Litinium gludi sp. nov. (Nematoda, Oxystominidae) from Kermadec Trench, Southwest Pacific Ocean

Recent work on the taxonomy of nematodes in Southwest Pacific Ocean trenches has led to the discovery of taxa which so far appear to be restricted to the oceans’ deepest environments. Here, Litinium gludi sp. nov. is described based on specimens obtained from a deep basin within the Kermadec Trench at 9540 m water depth. The new species differs from other species of the genus in having a conico-cylindrical tail, papillose labial sensilla, and heartor leaf-shaped amphideal fovea. Both SSU and LSU phylogenetic analyses provide strong support for the placement of the new species within a clade containing both Thalassoalaimus and Litinum, and within Oxystomininae, which is consistent with the structure of the female reproductive system with only one posterior ovary in this subfamily. Our molecular analyses also indicate that the new species is most closely related to Thalassoalaimus despite lacking a caudal capsule, the main trait characterizing the latter genus, and despite being most morphologically similar to Litinium, particularly in the unusual shape of the amphideal fovea. However, given the changing definitions of the closely-related genera Thalassoalaimus and Litinium in recent years, available GenBank sequences may have been misidentified, which makes the interpretation of molecular phylogenetic analyses problematic. Given the current morphological definitions of Litinium and Thalassoalaimus, we choose to place the new species within Litinium, despite the apparently contradictory findings of molecular phylogenetic analyses. The placement of Cricohalalaimus in a clade with Thalassoalaimus and Litinium in both SSU and LSU analyses indicates that this genus should be placed within the Oxystomininae and not the Halalaiminae as in the current classification. This new proposed grouping is consistent with variation in the structure of the female reproductive system, a feature which appears more taxonomically informative than amphid shape for subfamily-level classification.


Introduction
The Oxystominidae Chitwood, 1935(order Enoplida Filipjev, 1929) is among the most common nematode families in deep-sea environments (Gambi et al. 2003;Bik et al. 2010a;Miljutin et al. 2010). This largely marine group is characterized by a thin elongated body, the outer labial and cephalic sensilla in separate circles, and the buccal cavity without teeth (Lorenzen 1981). The family includes three subfamilies: the Oxystomininae Chitwood, 1935, Halalaiminae De Coninck, 1965and Paroxystomininae De Coninck, 1965, which are diff erentiated based on the amphid shape, the structure of the female reproductive system and the presence/structure of the precloacal supplements. The Oxystomininae comprises fi ve genera, all of which except Oxystomina Chitwood, 1935 are exclusively marine; the presence of only the posterior ovary is considered a holapomorphy establishing the holophyly of the Oxystomininae (Lorenzen 1981). The Halalaiminae originally comprised a single genus, Halalaimus de Man, 1888, which is characterized by an elongated longitudinal amphideal aperture, an holapomorphic trait establishing the holophyly of the subfamily. The genus Cricohalalaimus Bussau, 1993 was later included in the subfamily by Smol et al. (2014) based on the similar structure of the amphideal aperture; this genus, however, is also characterized by the presence of a single posterior ovary in females, a feature of the Oxystomininae. There is no holapomorphy for the Paroxystominae, which comprises Paroxystomina Micoletzky, 1924and Maldivea Gerlach, 1962(Lorenzen 1981. Phylogenetic analyses based on SSU sequences do not provide support for the monophyly of the Oxystominae or Halalaminae (Bik et al. 2010a), and no molecular sequences are yet available for the Paroxystomininae.
The majority of marine free-living nematode genera have wide water depth distributions spanning both coastal and deep-sea environments (> 200 m water depth), but some are mostly found in the deep sea (e.g., Acantholaimus Allgén, 1933, Metasphaerolaimus Gourbault & Boucher, 1981, while others are found exclusively in continental slope and/or deeper environments (e.g., Bathyeurystomina Lambshead & Platt, 1979, Thelonema Bussau, 1993, Manganonema Bussau, 1993. Phylogenetic analyses based on SSU data suggest repeated and recent interchanges between the deep sea and shallow water environments, with habitat transitions thought to be frequent for free-living nematodes (Bik et al. 2010b). In Halalaimus and Oxystomina, shallow water species appear to have evolved from deep water ancestors (Bik et al. 2010b); however, within the Oxystominidae, Cricohalalaimus Bussau, 1993 is the only genus with a distribution strictly restricted to the deep sea.
Limited core sampling has been conducted in hadal trenches to date (> 6000 m water depth), resulting in scant knowledge of nematode systematics in the deepest parts of the world's oceans. Taxonomic work on Tonga and Kermadec trench material has led to the discovery of new species and genera which so far appear to be restricted to hadal environments (Leduc 2015(Leduc , 2016. More recently, sampling within the deep Kermadec Trench axis provided core samples from the trench's deepest basins, and the examination of specimens from a site located at 9540 m depth led to the discovery of a species of nematode, Litinium gludi sp. nov., which is described here. Phylogenetic relationships of this new species within the Oxystominidae are investigated using SSU and D2-D3 of LSU rDNA sequences.

Study area and sampling
The Kermadec Trench extends from approximately from 26 to 36°S near the northeastern tip of New Zealand's North Island, Southwest Pacifi c Ocean. Sediment samples were obtained from the deep axis of Kermadec Trench at 9540 m depth during RV Tangaroa cruise TAN1711 in December 2017. Sediments were obtained using a USNEL-type box corer (dimensions: 0.5 × 0.5 × 0.5 m, 0.125 m 3 capacity). Subsamples were obtained using a cut-off syringe with 29 mm internal diameter to a depth of 10 cm. Subsamples for morphological analyses were sliced into 1 cm layers and fi xed in 10% buff ered formalin. Nematodes were extracted from the remaining sediments by Ludox fl otation and transferred to pure glycerol (Somerfi eld & Warwick 1996). Species descriptions were made from glycerol mounts using diff erential interference contrast microscopy and drawings were made with the aid of a camera lucida. Measurements were obtained using an Olympus BX53 compound microscope with cellSens Standard software for digital image analysis.
Subsamples for molecular analyses were obtained as described above but the samples were frozen at -80°C instead of being fi xed in formalin. In the laboratory, frozen sediment samples were thawed overnight, then sieved trough a 20 μm mesh to retain nematodes. Nematodes were extracted using the Ludox fl otation method (Somerfi eld & Warwick 1996) and sorted under a dissecting microscope. One juvenile Litinium gludi sp. nov. specimen was isolated and transferred to a temporary slide to confi rm its identity. This specimen was identifi ed based on the characteristic shape of the cephalic region, cephalic sensilla, amphids, and tail.
All measurements are in μm (unless otherwise stated), and all curved structures are measured along the arc. Type specimens are held in the NIWA Invertebrate Collection, Wellington, New Zealand.

List of abbreviations
a = body length/maximum body diameter b = body length/pharynx length c = body length/tail length c' = tail length/anal or cloacal body diameter cbd = corresponding body diameter L = total body length; n, number of specimens V = vulva distance from anterior end of body %V = V/total body length × 100

DNA extraction, PCR and sequencing
Following observation under a compound microscope, the specimen was transferred to lysis buff er and kept frozen at -80°C prior to molecular analyses. DNA was extracted by the method of Zheng et al. (2002) with minor modifi cations (i.e., the nematode was not cut prior to being transferred to the lysis buff er solution). The DNA extract was stored at -20°C until used as PCR template.

Sequence alignment and phylogenetic inference
The ribosomal DNA SSU and LSU D2-D3 sequences were deposited in GenBank under accession numbers MW209715 and MW209714, respectively. The placement of the new SSU and D2-D3 of LSU sequences was investigated through a phylogenetic analysis of the representative genera of the Oxystominidae, as well as the Rhaptothyreidae Hope &Murphy, 1969 andOncholaimoidea Filipjev, 1916, which have been shown to be closely related (Bik et al. 2010a;Leduc et al. 2018), and rooted using Ironidae de Man, 1876 sequences (all SSU sequences > 1300 bp except four ca 800 bp, and D2-D3 of LSU sequences > 600 bp).
DNA sequences were aligned using MUSCLE (Edgar 2004a(Edgar , 2004b with default parameters. Phylogenies were built in Geneious ver. 10.2.6 (http://www.geneious.com, Kearse et al. 2012). MrModelTest ver. 2.3 (Nylander 2004) in conjunction with PAUP* ver. 4.0b10 (Swoff ord 2002) were used to select the best model using the Akaike Information Criterion. The substitution model [GTR (general time-reversible) + I (proportion of invariable sites) + G (gamma distribution)] was selected as the best-fi t model for SSU alignments (1522 bp) and LSU alignments (722 bp), respectively. The trees were run with chain length of 1 100 000, and burn-in length of 100 000. The perimeter fi les from multiple runs were inspected for chain convergence in Tracer ver. 1.5 (Rambaut & Drummond 2007), and the trees were edited in FigTree ver. 1.4.2 (http://tree.bio.ed.ac.uk/software/fi gtree) and PowerPoint. These analyses were also conducted with PhyML ver. 3.0 using the default settings in Geneious ver. 10.2.6. The substitution model GTR, the NNI (default, fast) topology search and 1000 bootstrap replicates (Guindon et al. 2010) were selected for building the tree. Phylum Nematoda Diesing, 1861Order Enoplida Filipjev, 1929Suborder Ironina Siddiqi, 1983Superfamily Ironoidea de Man, 1876 Family Oxystominidae Chitwood, 1935 Diagnosis (emended from Smol et al. 2014) Body elongated and very thin at anterior end. Anterior sensilla in three separate circles, second and third circles clearly separated; inner labial sensilla papilliform or setiform, outer labial sensilla usually setiform (except in some species of Litinium), cephalic sensilla setiform. Buccal cavity narrow, tubular or funnelshaped and without teeth. Sexual dimorphism in amphid shape sometimes present. Only orthometanemes with very short caudal fi laments present. Pharynx inserts into body cuticle in region of buccal cavity; however, cephalic capsule not well developed. Posterior section of pharynx with undulating outline. Females didelphic-amphidelphic with antidromously refl exed ovaries or monodelphic-opisthodelphic. Males diorchic with opposed testes or only one anterior testis. Position of caudal glands variable.

Other valid genera
Litinium Cobb, 1920Nemanema Cobb, 1920Thalassoalaimus de Man, 1893Wieseria Gerlach, 1956 Litinium Cobb, 1920 Diagnosis (modifi ed from Tchesunov et al. 2014) Circles of six inner and six outer labial papillae (< 2 μm) or setae (≥ 2 μm) situated close together on anterior end, subapically, with circle of four cephalic papillae or setae posterior to circles of inner and outer labial sensilla. Amphideal fovea situated between circles of outer labial and cephalic sensilla. Amphideal fovea varies in shape between species and may diff er in males and females of same species: may be ovoid with anterior round aperture, horseshoe-like or crescent contoured or more complex. Only one posterior antidromously refl exed ovary present; vulva shifted anteriorly. Tail never clavate, more or less short, cylindrical with rounded tip, occasionally conical or conico-cylindrical with pointed tip; terminal caudal capsule (thick cuticular lining at tip of tail) absent or weakly developed.

Remarks
The amphid of some species of Litinium Cobb, 1920 is described as being "horseshoe-shaped". In the new species, our observations indicate that it is the superimposition of the amphideal aperture contour over the larger amphideal fovea contours that creates the appearance of a horseshoe-shaped structure. Whether this is also the case in other species of Litinium remains to be clarifi ed.
There has been some confusion regarding the defi nitions of Litinium and Thalassoalaimus. The diagnoses provided by Smol et al. (2014) indicate that the two genera diff er in the shape of the amphids (pocketshaped amphideal fovea and slit-like aperture in Thalassoalaimus vs horseshoe-or heart-shaped fovea and round aperture in Litinium) and presence (Thalassoalaimus) or absence (Litinium) of a caudal capsule. However, we note that the pocket-shaped amphideal fovea and slit-like aperture in  Table 1 Diagnosis Litinium gludi sp. nov. is characterised by a slender body (a = 58-63), a body length of 895-1066 μm, papilliform labial sensilla, short cephalic sensilla 1.5-2.0 μm or ~0.2 cbd long, a large heart-or leafshaped amphideal fovea 67-88% cbd wide, a smaller amphideal aperture with two central longitudinal ridges, males with dimorphic sperm, spicules 1.1-1.4 times cloacal body diameter, a short and simple gubernaculum, and two precloacal setae.

Etymology
The species is named after Professor Ronnie Glud, leader of the HADES-ERC trench project and covoyage leader.

Males
Body slender, cylindrical, widest at level of nerve ring, tapering slightly towards anterior extremity. Cuticle smooth, without ornamentation; somatic setae absent, except for sparse sublateral setae on tail. Cephalic region demarcated by slight constriction near posterior edge of amphideal fovea. Inner labial sensilla not observed; six outer labial papillae, ~0.5 μm long, located slightly anterior to amphideal fovea and four short cephalic sensilla, 1.5-2.0 μm or ~0.2 cbd long, situated far posteriorly at level of posterior edge of amphideal fovea. Amphideal fovea large, heart-or drop-shaped, with lightly cuticularized outline; amphideal aperture smaller, oval-or drop shaped, with two lightly cuticularized longitudinal central ridges spanning long axis of amphideal fovea. Buccal cavity narrow, cylindrical, with cuticularized portion 3-5 μm deep. Pharynx muscular, cylindrical, surrounding posterior portion of buccal cavity, widening towards posterior extremity but not forming true bulb. Nerve ring located slightly anterior to middle of pharynx. Secretory-excretory system not observed. Cardia 2-4 μm long, partially embedded in intestine.

Female
Similar to males but without any somatic setae on tail. Reproductive system with single posterior genital branch and refl exed ovary. Oocyte up to 160 × 10 μm. Vulva located far anteriorly at about ⅓ of body length from anterior extremity. Vagina perpendicular to body surface; vaginal glands not observed.

Molecular phylogenetic relationships
Near full-length SSU (1516 bp) and D2-D3 of LSU sequences (756 bp) were obtained for Litinium gludi sp. nov. The Oxystominidae did not form a monophyletic clade in the consensus SSU tree due to the placement of Rhaptothyerus typicus Hope & Murphy, 1969 and oncholaimoid sequences among the oxystominid sequences (Fig. 5). Neither the Halalaiminae (represented by Halalaimus and Oxystomina) nor the Oxystomininae (all other oxystominid genera) were recovered as monophyletic. Litinium  Fig. 5. Bayesian tree inferred from SSU sequences, aligned using the MUSCLE alignment algorithm under the general time-reversible (GTR) + I (proportion of invariable sites) + gamma distribution (G) model. The new sequence is shown in bold font on a grey background. Posterior probabilities (left) and bootstrap values (right) greater than or equal to 50% are given on appropriate clades. The scale stands for substitutions per site. -: less than 50% bootstrap support; # : sequence FJ040500 is labelled Thalassoalaimus pirum in GenBank but is labelled Litinium pirum here to refl ect the new classifi cation of this species proposed by Martelli et al. (2017).
European Journal of Taxonomy 748: 138-154 (2021) 150 Fig. 6. Bayesian tree inferred from D2-D3 of LSU sequences, aligned using the MUSCLE alignment algorithm under the general time-reversible (GTR) + I (proportion of invariable sites) + gamma distribution (G) model. Posterior probabilities (left) and bootstrap values (right) greater than or equal to 50% are given on appropriate clades. The new sequence is shown in bold font on a grey background. The scale stands for substitutions per site.
gludi sp. nov. was placed in a well-supported clade (100% posterior probability and 82% bootstrap support) with sequences of Thalassoalaimus. The similarity between SSU sequences of Litinium gludi sp. nov. and Thalassoalaimus sp. was 94-95% with a diff erence of 69 in 1508 nucleotides with 4 gaps (MN250034), 83 in 1504 nucleotides with 1 gap (HM564634) and 43 in 798 nucleotides with 1 gap (MN250041). The Litinium gludi sp. nov. + Thalassoalaimus clade was part of a wider, strongly supported oxystominid clade including all sequences of Litinium and Cricohalalaimus (100% posterior probability and bootstrap support), although neither Litinium nor Thalassoalaimus were monophyletic. The similarity between SSU sequences of Litinium gludi sp. nov. and Litinium sp. was 87-94% with a diff erence of 84 in 1513 nucleotides with 5 gaps (FJ040500), 116 in 1516 nucleotides with 3 gaps (HM564629), 113 in 1514 nucleotides with 2 gaps (HM564649), and 46 in 786 nucleotides with 1 gap (MK626773). Sequences of Halalaimus formed a strongly supported (100% posterior probability and bootstrap support) monophyletic sister group to the Oncholaimoidea Filipjev, 1916, the latter also forming a strongly supported monophyletic group (100% posterior probability and bootstrap support). The sequences of Oxystomina formed a strongly supported group (100% posterior probability and bootstrap support), however this genus was not monophyletic due to the inclusion of a sequence of Nemanema Cobb, 1920 in the clade. The placement of another sequence of Nemanema as a sister group to the Halalaimus + Oncholaimoidea clade was poorly supported (56% posterior probability and <50% bootstrap support).
The topology of the D2-D3 of LSU consensus tree was similar to that of the SSU tree, with the Oxystominidae again not recovered as monophyletic (Fig. 6). Sequences of Halalaimus and Oxystomina each formed strongly supported monophyletic clades (100% posterior probability and bootstrap support). As in the SSU tree, Litinium, Cricohalalaimus, and Thalassohalaimus formed a strongly supported clade (100% posterior probability and bootstrap support), with Litinium gludi sp. nov. most closely related to Thalassoalaimus (100% posterior probability and 98% bootstrap support). The similarity between LSU sequences of Litinium gludi sp. nov. and Thalassoalaimus sp. was 92% with a diff erence of 118 in 664 nucleotides with 7 gaps (HM564880); the similarity between LSU sequences of Litinium gludi sp. nov. and Litinium sp. was 71-72% with a diff erence of 188 in 671 nucleotides with 13 gaps (HM564858), 184 in 671 nucleotides with 12 gaps (HM564874) and 183 in 671 nucleotides with 13 gaps (HM564875).

Discussion
The topologies of our SSU and D2-D3 of LSU trees are very similar to the phylogeny of Bik et al. (2010a), which recovered four sub-clades: Oncholaimoidea, Thalassoalaimus + Cricohalalaimus + Litinium, Oxystomina, and Halalaimus. Bik et al. (2010a) noted that Halalaimus was consistently recovered as a long-branch clade, which may have a destabilizing eff ect on the internal tree topology. The inclusion of Rhaptothyreus typicus in the present study, which was also recovered as a long-branch clade, may have had a similar destabilizing eff ect, and the placement of this taxon in relation to the Oxystominidae as well as wider Enoplida remains uncertain (Leduc et al. 2018). Molecular phylogenetic analyses have not yet provided support for the monophyly of the Oxystomininae or Halalaiminae (Bik et al. 2010a;present study).
The morphology of Litinium gludi sp. nov. is consistent with other species of Litinium due to the shape of the amphids, the posterior position of the cephalic setae and the lack of a caudal capsule. However, both SSU and LSU trees indicate that the new genus is most closely related to Thalassoalaimus. The latter genus is similar to Litinium in the structure and position of cephalic sensilla, but diff ers from Litinium in having a caudal capsule. It should be noted that most of the sequences of Thalassoalaimus in GenBank were not identifi ed to the species level, and that given the recent changes in the defi nitions of Litinium and Thalassoalaimus (Tchesunov et al. 2014;Martelli et al. 2017), some or possibly all of the sequences of Thalassoalaimus from GenBank included in our phylogenetic analyses may in fact belong to Litinium or vice versa. For example, the SSU sequence of Thalssoalaimus pirum, the only sequence of Thalassoalaimus identifi ed to species level in GenBank, has since been transferred to Litinium by Martelli et al. (2017) (see Fig. 4). This uncertainty puts into question the apparent close relationship between the new species and Thalassoalaimus in molecular analyses. Therefore, given the current morphological defi nitions of Litinium and Thalassoalaimus, we choose to place the new species within Litinium, despite the apparently contradictory fi ndings of molecular analyses.
Our phylogenetic analyses provide strong support for the placement of the new genus within a clade containing both Thalassoalaimus and Litinum, and within the Oxystomininae. Interestingly, the placement of Cricohalalaimus together with Thalassoalaimus and Litinium indicates that this genus, which is characterized by confl icting features used to defi ne the Halalaiminae (elongated amphideal aperture) and Oxystomininae (female with single posterior ovary), should be placed within the Oxystomininae and not in the Halalaiminae as is currently the case (Smol et al. 2014). An analogous change to the classifi cation of the Rhabdocoma Cobb, 1920(Enoplida, Trefusiidae Gerlach, 1966(De Ley & Blaxter 2004) was suggested by Shi & Xu (2017), who found that SSU phylogenetic relationships supported the classifi cation of this genus with the Halononchinae Wieser & Hopper, 1967, a classifi cation consistent with variation in the structure of the female reproductive system, and not in the Trefusiinae Gerlach, 1966, as previously suggested based on buccal morphology. Similarly, in the case of Cricohalalaimus, it appears that the shape of the amphideal aperture is not as informative as the structure of the female reproductive system for subfamily-level classifi cation, and we therefore propose that Cricohalalaimus be moved to the Oxystomininae.