Four new species and further records of Dorvilleidae (Annelida, Polychaeta) from deep-sea organic substrata, NE Atlantic

Eight species of Ophryotrocha and one of Parougia were identified from organic substrata (wood and alfalfa) sampled at the Gulf of Cadiz and Western Iberian Margin (NE Atlantic). Morphological examination and molecular phylogenetic analyses, based on the nuclear gene H3 and the mitochondrial gene 16S, indicate the presence of four species new to science: Ophryotrocha chemecoli sp. nov., O. nunezi sp. nov., O. geoffreadi sp. nov. and Parougia ougi sp. nov. The geographic and/or bathymetric distribution is extended for four previously known species: O. cantabrica, O. hartmanni, O. mammillata and O. scutellus. Another species may also have its distribution extended, pending the molecular confirmation of its identity: O. lipscombae. Full descriptions and figures are given for all the new species and, when justified, also for the previously known ones. The original description of O. scutellus is amended. New DNA sequences are given for eight of the nine species studied here. The ecology and geographic distribution of the reported species is discussed.

are the main dorvilleid genera occurring in the deep sea (Claparède & Mecznikov 1869;Jumars 1974;Wolf 1986;Thornhill et al. 2012;Levin et al. 2013). Together, these three genera include more than 48% of the species within the dorvilleid family. Ophryotrocha is by far the most specious genus, including at present 77 species (Read & Fauchald 2020), most of them (~80%) described from deep waters. The other two genera, Parougia and Exallopus, include nineteen and five species, respectively (Read & Fauchald 2020; Yen & Rouse 2020). These genera, and Ophryotrocha in particular, are known to thrive in all kinds of organically enriched habitats (e.g., Wiklund et al. 2009Wiklund et al. , 2012Ravara et al. 2015;Taboada et al. 2016;Zhang et al. 2017;Vedenin et al. 2020). As new deep-sea organic-enriched places are studied (e.g., whale, wood or other organic falls), the rate of novel species of Ophryotrocha and related genera greatly increase when compared to the new records of already known species (e.g., Wiklund et al. 2009Wiklund et al. , 2012Taboada et al. 2013;Salvo et al. 2014;Ravara et al. 2015). This denotes the incipient and fragmentary knowledge of these ephemeral habitats, thus limiting a comprehensive understanding of the diversity, distribution and dispersal of the species they support. For the NE Atlantic Ocean, only a couple of studies of dorvilleids on organic falls have previously been published, both reporting five Ophryotrocha species, three of which were new to science (Wiklund et al. 2009;Ravara et al. 2015). The present study adds three new species of Ophryotrocha and one new species of Parougia to the NE Atlantic organic-falls fauna. Five previously known species of Ophryotrocha are also reported for the first time in the Gulf of Cadiz (Moroccan margin) and the West Iberian margin, thus extending their geographic and bathymetric distributions. Full descriptions and images are given for the new species and, whenever considered necessary, also for some of the previously known species, based on the specimens studied here. Observed discrepancies regarding the original descriptions and distributions of the species are discussed. The original description of O. scutellus is amended based on the specimens studied here and a re-examination of some specimens from the original material (Wiklund et al. 2009). Molecular data is provided for eight out of the nine taxa studied.

Material and methods
Most of the biological material examined in this study was obtained from wood and alfalfa substrata experimentally deployed in the Gulf of Cadiz (NE Atlantic) from 2007 to 2009, with the aim of studying colonization and dispersal processes in deep-sea reduced habitats (Cunha et al. 2013). The experimental colonization devices (CHEMECOLI) contained two types of organic substrata -dried alfalfa grass and Douglas fir wood cubes -enclosed by a 2 mm mesh net and were deployed at three mud volcanoes (Mercator, Mèknes and Darwin), within a depth range of 354-1100 m, for a period of one to two years. Carbonate cubes were used as control samples. Table 1 includes site coordinates and cruises involved in the deployment and recovery of the colonization devices and Fig. 1 locates the sites on the map. Details on the experimental devices and sample processing can be found in Cunha et al. (2013). Additional material was retrieved from a small piece of sunken wood collected using the ROV hydraulic arm during a survey of a seemingly inactive pockmark field at the Estremadura Spur, West Iberian Margin, at a depth of 327 m (the sample was stored in the ROV biobox during ascent and immediately processed onboard). All the polychaete specimens were fixed either in 96% ethanol or in a 4% formaldehyde solution.
After a preliminary sorting into putative species, DNA of selected ethanol fixed specimens was extracted using a DNeasy Blood and Tissue Kit (Qiagen) following the protocol supplied by the manufacturer. About 450bp of the mitochondrial gene 16S were amplified with the primers Ann16SF 5′-GCGGTATCCTGACCGTRCWAAGGTA-3′ (Sjölin et al. 2005) and 16SbrH 5′-CCGGTCTGAACTCAGATCACGT-3′ (Palumbi 1996), and about 350bp of the nuclear coding gene H3 were amplified with the primers H3F 5′-ATGGCTCGTACCAAGCAGACVGC-3′ (Colgan et al. 2000) and H3R 5′-ATATCCTTRGGCATRATRGTGAC-3′ (Colgan et al. 2000). Attempts were made to amplify a fragment of the mitochondrial gene COI, but with low success. Only three sequences were obtained with the primers polLCO 5′-GAYTATWTTCAACAAATCATAAAGATATTGG-3′ as RAVARA A. et al., New Dorvilleidae from the NE Atlantic well as PolyLCO and PolyHCO 5′-TAMACTTCWGGGTGACCAAARAATCA-3′ (Carr et al. 2011). PCR mixtures contained 1 μl of each primer (10 μM), 2 μl of template DNA and 21 μl of Red Taq DNA Polymerase 1.1X MasterMix (VWR). The temperature profile was as follows: 96°C for 240 s, followed by (94°C for 30 s, 50°C for 30 s then 72°C for 60 s)*35 cycles, followed by 72°C for 480 s. PCR purification was performed using a Millipore Multiscreen 96-well PCR Purification System, and sequencing was performed on an ABI 3730XL DNA Analyser (Applied Biosystems) at the Natural History Museum Sequencing Facility, using the primers mentioned above. All sequences were deposited in GenBank (Table 2).    Overlapping sequence fragments were merged into consensus sequences using Geneious (Kearse et al. 2012). The sequences obtained in this study were aligned together with sequences from Genbank (Table 2) using MAFFT (Katoh et al. 2002) with default settings, provided as a plug-in in Geneious. Only 16S and H3 sequences were used in the analyses due to the greater coverage of those sequences. The 16S alignment consisted of 526 characters and the H3 alignment of 341 characters. In total, 77 terminal taxa were used in the phylogenetic analyses, with nine from Parougia, 57 from Ophryotrocha, ten from other Dorvilleidae genera and Eunice pennata (Eunicidae) as root ( Table 2). The program jModelTest (Posada 2008) was used to assess the best model for each partition (16S and H3) with BIC, which suggested GTR+I+G as the best model for both genes. The data was partitioned into the two genes (16S and H3), and the evolutionary model mentioned above was applied to each partition. The parameters used for the partitions were unlinked. Bayesian phylogenetic analyses (BAs) were conducted with MrBayes ver. 3.2.6 (Ronquist et al. 2012). Analyses were run three times for 10 000 000 generations. Of these, 2 500 000 generations were discarded as burn-in. The tree files were interpreted with FigTree ver. 1.4.2 (available from http://tree.bio.ed.ac.uk/software/figtree/). The consensus tree is presented in Fig. 2.
Detailed morphological observations were carried out with stereo and compound microscopy. Stereo microscope images were taken with a Canon EOS1100D camera. Jaw and chaetal morphology was examined and imaged with a Zeiss Axioplan 2 imaging light microscope (Carl Zeiss, Oberkochen, Germany), equipped with a DP70 Olympus camera (Olympus Corp., Tokyo, Japan), after mounting the specimen on a slide preparation using Aquamount (Gurr) liquid. Measurements of body width were recorded from the widest part of the body (anterior region), excluding parapodia and chaetae, and body   length excludes anal cirri. A 3D scatter plot was used to illustrate the relationship between the body length, width and the number of chaetigers for each species, including data from type material (Fig. 3). The plot was generated using the R program (R Core Team 2020) and the "scatterplot3d" package (ver. 0.3-41; Ligges & Mächler 2003). Whenever necessary, type material was requested from museum collections for comparison. The holotypes of the newly described species are deposited at the Natural History Museum of London (NHMUK). The remaining material is deposited at the Biological Research Collection (Marine Invertebrates) of the Department of Biology of the University of Aveiro (COBI-DBUA). DNA voucher terminology follows Pleijel et al. (2008).
All the material mentioned in this paper was examined by us. The notation 'Other material' refers to non-type material collected by us, whereas 'Additional material' refers to material loaned from other institutions.

Description
Relatively large specimens in comparison to the majority of the species of Ophryotrocha (Fig. 3). Largest complete specimen 5.4 mm long and 0.8 mm wide for 35 chaetigers. Body compressed dorsoventrally, wider sub-anteriorly and tapering posteriorly (Fig. 4A). A transversal band of cilia present in the middle of each segment. Prostomium broadly rounded, without eyes. Antennae and palps digitiform, sometimes retracted and with a more globulous appearance, subequal in length. Peristomium achaetous, with two rings subequal in length to the following segments. Mouth opens ventrally on the first peristomial ring (Fig. 4B). Mandibles rod-like; cutting plates straight and toothed; apophyses triangular, extending laterally right beneath the tip of the mandibula (Fig. 4B, D). Maxillae of P-type with a pair of forceps and seven pairs of free denticles (D1-7) (Fig. 4E); forceps and D1 comb-like with 13 large teeth alternating with two small teeth; D2-3 shovel-like, with long distal fang and serrated edge; D4-6 with coarsely serrated edge (~11 teeth); D7 shovel-like, broader than the others and with smaller and more numerous teeth (~40). Parapodia uniramous, with broadly conical acicular lobes, long digitiform pre-chaetal lamellae (in median parapodia), conical sub-acicular lobes slightly shorter than pre-chaetal lamellae, conical ventral cirri shorter than sub-acicular lobes, and ovoid to piriform dorsal cirri as long as pre-chaetal lamellae in anterior parapodia and greatly increasing in size posteriorly (Fig. 4A). Protruding acicula of sub-acicular lobes very thin and short. Supra-acicular chaetae simple, serrated, tapering distally to a thin pointed hook (Fig. 4F). Sub-acicular chaetae compound with serrated bidentate blades; secondary tooth tiny and difficult to see (Fig. 4G). Pygidium with terminal anus, two cirriform anal cirri inserted laterally and one short median cirrus (Fig. 4C).

Remarks
Ophryotrocha cantabrica was originally described from the Cantabrian Sea at 72-74 m depth (Nuñez et al. 2014. The holotype and paratypes were examined during this study and found to be similar in size (Fig. 3) and morphology to the material from GoC and WIM. Only the sub-acicular chaetae of the GoC and WIM specimens seem to have somewhat thicker and shorter blades (Fig. 5D) but this was not considered significant enough to establish a different species. Nuñez et al. (2014) described and illustrated the mandibulae of this species as being distally slightly bifid without serration on the anterior edge. However, a thorough examination of all the specimens (including type material) revealed that the mandibles are often broken at the tip (Fig. 5A), giving the appearance illustrated in the original description (Nuñez et al. 2014: fig. 3a). In fact, when not broken, the mandibles present a ridge of small teeth on its distal edge (Fig. 4D). Also in the forceps and denticle 1 of the maxillary apparatus, the large teeth were found to alternate with two small teeth (Figs 4E, 5B). The phylogenetic analysis segregates this species from the others within the tree (Fig. 2), but there are no DNA sequences from the type material to compare with. The two sequenced specimens were from WIM, the K2P value from the H3 alignment between the two specimens is 0.00 and the K2P value to the nearest species in the tree, O. batillus Wiklund et al., 2012, is 0.10. This study extends the geographic and bathymetric distribution of O. cantabrica to WIM and GoC, where it was the most abundant species, and to a depth of 698 m. It is noteworthy that all our specimens were associated with wood samples only and not with alfalfa grass.

Etymology
The species name is an allusion to the colonization devices (CHEMECOLI) deployed in the GoC within the scope of the project CHEMECO ("Monitoring colonization processes in chemosynthetic ecosystems"). This species occurred in five of those devices containing wood and alfalfa, at all three mud volcanoes where the experiment was carried out. CHEMECOLI stands for "CHEMosynthetic Ecosystem COlonization by Larval Invertebrates".

Description
Relatively small specimens, 0.64-1.49 mm long and 0.16-0.26 mm wide for 10-19 chaetigers (Fig. 3). Body dorso-ventrally flattened, with the same width in the anterior half, gradually tapering posteriorly (Fig. 6A). Prostomium short and broadly rounded with a flattened anterior rim, without eyes (Fig. 6A). Antennae and palps long, digitiform; antennae inserted mid-dorsally on the prostomium; palps inserted laterally. Peristomium achaetous, with two short rings. Jaw apparatus heavily sclerotized, well visible through the specimen body . Mandibles rod-like, with straight and clearly dentate anterior edge; apophyse long, slightly surpassing the cutting edge, narrow and curved laterally (Fig. 6C). Maxillae of P-type (Fig. 6D); forceps falcate, comb-like with about 13 teeth, wider with thicker teeth on the right side and narrower with thinner teeth on the left side; eleven pairs of free denticles (D1-11), D1 similar to forceps with thick teeth on both sides, D2 and D3 broad, shovel-like, D4 to D11 shovel-like, narrower, usually directed inwards giving a rhomboid appearance to the maxillae; carrier-like structure with a toothed ridge on each side near the forceps, posterior end with a fimbriate handle. Parapodia  6E); pre-chaetal lamellae of the median parapodia long, conical; dorsal and ventral cirri conical (dorsal longer than ventral); sub-acicular lobes conical, similar in size to pre-chaetal lamellae, with a short, needle-like acicula protruding. Chaetae long and stiff; supra-acicular chaetae simple, slightly flattening and tapering distally to a fine tip, very lightly serrated (Fig. 6F); sub-acicular chaetae compound, with bifurcate, sub-distally serrated shafts and falcate, very lightly serrated blades (Fig. 6G). Pygidium with terminal anus, a pair of cirriform anal cirri and a very short, rounded median stylet.

Remarks
Ophryotrocha chemecoli sp. nov. is very close at molecular (Fig. 2) and morphological levels to the species O. scutellus Wiklund, Glover & Dahlgren, 2009, O. lusa Ravara, Marçal, Wiklund & Hilário, 2015and O. batillus Wiklund et al., 2012. Ophryotrocha scutellus and O. lusa are known to occur in the NE Atlantic (Wiklund et al. 2009;Ravara et al. 2015), whereas O. batillus occurs in the NE Pacific (Wiklund et al. 2012). Like O. chemecoli sp. nov., all of those species have a dorso-ventrally flattened body, a long parapodial pre-chaetal lobe and roughly rhombus-shaped maxillae. Also the chaetae of O. scutellus and O. batillus are very similar to those of the new species. However, O. scutellus differs from O. chemecoli sp. nov. by the larger body size, the flatter prostomium, the much longer pre-chaetal lobe and the mandibular morphology of the adult specimens; whereas O. batillus differs in the very large size of the parapodia and chaetae, the flatter and wider prostomium and, to some extent, also the mandibular shape. As in O. chemecoli sp. nov., O. lusa has eleven pairs of maxillary free denticles, while the other Ophryotrocha species usually have only eight or less (Paxton 2004(Paxton , 2009Ravara et al. 2015). Furthermore, the size of the specimens of the new species, as well as the neurochaetae and the mandibulae shape, are similar to those of younger specimens of O. lusa, making it difficult to distinguish between the two species. The adult specimens of O. lusa, however, have very different neurochaetae and mandibular morphology (see Ravara et al. 2015). Such variation in the mandibular morphology was not observed in the specimens of O. chemecoli sp. nov., although globular masses, indicating reproductive maturity, were detected at the base of median parapodia of some larger specimens. In the phylogenetic analysis, O. cantabrica also groups in the same clade as O. chemecoli sp. nov. (Fig. 2). At the morphological level, these species are very different, having only the morphology of the chaetae in common.

Description
Relatively small specimens compared to most species of Ophryotrocha. Most of the larger specimens are damaged or incomplete and were not measured. Measured specimens 0.73 to 1.46 mm long and 0.19 mm wide, for 17 to 25 chaetigers (Fig. 3). Body dorso-ventrally flattened, wider anteriorly and tapering posteriorly. Prostomium broadly rounded (Fig. 7A), without eyes. Antennae and palps short, cirriform; antennae inserted mid-dorsally on the prostomium; palps inserted laterally. Peristomium achaetous, with two rings of length similar to the following segments. Jaw apparatus brown, well visible through the specimen body. Mandibles rod-like, with straight, serrated cutting edge (with around 8 teeth in smaller specimens, Fig. 7B) and weakly sclerotized apophyses, shorter than cutting edges. In larger specimens, the cutting edge teeth are usually worn and the apohyses are enlarged longitudinally (Fig. 7E). Maxillae of P-type (also after moulting) with asymmetric forceps; left forceps narrow and finely toothed (up to 35? teeth), right forceps wider and coarsely toothed (~15 teeth) (Fig. 7F); seven free denticles (D1-7; Fig. 7C, F), D1 comb-like coarsely toothed, similar to right forceps, D2 shovellike with an outer larger tooth and coarse teeth, D4-6 shovel-like with smaller teeth slightly increasing in number anteriorly, D7 shovel-like, much broader than the others; carrier-like structure with a toothed ridge on each side at the base of the bifurcation (Fig. 7F) and a posteriorly fimbriate handle (Fig. 7C). Parapodia uniramous, with conical acicular lobes, inconspicuous pre-chaetal lamellae, long sub-acicular lobes and conical dorsal cirri inserted sub-distally on the parapodia (Fig. 7D); ventral cirri absent. Subacicular lobes with a short needle-like protruding acicula (Fig. 7H). Chaetae relatively short and stiff; supra-acicular chaetae simple, tapering distally, smooth or very finely serrated, up to 7 per fascicle (Fig. 7I); sub-acicular chaetae compound with sub-distally serrated shafts and falcate, lightly serrated blades (Fig. 7G), up to 7 per fascicle. Pygidium with terminal anus, a pair of short cirriform anal cirri (spherical and almost inconspicuous in smaller specimens) and a median papilla.

Remarks
The original description of O. hartmanni given by Huth (1933) is based mostly on cytological features. The only morphological details included reference to the specimens' average length (4 -5 mm), to the body shape as being torpedo-like and to the position of the sperms and oocytes (sperms in the first three chaetigers and oocytes from the fourth chaetiger onwards). The specimens examined by Huth were collected at Plymouth and kept in laboratory conditions, but no information is given on type material or its deposition. Parenti (1961) provided a re-description of the same species based on material from Roscoff also kept in laboratory conditions. There is no mention to a re-examination of type material or other material from the type locality, but Parenti (1961) confirmed some of the cytological observations made by Huth (1933) and provided a complete morphological description, highlighting the differences between this and other species. Again, there is no indication regarding the deposition of the examined material. Åkesson (1973) studied the reproduction and larval morphology of O. hartmanni using specimens from Plymouth and Roscoff aquaria and from the harbours of Malaga (S Spain) and concluded that interbreeding of the three strains produced fertile progeny in all combinations, and also in subsequent generations. This denotes the widespread distribution of this species.
Only three specimens of O. hartmanni were found in the present study, two of which are not in good condition. The smallest (13 chaetigers, 0.76 mm long), and better preserved, and the larger (approximately 20 chaetigers, 2.3 mm long) specimens were mounted in a permanent slide where the jaw apparatus, parapodia and chaetae could be properly examined (Fig. 8). The molecular analysis performed with the posterior part of the third specimen failed, thus preventing molecular comparison. Except for the presence of small dorsal cirri on the parapodia (Fig. 8D), stated as absent for O. hartmanni, the GoC specimens have all the morphological diagnostic characteristics described by Parenti (1961 (2021) plates with tiny pointed teeth (around 14 teeth in the smaller specimen spread along all the cutting edge, including around the inner peak, Fig. 8A-B; in the larger specimen, the teeth are worn out and the inner peak is almost smooth, Fig. 8E). The juvenile and adult forms of maxillae are also similar to those of O. hartmanni. The P-type forceps have a large distal tooth, the distal half is comb-like with alternating large and small teeth, and the posterior part has a ridge of tiny teeth (Fig. 8C). The K-type forceps have bifid tips strongly bent inwards (more gently curved in O. hartmanni, as illustrated in Parenti 1961: fig. II-4); the sub-apical tooth has a straight spine superiorly and is finely denticulated inferiorly (Fig. 8F). Denticles 1 to 7 (D1-7) are similar in shape for P-and K-type maxillae. Denticle 1 is similar in shape to the P-type forceps; D2 is shovel-shaped but also with a large tooth on the inner edge; D3 is shovel-shaped, narrow, with coarse teeth; and D4-7 are shovel-shaped, wider and with smaller teeth. According to Parenti (1961), O. hartmanni is particularly abundant in muddy sediments rich in organic detritus. This study provides a new record of O. hartmanni for the North Atlantic (at the GoC) and extends the bathymetric distribution of the species to 1100 m depth. Ophryotrocha cf. lipscombae Lu & Fauchald, 2000 Fig. 9 Ophryotrocha lipscombae

Description of the material examined in this study
Body robust and compact (Fig. 9A), with the same width throughout, only tapering at the pygidium. Largest specimen 2.5 mm long and 0.43 mm wide for 38 chaetigers (Fig. 3). All segments with a transversal band of cilia. Prostomium short, broadly rounded, slightly trilobed (more evident in smaller specimens; Fig. 9B), without eyes. Antennae and palps spherical, very small, may appear to be absent (only visible under the microscope). Peristomium achaetous, with two rings half the length of the following segments. Delimitation between the prostomium and the first ring of the peristomium almost imperceptible. Jaw apparatus dark and relatively large (Fig. 9B). Mandibles rod-like; in larger specimens, the anterior end is cut in two levels, with the inner lower level toothed (teeth often worn) (Fig. 9C-E); in smaller specimens, the anterior end of mandibulae is mainly straight with small elevations at the outer tips and distinctly toothed (Fig. 9I). Maxillae with a pair of falcate forceps (Fig. 9F), seven pairs of free denticles (D1-7) (Fig. 9G-H) and five pairs of accessory plates (Fig. 9; the fifth plate was lost when detachjing the denticles, Fig. 9G-H)); forceps with a main fang followed by 1-2 large and several very small teeth (Fig. 9J); D1 comb-like, with about 18 large and small teeth alternating (Fig. 9F); distal denticles plate-like, with up to 8 large and small teeth irregularly alternating (Fig. 9G-H (2021) with broadly conical acicular lobes, sub-acicular lobes and ovoid dorsal cirri fused to the acicular lobe (Fig. 9N); ventral cirri apparently absent. Aciculae slightly protruding in all parapodia. Sub-acicular lobes with harpoon-like protruding acicula with acute tip and subdistal tooth (Fig. 9M). Supra-acicular chaetae simple, subdistally serrated, tapering abruptly into a large main fang (Fig. 9K-L). Sub-acicular chaetae compound with very short serrated blades abruptly tapering into a main fang; heterogomph shafts distally serrated (Fig. 9O). Pygidium with terminal anus and 2 very small ovoid anal cirri.

Remarks
Ophryotrocha lipscombae is only known from its original description based on specimens collected from Woods Hole (NW Atlantic) at a depth of 1830 m (Lu & Fauchald 2000). This species is considered to be unique among the dorvilleids based on the presence of five pairs of accessory plates associated with the denticles of the maxillae (Lu & Fauchald 2000). Other distinctive characteristics are the presence of teeth below the main fang of the maxillary forceps and the clear fusion between the prostomium and the peristomium. All these characteristics are present in the specimens from the Estremadura Spur (WIM, NE Atlantic) and the single specimen from Mercator MV (GoC, NE Atlantic) examined in this study. The minor differences registered for the specimens from the NW Atlantic concern features that may have been overlooked or misinterpreted, such as: the absence of distal teeth on the mandibulae of larger specimens, which may be worn (in our specimens these teeth are mostly worn and the ones that are present are difficult to detect); the absence of prostomial palps (in our specimens the antennae and palps are tiny and could only be detected under the optical microscope); the absence of dorsal cirri on parapodia (in our specimens the dorsal cirri appear almost entirely fused with the acicular lobes and thus are difficult to distinguish (Fig. 9N); figure 1d of the original description illustrates a similar structure); and the tips of the supra-and sub-acicular chaetae, which in our specimens are more abruptly tapered into a distinct main fang (only visible under high magnification). The phylogenetic analysis segregates this species from the others within the tree (Fig. 2). The two sequenced specimens were from WIM, the K2P value from the H3 alignment between the two specimens is 0.00 and the K2P values to the nearest species in the tree, O. maculata and O. hartmanni, is 0.09. However, there are no DNA sequences for the West Atlantic O. lipscombae specimens to compare with. Furthermore, it was not possible to examine the type material that is deposited in the Smithsonian National Museum of Natural History. If the similarity at the morphological and molecular level is confirmed, then this study extends the geographic and upper limit of the bathymetric distribution of O. lipscombae to the NE Atlantic and to a depth of 327 m.

Remarks
This species was recently described from an experimentally deployed mammal carcass for a similar amount of time (approximately 18 months) at the Setúbal Canyon (WIM) (Ravara et al. 2015). The morphological identification was confirmed with molecular analyses (Fig. 2). This study extends its distribution to the Gulf of Cadiz where it occurred associated with experimentally deployed wood and alfalfa substrata. Curiously, O. mammillata was not found in the wood-fall collected at Estremadura Spur (WIM), a site very close to its type locality. It is worth mentioning here that in the GoC only five specimens of O. mammillata were retrieved, compared to 198 specimens previously reported from the Setúbal Canyon (Ravara et al. 2015).

Ecology and distribution
NE Atlantic: from Setúbal Canyon (West Iberian Margin) to the Gulf of Cadiz. Found in experimentally deployed organic falls (mammal carcasses, wood and alfalfa substrata), at a depth of 354-1100 m (Ravara et al. 2015;present study).

Etymology
The species is named after Dr Jorge Nuñez, who described several of the currently known Ophryotrocha species and first examined the specimens from the Cantabrian Sea. Dr Nuñez has always been very helpful in our requests for assistance on these taxa.

Remarks
Ophryotrocha nunezi sp. nov. can be distinguished from all other species by the combination of four characteristics: bi-articulated palps, winged mandibulae, falcate maxillary forceps with internal tooth, and clavate anal cirri. According to the phylogenetic analysis (Fig. 2), six described species of Ophryotrocha and one of Exallopus are placed in the same clade as O. nunezi sp. nov.: O. longidentata Josefson, 1975, O. nauarchus Wiklund et al., 2012, O. magnadentata Wiklund et al., 2012, O. longicollaris Wiklund et al., 2012, O. globopalpata Blake & Hilbig, 1990, O. flabella Wiklund et al., 2012and E. jumarsi Blake, 1985(Josefson 1975Blake 1985;Blake & Hilbig 1990;Wiklund et al. 2012). All these species of Ophryotrocha have bi-articulated palps, but, unlike the others, O. longicollaris does not have winged mandibulae. Furthermore, of all of them, only O. flabella has toothed K-type forceps and only O. magnadentata has clavate anal cirri. The latter species also has sub-acicular lobes sustained by a compound falciger similarly to O. nunezi sp. nov. However, all the other characteristics of both species are different. There are two more species of Ophryotrocha with bi-articulated palps and winged mandibulae (not included in the phylogenetic analysis), O. paragerlachi Brito &Nuñez, 2003 andO. platykephale Blake, 1985, but both these species have long cirriform anal cirri and different parapodial and chaetal morphology (Blake 1985;Brito & Nuñez 2003). Two other species, O. mammillata Ravara, Marçal, Wilund &Hilário, 2015 andO. wubaolingi Miura, 1997, have bi-articulated palps but different mandibulae and anal cirri as well as other morphological features (Miura 1997;Ravara et al. 2015). Ophryotrocha labidion Hilbig & Blake, 1991, from the US Atlantic coast, have bi-articulated palps, winged mandibulae and clavate anal cirri, but strongly differs in the maxillary, parapodial and chaetal morphology (Hilbig & Blake 1991). Another species from the West Atlantic coast, O. atlantica Hilbig & Blake, 1991, with winged mandibulae, falcate forceps with teeth and clavate anal cirri, have simple palps and different maxillary and chaetal morphology (Hilbig & Blake 1991). Paxton & Morineaux (2009) gave a brief description of an undetermined species (Ophryotrocha sp.) reported from the Mid-Atlantic Ridge (Menez Gwen vent field, 865 m depth), that although larger in size, had a similar appearance to this species but with simple palps and mandibulae with a smooth cutting edge. The genus Exallopus also includes species with bi-articulated palps, although this is not the case for the species E. jumarsi, which RAVARA A. et al., New Dorvilleidae from the NE Atlantic  (2021) further differs from O. nunezi sp. nov. by having the antennae weakly annulated, the P-type maxillae and by the presence of strongly modified chaetae in the first chaetigerous segment. This latter characteristic is the main morphological difference between the genera Exallopus and Ophryotrocha. Nevertheless, based on the molecular analyses done to date, the two genera cannot be separated at the molecular level (Fig. 2;Wiklund et al. 2012). The lack of molecular data for the type species of the genus Exallopus (Exallopus cropion Jumars, 1974) has prevented concluding the synonymy of the two genera. Four other non-described species (Ophryotrocha 'Seep 3, 4 and 5' and Exallopus 'seep', Fig. 2)

Description (amended)
Size of WIM and GoC specimens varies within 1.55-2.70 mm long and 0.24-0.39 mm wide for 18-24 chaetigers. Skagerrak specimens are larger, up to 3.60 mm long and 0.75 mm wide for 31 chaetigers (Fig. 3). Body dorso-ventrally flattened, with similar width throughout the body, abruptly ending with pygidium in smaller specimens (Fig. 11A) or tapering slightly at posterior end in larger ones. Prostomium broadly rounded, dorso-ventrally flattened, with a transverse ridge between the antennae, without eyes (Fig. 11A). Antennae and palps long, digitiform; antennae inserted mid-dorsally on the prostomium; palps inserted laterally. Peristomium achaetous, with two rings slightly narrower and shorter than the RAVARA A. et al., New Dorvilleidae from the NE Atlantic following segments. Jaw apparatus heavily sclerotized, well visible through the specimen body, usually with an apparent rhombus shape (Fig. 11A-B). The morphology of mandibles and maxillary forceps varies with the specimen size (Fig. 12A-J). Mandibles rod-like; smaller specimens with straight and clearly dentate anterior end and long apophyse, well surpassing the cutting edge, with a diagonal connection to the shaft (Fig. 12A); with growth, the teeth wear out (Fig. 12B) and the cutting edge becomes short and more curved forward, without teeth, the apophyse becomes thicker with an almost vertical connection to the shaft (Fig. 12D-E). Maxillae of P-type; forceps falcate, comb-like, slightly wider with up to 20 large teeth on the right side, and narrower with up to 26 thinner teeth on the left side (Fig. 12F-G); with growth, the teeth of the left forcep become irregular (Fig. 12H) resulting in a clear alternation in size in larger specimens (Figs 11D, 12I-J); eleven free denticles (D1-11), D1 similar to forceps (always with even teeth), D2 to D11 shovel-like, D4 to D11 usually directed inwards (Fig. 11H-I); carrierlike structure with a toothed ridge on each side near the forceps (see details in Fig. 11D) and with a posteriorly fimbriate handle (Fig. 11D). Parapodia uniramous (Fig. 11C); pre-chaetal lamellae of median parapodia very long, cirriform; dorsal and ventral cirri digitiform, long (dorsal longer than ventral); sub-acicular lobes conical, about two-thirds the length of pre-chaetal lamellae in smaller specimens, becoming shorter in larger specimens , with a needle-like acicula (Fig. 11E). Chaetae long and stiff; supra-acicular chaetae simple, slightly flattening and tapering distally to a fine tip, very lightly serrated, 7 per fascicle (Fig. 11G); sub-acicular chaetae compound with bifurcate, sub-distally serrated shafts, and falcate, very lightly serrated blades (Fig. 11F), 7-9 per fascicle. Pygidium with terminal anus, a pair of cirriform anal cirri and a very short (almost imperceptible) median stylet.

Remarks
This species was originally described from a minke whale carcass deployed at a depth of 125 m off Sweden and organically enriched sediments beneath a fish farm in Norway, at a depth of 104 m (Wiklund et al. 2009). Later, seven specimens of the same species were retrieved from an experimentally implanted cow carcass at the Setúbal Canyon (WIM), 1000 m depth (Ravara et al. 2015). The present study extends the distribution of O. scutellus to GoC where it occurred associated with experimentally deployed alfalfa and wood substrata and control samples (carbonate cubes), at a depth of 354-1100 m. The specimens from the GoC and WIM are overall smaller than the ones originally described from Sweden and Norway (Fig. 3) and exhibit some variation in the mandibular and maxillary morphology, apparently associated with growth (Fig. 12). However, the larger specimens of the southern locations entirely match the morphology of the northern ones. The morphological identification was furthermore confirmed with molecular analyses for both the larger and the smaller specimens. The specimen from GoC (DBUA0002287.01) sequenced here falls among previously published O. scutellus sequences (Genbank accession numbers GQ415506 and KP731544-48) with within-species K2P values from the H3 alignment of 0.009-0.01, and a K2P value of 0.10 to the nearest species in the tree, O. chemecoli sp. nov. A similar variability in length and corresponding variation in the mandible morphology has earlier been described for other species, such as O. sadina and O. lusa (Ravara et al. 2015: figs 15, 25, respectively). Differing from what was stated in the original description, the specimens of O. scutellus studied here have eleven pairs of free denticles (instead of seven) in the maxillary apparatus, and the left forcep of the larger specimens have uneven teeth. These characters were also found in the specimens from off Sweden examined here (Figs 11-12). Thus, the original description is here amended accordingly.

European Journal of Taxonomy
The species is named after Dr Eivind Oug, who made a thorough revision and described several species within the genus Parougia.

Remarks
The genus Parougia includes, at present, nineteen nominal species (Yen & Rouse 2020). Five of these species (P. albomaculata (Åkesson & Rice, 1992), P. bermudensis (Åkesson & Rice, 1992), P. diapason Taboada, Bas & Avilla, 2015, P. eliasoni (Oug, 1978 and P. oregonensis Hilbig & Fiege, 2001) are represented in the phylogenetic reconstruction (Fig. 2) and are found to be molecularly different from the one studied herein (Oug 1978;Åkesson & Rice 1992;Hilbig & Fiege 2001;Taboada et al. 2015). Twelve species of Parougia are known to occur in chemosynthetic habitats: one in whale-falls in Antarctica (P. diapason, 10 m depth), eight in methane seeps in the NE Pacific (P. batia (Jumars, 1974), also in whale-falls, 587-1229 m depth; P. billiemiroae Yen & Rouse, 2020, 587 m depth;P. ceruleibohnorum Yen & Rouse, 2020, 588-1433  Rouse, 2020, 1991 m depth; P. jessieae Yen & Rouse, 2020, 1821-1845 and in the NE Pacific (P. wolfi Blake & Hilbig, 1990, 1545-2200 m depth) (Jumars 1974;Blake & Hilbig 1990; Yen & Rouse  (2021) 2020). Three other species are known from shallow waters of the North Atlantic: P. caeca (Webster & Benedict 1884), P. macilenta (Oug, 1978) and P. nigridentata (Oug, 1978) and another species was described from soft sediments of the southern ocean: P. furcata (Hartman, 1953) (Hartman 1953;Oug 1978;Webster & Benedict 1884). All of these species differ from P. ougi sp. nov. at the morphological level, especially in the maxillary apparatus. In P. ougi sp. nov. all free denticles are shovel-shaped with numerous small teeth of similar size, while in the majority of the other species, at least the posterior free denticles of the superior row have a large main tooth followed by several smaller ones. The only exception is P. jessieae, for which the free denticles are very similar to those of P. ougi sp. nov. Also, the mandibles of this species have smooth anterior margins as in P. ougi sp. nov., although with a prominent median peak and slightly different shape. However, P. ougi sp. nov. differs from P. jessieae by the coarsely serrated supra-acicular chaetae and the presence of furcate chaetae (absent in P. jessieae). Another species from a closely related genus, Ougia tenuidentis Wolf, 1986, described from shallow sandy sediments from Florida, presents the same kind of mandibulae as P. ougi sp. nov. and is otherwise very similar (Wolf 1986). But again, the free denticles of that species bear both small and large teeth. Ougia tenuidentis also differs from P. ougi sp. nov. in the basal plates of the maxillary superior row, which are serrated rather than composed of fused free denticles, and the presence of maxillary carriers, which are distinctive characteristics of the genus Ougia Wolf, 1986(Wolf 1986. Parougia ougi sp. nov. presents what apparently are basal plates on the inferior rows of maxillary denticles. Yen & Rouse (2020) recently referred to these structures as being possible moults, although their identification as either the inferior or a replacement row was considered to be ambiguous for some species. None of these hypotheses can be confirmed in the single specimen examined in this study. Also P. wolfi was originally described as having superior basal plates with double vertical rows of teeth fused to each other and to a horny base (Blake & Hilbig 1990). This structure may be equivalent to the presence of both inferior and superior basal plates, but further studies are needed to confirm this. Parougia wolfi clearly differs from P. ougi sp. nov. by the distinct shape of the anterior free denticles of the maxillary apparatus.

Ecology and distribution
NE Atlantic: Gulf of Cadiz (Moroccan margin). Found in experimentally deployed wood at a depth of 354 m.

Phylogenetic results
A total of 477 dorvilleid specimens were examined in this study: 284 from GoC and 193 from WIM (Table 3). These specimens were ascribed to eight Ophryotrocha and one species of Parougia. Four of these species are described here as new to science. Seventeen new DNA sequences (nine H3, five 16S and three COI) were obtained, ten for the four newly described species, six for the new records of three previously known species and one for a specimen herein referred as O. cf. lipscombae (Table 2). Although morphologically similar to O. lipscombae, known from Woods Hole (NW Atlantic), a putative extension of the species distribution to the NE Atlantic, combined with our inability to examine the type material and the absence of sequences from the original location to compare with the one obtained herein, all add uncertainty to the identification that must be considered with caution. Only the specimens of O. hartmanni failed to produce viable sequences. The Bayesian phylogenetic analyses (including the genetic markers 16S and H3) supported the species O. chemecoli sp. nov., O. nunezi sp. nov., O. geoffreadi sp. nov. and P. ougi sp. nov. as being new to science, and confirmed the identity of the previously known species O. mammillata and O. scutellus (Fig. 2). No sequences were previously known for the species O. cantabrica.
Similarly to other studies (Wiklund et al. 2012;Taboada et al. 2013;Salvo et al. 2014;Zhang et al. 2017), the phylogenetic analysis retrieved a well-supported 'Ophryotrocha' clade which also includes species from the genera Exallopus, Iphitime Marenzeller, 1902 andPseudophryotrocha Hilbig &Blake, 1991 RAVARA A. et al., New Dorvilleidae from the NE Atlantic ( Fig. 2), but the non-existence of DNA data for the type species of these genera prevents their synonymy with Ophryotrocha, as previously discussed by Wiklund et al. (2012) (Marenzeller 1902;Hilbig & Blake 1991). The 'Parougia' group, where the new species P. ougi sp. nov. is included, is well supported and sister to the 'Ophryotrocha' group. Within the Ophryotrocha clade, the tree shows a low degree of resolution with only 58.8% of the nodes having posterior probability support values of 95 or above. Still, the analysis retrieved the three main clades known within Ophryotrocha, namely 'hartmanni', 'lobifera' and 'labronica' (Salvo et al. 2014;Taboada et al. 2017;Zhang et al. 2017), along with a fourth unnamed clade previously defined by Wiklund et al. (2012) as clade 'D' (Fig. 2). All these clades are well supported, although the 'hartmanni' group seems to be paraphyletic, as already shown by Dahlgren et al. (2001).

Ecological results
Previous studies have suggested a possible bathymetric segregation of the fauna from reduced environments, with the shallow-water naturally reduced habitats being more similar to other shallowwater organic-rich sulphidic substrata, such as fish farms or sewage discharges (Wiklund et al. 2009;Danise et al. 2014;Taboada et al. 2016), while deep-water organic falls would have strong affinities with hydrothermal vents and cold seeps (Bennett et al. 1994;Smith 2006). This is not validated by our analysis, where close affinities between species of various habitat types, depths and biogeographical regions are patent in several clades (Fig. 2, see also Table 2). In fact, species from reduced environments are present in several clades, indicating that the adaptation to such environments happened multiple times independently, as already suggested by Thornhill et al. (2012) and Wiklund et al. (2012) for the dorvilleids and by Eilertsen et al. (2017) for the NE Atlantic ampharetids.
All the nine species reported herein were found in the colonization devices from one or more mud volcanoes of GoC, but only three species (O. cantabrica, O. chemecoli sp. nov. and O. cf. lipscombae) were also present in the sunken wood fragment sampled at WIM (Table 3). The geographic and bathymetric distribution is extended for four species (O. cantabrica, O. hartmanni, O. mammillata and O. scutellus). As mentioned before, the putative extension of O. lipscombae must be considered with caution. However, both O. scutellus and O. hartmanni were already known to have a wide distribution (from Norway to Portugal and from the English Channel to the Mediterranean, respectively) (Åkesson 1973;Ravara et al. 2015), compatible with the new records from the Iberian and Moroccan margins. Although uncommon, these wide geographic distributions are not unprecedented, and there are similar reports for other species (e.g., O. orensanzi Taboada et al., 2013in Paxton et al. 2017, O. bifida Hilbig & Blake, 1991and P. ceruleibohnorum Yen & Rouse, 2020. Given the known limited larval dispersal capability of Ophryotrocha species (Paxton & Åkesson 2007), a possible explanation for such wide   (2021) distributions is the existence of ephemeral or long-lasting spots of organic enrichment across those distribution areas, capable of sustaining viable populations of those species (Cunha et al. 2013). Species of Ophryotrocha are considered as omnivores or unspecialized opportunists, able to inhabit nonenriched environments and to rapidly increase their abundances when nutrient conditions become more favourable (Paxton et al. 2017). This seems to be the case for O. orensanzi, which was described from a whale-fall in W Antarctica and later reported from non-enriched sediments at the other side of the continent (Taboada et al. 2013;Paxton et al. 2017). The species O. cantabrica was first reported from non-enriched sediments and it is reported here as the dominant dorvilleid both in naturally occurring and experimentally deployed sunken wood at Estremadura Spur (327 m deep) and Mercator MV (354 m deep), respectively. These two shallower locations also showed overall the highest number of dorvilleids (177 specs in Mercator MV and 193 in Estremadura Spur), mostly in association with wood substrata colonised by the wood-borer bivalve Xylophaga W. Turton, 1822, known for its role in the breakdown of refractory organic materials into more readily available food sources (Voight 2015).
The genus Ophryotrocha was always represented by several species in all samples: Mercator MV presented the highest species diversity, with the presence of eight different species, two of which were only found at this location (O. nunezi sp. nov., Parougia ougi sp. nov.); at the deeper sites five species were retrieved in each mud volcano, represented by a total of 61 and 44 specimens, respectively, in Mèknes MV (698 m) and Darwin MV (1100 m). Three species, O. chemecoli sp. nov., O. mammillata and O. scutellus, were found at all three mud volcanoes, in wood and alfalfa substrata. Ophryotrocha chemecoli sp. nov. was also found at Estremadura Spur in sunken wood. Only one species, O. geoffreadi sp. nov., was found exclusively in the deeper mud volcanoes (Mèknes and Darwin), both in alfalfa and wood substrata. The outstanding diversity of Ophryotrocha taxa in organically-enriched, reduced environments is patent in the frequent occurrence of several sympatric congeners, also reported in previous studies (e.g., California methane seeps - Levin et al. 2003; an Italian harbour - Prevedelli et al. 2005; NE Atlantic and NE Pacific organic falls - Wiklund et al. 2009Wiklund et al. , 2012; Antarctic whale falls and eutrophic sediments - Taboada et al. 2013Taboada et al. , 2017; experimentally deployed organic falls in the NE Atlantic - Cunha et al. 2013;Ravara et al. 2015;this study). This co-existence is often explained by adaptive mechanisms leading to niche differentiation via specialization and resource partitioning (Cunha et al. 2013;Ravara et al. 2015 and references therein). Previous studies indicate that organically-enriched habitats are heterogeneous environments supporting diverse food sources including different bacterial communities that, in turn, feed distinct species of Ophryotrocha, as revealed by isotopic analyses (Levin et al. 2013). Resource partitioning (e.g., food, space occupancy) may arise from differences in body size, as appears to be the case in our samples, where the most abundant species are segregated by size (Fig. 14; see also Table 3  In most cases, other species are represented by a single or few specimens in each sample. Studies over time have revealed an overlap of successional stages in the colonization process of whale-falls (Ravara et al. 2015;Taboada et al. 2017) that may also explain the locally high species richness encountered in organically-enriched habitats. Both adaptive (e.g., niche differentiation) and non-adaptive mechanisms (founder effect, genetic drift, hybridization) have been invoked as contributors to the global speciation of Ophryotrocha in organic falls and other reducing environments (Cunha et al. 2013 and references therein). Life history traits of species of Ophryotrocha play an essential role in these processes but little is known about their reproductive strategies (e.g., fecundity, dispersal potential, population growth rates) or feeding behaviour (e.g., type and size of food particles). Although there are some studies based on live observations of single species of Ophryotrocha feeding behaviour (e.g., Mercier et al. 2014), more comprehensive studies are crucial to better understand the ecology and distribution of this genus.