, 2015 subtidal subtropical bermudensis Bermuda -64.738267 coarse

Paralaxus Harrington Sound, 32.324038, medium BER364 / 2028C N Atlantic May, 2015 subtidal subtropical bermudensis Bermuda -64.738267 coarse The copyrightholderforthispreprint(whichwasnot

Paralaxus Bermuda, Nonsuch 32.3477777, . JO93-1 / - N Atlantic 1993 subtidal medium fine subtropical - - - bermudensis Island -64.665

21.280897, HW15020 / 2028J Paralaxus "oahu 1" Oahu, Hawaii Central Pacific Oct, 2015 intertidal tropical -157.728126 certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

21.280897, HW15003 / 2028H Paralaxus "oahu 1" Oahu, Hawaii Central Pacific Oct, 2015 intertidal tropical -157.728126 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

24.705277777,- JO70-1 / - Paralaxus columbae Florida Atlantic May, 1970 subtidal medium fine subtropical − − − 81,1555555

HGV13CBC-01 / Robbea W Atlantic, 16.81311, Carrie Bow Cay, Belize Apr, 2013 subtidal coarse tropical no sequence no sequence 1385H hypermnestra Caribbean Sea -88.08299 The copyrightholderforthispreprint(whichwasnot

HGV13CBC-02 / Robbea W Atlantic, 16.81311, . Southwater Cay, Belize Apr, 2013 subtidal coarse tropical no sequence no sequence 1385F hypermnestra Caribbean Sea -88.08299

HGV13CBC-03 / Robbea W Atlantic, 16.81311, Southwater Cay, Belize Apr, 2013 subtidal coarse tropical 1385G hypermnestra Caribbean Sea -88.08299 certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

55.01467, JZ_W217 / 1094A Leptonemella vicina List, Sylt North Sea June, 2013 intertidal fine temperate KU921488 KU921521 8.43752 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

Leptonemella 55.01467, JZ_W377/ 1855C List, Sylt North Sea Sept, 2015 intertidal fine temperate aphanothecae 8.43752

W Atlantic, 16.8030556, NL14CBC-01 / 1385P Stilbonema majum Carrie Bow Cay, Belize Apr, 2014 subtidal coarse tropical no sequence Caribbean Sea -088.0819444 The copyrightholderforthispreprint(whichwasnot

W Atlantic, 16.8030556, . NL14CBC-02 / 1385O Stilbonema majum Carrie Bow Cay, Belize Apr, 2014 subtidal coarse tropical no sequence Caribbean Sea -088.0819444

W Atlantic, 16.8030556, NL14CBC-03 / 2348E Stilbonema majum Carrie Bow Cay, Belize Apr, 2013 subtidal coarse tropical no sequence Caribbean Sea -088.0819444 certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

W Atlantic, 16.8030556, NL14CBC-04 / 1385A Laxus oneistus Carrie Bow Cay, Belize March, 2014 subtidal coarse tropical Caribbean Sea -088.0819444 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

HGVE13-03 / Laxus cf. Mediterranean 42.808366, medium Elba, Italy Aug, 2013 subtidal temperate 811Aand 703A cosmopolitus Sea 10.141731 coarse

Laxus cf. Mediterranean 42.808366, medium HGVE13-04 / 811B Elba, Italy Aug, 2013 subtidal temperate no sequence no sequence cosmopolitus Sea 10.141731 coarse The copyrightholderforthispreprint(whichwasnot

Mediterranean 35.515099, . NL15CR-01 / 2349A Catanema "crete 1" Crete, Greece Aug, 2015 subtidal medium fine temperate Sea 23.989591

Mediterranean 35.515099, NL15CR-02 / 2349B Catanema "crete 1" Crete, Greece Aug, 2015 subtidal medium fine temperate no sequence Sea 23.989591 certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

Eubostrichus cf. Mediterranean 39.516388, MA16_017 / 2344_L Mallorca, Spain May, 2016 subtidal fine temperate topiarius Sea 2.741944 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

Eubostrichus cf. Mediterranean 39.757055, MA16_025 / 2344_J Mallorca, Spain May, 2017 subtidal fine temperate topiarius Sea 3.219944

Piran16SB_N6 / Eubostrichus cf. Mediterranean 45.517656, Piran, Slovenia August, 2016 subtidal fine temperate 2349_L topiarius Sea 13.568039 The copyrightholderforthispreprint(whichwasnot

Eubostrichus cf. Mediterranean 42.808366, medium . HGVE13-01 / 1045A Elba, Italy Aug, 2013 subtidal temperate topiarius Sea 10.141731 coarse

Eubostrichus cf. Mediterranean 42.808366, medium HGVE13-02 / 1045B Elba, Italy Aug, 2013 subtidal temperate topiarius Sea 10.141731 coarse certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

Accession Accession doi: Habitat Habitat Accession no. Sample ID / Location Sampling no. no. https://doi.org/10.1101/728105 Host species Sampling region Ocean region (marine (sediment Climate symbiont 16S Library (Long/Lat) date host 18S host COI zone) grain size) gene rRNA gene gene

W Atlantic, 16.823702, BZ15NL_032 / 2348I Eubostrichus dianeae Twin Cays, Belize May, 2015 subtidal fine tropical Caribbean Sea -88.106056 a CC-BY-NC-ND 4.0Internationallicense ; this versionpostedAugust7,2019.

Harrington Sound, 32.324038, medium BER15JZ649 / 2349H Eubostrichus dianeae N Atlantic May, 2015 subtidal subtropical Bermuda -64.738267 coarse

1139 1140 The copyrightholderforthispreprint(whichwasnot . certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

1141 doi: 1142 Table S2. Morphological details and measurements taken from microphotographs of live specimens of Paralaxus cocos, P. bermudensis https://doi.org/10.1101/728105 1143 and P. columbae.

Paralaxus cocos Paralaxus bermudensis Paralaxus columbae

Holotype Paratype male Paratype female Holotype Paratype male Holotype Paratype male Paratype female a CC-BY-NC-ND 4.0Internationallicense ; n 1 3 3 1 1 1 1 3 this versionpostedAugust7,2019.

Length 3822 3135 -4602 4336 - 5069 4115 4336 3568 2801 3935 - 4669

a 73.5 56.9 - 83.6 72.3 - 94.3 68.6 66.7 64.9 50.9 66 - 78

b 42.9 28.5 - 43 36.7 - 43.4 35.5 34.7 32.4 28 35.8 - 44.5

c 31.8 27.3 - 41.7 40.8 - 42.9 39.2 34.7 31.8 25.9 36.4 - 39.7 The copyrightholderforthispreprint(whichwasnot .

max. diameter 52 50 - 55 46 - 66 60 65 55 55 55 - 60

pharynx length 89 98 - 110 100 - 130 116 125 110 100 103 - 110

tail length 120 108 - 115 102 - 118 105 125 112 108 100 - 120

anal/cloacal diameter 50 50 - 55 40 - 42 60 65 55 52 50 - 55 certified bypeerreview)istheauthor/funder,whohasgrantedbioRxivalicensetodisplaypreprintinperpetuity.Itmadeavailableunder bioRxiv preprint

doi:

Paralaxus cocos Paralaxus bermudensis Paralaxus columbae https://doi.org/10.1101/728105

Holotype Paratype male Paratype female Holotype Paratype male Holotype Paratype male Paratype female

c' 2.4 2.0 - 2.3 2.6 - 2.8 1.75 1.9 2 1.96 1.8 - 2.2

vulva % n.a. n.a. 44 - 45 n.a. n.a. n.a. n.a. 44.5 - 48 a CC-BY-NC-ND 4.0Internationallicense ; amphidial fovea width 14 14 - 15 11 bis 12 15 12 12 12 12 bis 13 this versionpostedAugust7,2019.

pharynx bulbus l/w 30/28 32-37/32-35 30-45/27-43 37/35 42/40 35/32 30/35 30-35/37-45

bulbus % of pharynx l 33.7 31.8 - 34.5 30 - 34.6 31.9 32 31.8 30 29 - 33

diam. end of pharynx 47 46 - 57 46 - 50 60 65 55 54 55 - 60

spicula length, arc/chord 61/50 60-65/50-51 n.a. 57/45 63/45 60/50 60/50 n.a. The copyrightholderforthispreprint(whichwasnot .

gubernaculum length 33 31 - 35 n.a. 30 30 30 32 n.a.

velum % of tail length 47 43 - 46 n.a. 40 36 35.7 37 n.a.

1144 1145 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1146 SI Figures

1147 1148 Fig. S1. Scanning electron micrographs of Paralaxus cocos nov. spec.female (a) Anterior in en 1149 face view, (b) buccal field, (c) detail of symbiotic coat, (d) midbody region, (e) tip of tail. SEM 1150 (ann annulation, bac symbiotic bacteria, bf buccal field, cs cephalic setae, fa fovea amphidalis, 1151 ls labial setae, na non-annulated tip of tail, scs subcephalic setae, ss somatic setae. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1152 1153 Fig. S2. Light micrographs of Paralaxus cocos nov. spec. male. (a) anterior end (b) anterior 1154 end surface view (c) anterior end, optical section (d) midbody region, optical section of 1155 bacterial coat (e) midbody region, surface view of bacterial coat (f) cloacal region (g) tip of tail. 1156 LM of live specimen. (ann annulation, b pharyngeal bulbus, bac symbiotic bacteria, cc cephalic 1157 capsule, cgl caudal gland, cs cephalic setae, d ductus ejaculatorius, fa fovea amphidalis, gb 1158 gubernaculum, scs1 subcephalic setae of first circle, scs2 subcephalic setae of second circle, sp 1159 spiculum, vel velum). bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1160 1161 Fig. S3. Light micrographs of the reproductive system of Paralaxus cocos nov. spec. male. (a) 1162 junction of testis and vas deferens (b) junction of vas deferens and ductus ejaculatorius; (c) 1163 posterior end of ductus ejaculatorius. LM of live specimen. (de ductus ejaculatorius, s sperm, t 1164 testis, vd vas deferens) bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1165 1166 Fig. S4. Light micrographs of Paralaxus bermudensis nov. spec. male. (a) anterior end, optical 1167 section (b) posterior end (c) anterior tip, surface view (d) midbody region. LM of preserved 1168 specimen. (ann annulation, b pharyngeal bulbus, cc cephalic capsule, cs cephalic setae, fa 1169 fovea amphidalis, gb gubernaculum, na non-annulated tip of tail, nr nerve ring, scs1 1170 subcephalic setae of first circle, scs2 subcephalic setae of second circle, sp spiculum, vel 1171 velum). 1172 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1173

1174 1175 Fig. S5. Light micrographs of the posterior end of Paralaxus columbae nov. spec. (a – c) male 1176 (a) sperm in testis (b) cloacal region (c) tail (d) female tail. LM of preserved specimen (bac 1177 symbiotic bacteria, cgl caudal gland, d ductus ejaculatorius, gb gubernaculum, na non- 1178 annulated tip of tail, sp spiculum, vel velum). bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1179

1180 1181 Fig. S6. Light micrographs of Paralaxus sp. “heron” (a) anterior region (b) anterior end (c) 1182 midbody region with dissociated bacterial coat. LM of preserved specimen. (b pharyngeal 1183 bulbus, bac bacteria, cc cephalic capsule, fa fovea amphidialis, g gut, gso glandular sense 1184 organ). bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1185 1186 Fig. S7. Light micrographs of Paralaxus “oahu” male. (a) total view (b) anterior region (c) 1187 optical section of anterior region (d) cloacal region. LM of live specimen (ann annulation, b 1188 pharyngeal bulbus, bac bacteria, cc cephalic capsule, cs cephalic setae, gb gubernaculum, gso 1189 glandular sense organ, scs1 subcephalic setae of first circle, sp spiculum). bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1190

1191 1192 Fig. S8. Light micrographs of Paralaxus species and its morphological adaptation to its 1193 symbiont coat. (a) Paralaxus columbae nov. spec, juvenile. Bacterial coat starting at the 1194 transition from coarse to finer annulation; (b – d) Transition in adult specimens (b) P. cocos 1195 nov. sp. (c) P. columbae nov. sp. (d) P. bermudensis nov. sp. The transition points are marked 1196 by white arrows. LM of preserved specimen. (bac bacteria, bbac begin of bacterial coat) 1197 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1241 Fig. S9. Host 18S rRNA phylogeny of extended dataset (non-collapsed tree version) 1242 The tree was calculated using IQ-Tree. Support values are given in the following order: SH-aLRT 1243 support (%) / aBayes support / ultrafast bootstrap support (%). Species described in this study 1244 are highlighted in bold. Species from Armenteros et al. (2014b) are highlighted in italic and 1245 light grey, with quote marks. Provisional working names for undescribed genera or species are 1246 given in quotes. The scale bar represents average nucleotide substitutions per site. 1247 1248 1249

1250 1251 Fig. S10. Host COI phylogeny of extended dataset (non-collapsed tree version) 1252 The tree was calculated using IQ-Tree. Support values are given in the following order: SH-aLRT 1253 support (%) / aBayes support / ultrafast bootstrap support (%). Species described in this study 1254 are highlighted in bold. Species from Armenteros et al. (2014b) are highlighted in italic and 1255 with quote marks. Outgroup in light grey and italic. Provisional working names for undescribed bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1256 genera or species are given in quotes. The scale bar represents average nucleotide 1257 substitutions per site. bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1258 1259 1260 bioRxiv preprint doi: https://doi.org/10.1101/728105; this version posted August 7, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1261 Fig. S11. Symbiont 16S rRNA phylogeny (non-collapsed tree version) 1262 The tree was calculated using IQ-Tree. Support values are given in the following order: SH-aLRT 1263 support (%) / aBayes support / ultrafast bootstrap support (%). Symbiont sequences generated 1264 in this study are highlighted in bold. Outgroup in light grey. Provisional working names for 1265 undescribed genera or species are given in quotes. The scale bar represents average 1266 nucleotide substitutions per site. 1267 1268