New species of Eurythenes from hadal depths of the Mariana Trench, Pacific Ocean (Crustacea: Amphipoda)

Eurythenes S. I. Smith in Scudder, 1882 are one of the largest scavenging deep-sea amphipods (max. 154 mm) and are found in every ocean across an extensive bathymetric range from the shallow polar waters to hadal depths. Recent systematic studies of the genus have illuminated a cryptic species complex and highlighted the benefits of using a combination of morphological and molecular identification approaches. In this study, we present the ninth species, Eurythenes plasticus sp. nov., which was recovered using baited traps between the depths 6010 and 6949 m in the Mariana Trench (Northwest Pacific Ocean) in 2014. This new Eurythenes species was found to have distinct morphological characteristics and be a well-supported clade based on sequence variation at two mitochondrial regions (16S rDNA and COI). While this species is new to science and lives in the remote hadal zone, it is not exempt from the impacts of anthropogenic pollution. Indeed, one individual was found to have a microplastic fibre, 83.74% similar to polyethylene terephthalate (PET), in its hindgut. As this species has a bathymetric range spanning from abyssal to hadal depths in the Central Pacific Ocean basin, it offers further insights into the biogeography of Eurythenes . Eurythenes . Eurythenes plasticus nov. differ-entiated with article 2 of the mandible palp being narrow (instead of expanded), four protruding nodular spines on the inner plate of the maxilliped (versus three non-protruding), and straight ventral border of coxa 4 (opposed to curved). Eurythenes plasticus sp. nov. can be separated from E. magellanicus with a long gnathopod 1 palm (instead of short), a straight ventral border of coxa 4 (opposed to curved), a subquadrate posteroventral corner in epimeron 3 (instead of bearing a small tooth), and the rami of uropod 1 and 2 being subequal (opposed to uropod 2 outer ramus being shorter than inner ramus and uropod 1 outer ramus being longer than inner ramus). Eurythenes plasticus nov. can also be distinguished from E. aequilatus by its eyes with a variable width (opposed to constant width), the outer plate of maxilla 1 with 8/3 crown arrangement (instead of 9/3 arrangement), and a long gnathopod 1 palm (instead of short).


Introduction
While the deep sea is one of the largest ecosystems on Earth, it has traditionally been perceived as a homogenous environment, with few barriers to gene flow (Madsen 1961;Charette & Smith 2010). This led to the assumption that many deep-sea species are cosmopolitan, with several appearing to have large geographical and bathymetrical ranges (>3000 m; King & Priede 2008;Brandt et al. 2012;Jamieson et al. 2013). The deep sea, however, has a high degree of topographic complexity including mid-oceanic ridges, submarine canyons, seamounts, and subduction trenches, which could act as barriers. These barriers potentially restrain gene flow and promote allopatric speciation (Palumbi 1994). This cosmopolitan species concept has now been challenged on several occasions by genetic techniques, whereby widespread deep-sea species are in fact comprised of species complexes with several cryptic or pseudocryptic species (Garlitska et al. 2012;Cornils & Held 2014).
The lysianassoid amphipod, Eurythenes gryllus (Lichtenstein in Mandt, 1822), is a quintessential and abundant member of the deep-sea benthic community. Eurythenes gryllus has long been considered cosmopolitan with an extensive bathymetric range (184 to 8000 m), which spans the bathyal, abyssal, and hadal zones (Hessler et al. 1978;Ingram &Hessler 1987;Thurston et al. 2002). However, genetic diversity studies have indicated that E. gryllus is not a single species but a species complex (France & Kocher 1992;Havermans et al. 2013), with nuclear and mitochondrial DNA sequence data indicating the gryllus-complex to be composed of at least nine to twelve distinct clades (Havermans et al. 2013;Eustace et al. 2016;Havermans 2016). Our initial understanding of E. gryllus as a single cosmopolitan deep-sea species is reconceptualised when viewed as a species-complex. This provides a much more nuanced picture of their distribution, amphitropical at bathyal depths, and reveals a patchwork of distribution patterns with the complex's radiation. For example, Eurythenes maldoror d'Udekem d'Acoz & Havermans, 2015 (GEBCO 2015). Isobaths are added for every 1000 m and labelled between 5000 to 10,000 m.

Morphological Assessment and Digital Illustration
Whole specimens were photographed with a Canon EOS 750D DSLR camera, Tamron SP 90 mm f/2.8 VC USD Macro 1:1 VC Lens with polarising filter, and Falcon Eyes CS-730 copy stand and processed with Helicon Focus and Helicon Remote software (Helicon Soft). Body length was measured from the rostrum to the tip of telson. Appendages were dissected using a Wild Heerbrugg M8 stereomicroscope and imaged with a Leica DMi8 inverted microscope and DFC295 camera. Lengths of appendages and articles were measured following Horton & Thurston (2014) to provide consistency regardless of the degree of flexion. Images were converted into digital illustrations using Inkscape v0.92.2 (Coleman 2003;2009). Type and non-type specimens are deposited at the Smithsonian Institution National Museum of Natural History, Washington, D.C., USA (USNM).

Phylogenetics
Total genomic DNA was extracted from either the head or a pair of pleopods depending on size of the specimen using the Bioline ISOLATE II Genomic DNA Kit. Two partial regions of the mitochondrial DNA were amplified. The 16S (260 bp) was amplified with AMPH1 (France & Kocher 1996) and 'Drosophila-type' 16SBr (Palumbi et al. 2002) primers and COI (624 bp) was amplified with LCO1490 and HCO12198 (Folmer et al. 1994) primers. PCR protocols were as described in Ritchie et al. (2015). PCR products were purified enzymatically using New England Biolabs Exonuclease 1 and Antarctic Phosphatase and sequenced with an ABI 3730XL sequencer (Eurofins Genomics, Germany). Electropherograms were viewed and primers and any ambiguous sequences were trimmed in MEGA 7 (Kumar et al. 2016). Sequences were initially blasted using default parameters on NCBI BLASTn. COI sequences were translated into amino acid sequences to confirm that no stop codons were present. Nucleotide alignments with comparative sequences were made using MAFFT v7 (Table 2; Katoh et al. 2019). The optimal evolutionary models for each alignment were identified by model test in the phangorn 2.4.0 package (Schliep et al. 2017). The optimal Akaike Information Criterion and Bayesian Information Criterion indicated the HKY + I + G model for both alignments (Hasegawa et al. 1985). Phylogenetic relationships were inferred via the maximum-likelihood approach using PhyML v3.1 (Guidon et al. 2010) and the Bayesian approach using BEAST v1.8.4 (Drummond et al. 2012). Maximum-likelihood analyses were conducted with a neighbour-joining starting tree and using nearest neighbour interchange branch swapping using the model of sequence evolution and parameters estimated by PhyML. The stability of nodes was assessed from bootstrap support based upon 10,000 iterations. Bayesian analyses were performed for two independent runs of 40,000,000 generations sampling every 10,000 generations using the respective evolutionary models and an uncorrelated relaxed clock (Drummond et al. 2006). Outputs were assessed in Tracer v1.7 to ensure convergence (ESS < 200) (Rambaut et al. 2018) and combined in Log Combiner v1.8.4. The first 4,000,000 states were discarded. The maximum clade credibility tree was generated through Tree Annotator v1.8.4, viewed in FigTree v1.4.3, and annotated using Inkscape v0.92.2. Two independent methods were used to infer species delimitation on each dataset, specifically a Bayesian Poisson Tree Processes (bPTP) model (Zhang et al. 2013) and sequence divergence using the Kimura 2-parameter (K2P) distance model (Kimura 1980).

Sample Digestion and Analysis for Microplastic Ingestion
Preventive measures were taken to reduce and monitor for potential sources of contamination due to the ubiquity of microplastic fibres in the environment (Wesch et al. 2017). Samples were prepared and analysed in a clean laboratory with restricted access, where only one researcher, wearing a 100% clean lab coat at all times, was present conducting the experiment. Before any work session, benches were wiped with 70% ethanol on a 100% cotton cloth and allowed to dry fully. Only non-plastic equipment (glass and metal) were used to process the samples. Glass Petri dishes, graduated piston pipettes and test tubes were thoroughly washed with pre-filtered deionised water (DI), rinsed with acetone, covered with aluminium foil and allowed to dry at 70 °C in a drying oven. The digestion and filtration steps were conducted under a laminar flow cabinet (Purair, LS series, Air Science, USA LLC). The equipment and samples were covered wherever possible to minimize environmental exposure. Additionally, procedural blanks were run in parallel with samples to monitor environmental contamination. Meaning, a glass petri dish with a damped Whatman glass fibre filter was left open next to the microscope during the specimens' dissection (Murphy et al. 2016), while two empty glass tubes were processed as described below. The resulting three blanks filters were examined under a stereo microscope (Leica M205C, Leica Microsystems GmbH, Germany) to correct for potential air-borne and/or procedural plastic contamination.
Four E. plasticus sp. nov. specimens were selected for microplastic analysis: three juveniles (15.1, 15.6, and 23.1 mm body length) from 6865 m and one juvenile (15.6 mm body length) from 6949 m. Each specimen was individually rinsed with pre-filtered DI water and inspected under a stereo microscope (Leica M205C, Leica Microsystems GmbH, Germany), to ensure each specimen was free from external contamination. The hindgut was removed as described in Jamieson et al. (2019) and individually placed in 10 mL glass tubes. Aluminium foil was used to cover the tubes. After recording its wet mass, the hindgut was submerged in 10% m/v potassium hydroxide (KOH), using a volume at least three times greater than that occupied by the biological material (Foekema et al. 2013). The samples plus two procedural blanks (borosilicate tubes with 2 and 7 mL 10% KOH solution) were incubated for over a 36-hour period at 40 °C. After digestion, samples were left to cool inside a desiccator, following vacuum filtration through 0.6 μm glass fibre filters (Advantec Grade GA55, Advantec MSF Inc., Japan). Filters were individually placed onto a glass Petri dish until further microscopic inspection.
Once dried, glass fibre filters were examined under a stereo microscope. The physical appearance (e.g., colour, shape, size) of the putative particles (e.g., fibre, fragment) per filter was recorded. Said particles were then transferred onto gold platted slides (Thermo Fisher Scientific Inc., UK) for Fourier-transform infrared spectroscopy (FTIR) analysis. A Nicolet iN10 FTIR micro spectroscope (Thermo Fisher Scientific Inc., UK) was employed to obtain the particle's infrared transmittance spectra, using the liquid nitrogen cooled Mercury Cadmium Telluride detector. Results were then visualised and matched against a series of inbuilt reference spectra libraries using the instrument's software (OMNIC Picta v1.7) to determine the chemical identity of the analysed particles.

Phylogenetics and Species Delimitation Analysis
Three specimens of Eurythenes plasticus sp. nov. were successfully characterised across the two partial gene amplicons. The sequences have been annotated and deposited into GenBank (Table 2; 16S MT021437-39 and COI MT038070-72).
The phylogenetic relationship of E. plasticus sp. nov. within Eurythenes was investigated in separate 16S and COI datasets. These comparative datasets were constructed from sequences that are associated with either: type material, specimens identified high degree of confidence, or specimens from a known clade or undescribed lineage (Table 2; France & Kocher 1996;Escobar-Briones et al. 2010;Havermans et al. 2013;Ritchie et al. 2015;Eustace et al. 2016;Havermans 2016;Narahara-Nakano et al. 2017;Ritchie et al. 2017). For the 16S dataset, 26 individuals consisting of the eight species of Eurythenes and five genetic clades fit these criteria. For the COI dataset, 25 individuals consisting of seven species of Eurythenes and four genetic clades fit these criteria. Alicella gigantea Chevreux, 1899 was selected as the outgroup for both datasets. The 16S and COI datasets contained 191 and 394 positions of which 33 and 115 bases were parsimony-informative, respectively.
The Bayesian-based topology based on variation across 16S and COI is shown in Fig. 2. In general, the two topologies shared similar patterns and the differences were largely due to lacking both sets of sequences for a specimen. The COI topology showed E. plasticus sp. nov. to form a reciprocally monophyletic group. The 16S topology varied slightly with the inclusion of Eurythenes sp. (U40445; France & Kocher 1996) to the E. plasticus sp. nov. phylogroup. This Eurythenes sp. represents a singleton and recently distinguished as part of the species-level clade Eg7 (Havermans et al. 2013). In both topologies, Eurythenes plasticus sp. nov. was placed within a larger clade with E. magellanicus, E. aequilatus, and Eurythenes sp. 'PCT abyssal'. Eurythenes plasticus sp. nov. was consistently sister to E. magellanicus, with high support in the COI topology (0.99 posterior probability; Fig. 2B). Species delimitation analysis with bPTP for the COI datasets estimated the three specimens of E. plasticus sp. nov. to be the same species and distinct from all other Eurythenes taxon (mean: 14.33; acceptance rate: 0.0846; estimated number of species: 12 -17). The bPTP analysis of the 16S dataset did not delineated E. plasticus sp. nov. from E. magellanicus, E. andhakarae, E. sigmiferus, E. aequilatus, E. obseus, Eurythenes sp. 'PCT abyssal', and Eurythenes spp. Eg7 -9 (mean: 5.29; acceptance rate: 0.20456; estimated number of species: 3 -13).
With alternative delimitation method, the average K2P estimates of divergence between E. plasticus sp. nov. and E. magellanicus were 0.034 ± 0.007 for 16S and 0.074 ± 0.008 for COI. The levels of interclade divergence between E. plasticus sp. nov. and E. magellanicus were comparable to the levels of divergence that have been previously used to detect cryptic speciation within the gryllus-complex (Havermans et al. 2013;Eustace et al. 2016;Narahara-Nakano et al. 2017). Furthermore, the '4x' criterion was satisfied, whereby the interclade divergences were at least four times the maximum intraclade divergences (Birky et al. 2005).
Microplastics Three particles were observed between the four specimens. One particle was a 649.648 μm long, dark fibre extracted from the juvenile from 6949 m (Fig. 3). FTIR analysis determined this fibre to be 83.74% similar to polyethylene terephthalate (PET). FTIR analysis resolved the second and third particles to be of biological nature, likely undigested material. Additionally, one cotton fibre (74.08% similar to cellulose) was found in the filter used as a blank during the specimen dissection. No particles were present in the procedural blanks. Etymology. The species names, plasticus, stems from Latin for plastic. This name speaks to the ubiquity of plastic pollution present in our oceans.
PLEON AND UROSOME (Figs. 7, 8): Epimeron 1 anteroventral corner rounded with long slender setae; posteroventral corner produced into a small tooth. Epimeron 2 anteroventral margin lined with short fine setae; posteroventral corner produced into a strong tooth. Epimeron 3 ventral margin lined with long fine setae, weakly curved (Fig. 7D). Urosomite 1 with anterodorsal notch (Fig. 7D). Uropod 1 peduncle with one apicomedial setae; inner ramus subequal in length to outer ramus; outer ramus 0.85x as long as peduncle; outer ramus with 18 lateral and eight medial spine-like setae; inner ramus with 20 lateral and 11 medial spine-like setae (Fig. 8A). Uropod 2 peduncle with one apicomedial setae; inner ramus subequal in length (0.9x) to outer ramus; outer ramus subequal in length to peduncle outer ramus with 20 lateral and three medial spine-like setae; inner ramus with seven lateral and 16 medial spine-like setae (Fig. 8B). Uropod 3 inner ramus subequal in length to article 1 of outer ramus; article 2 of outer rami short, 0.05x length of article 1; setae of distolateral angle of peduncle of normal length and stoutness; medial margins of both rami with plumose setae (Fig. 8C). Telson 70% cleft, pair of apical setae on each lobe parallel with beginning of cleft, distal margin with a single apical seta on right lobe, distal end of left lob missing (Fig. 8D).
Variations. As with other species of Eurythenes, there appears to be very little sexual dimorphism. In part, this could be limited to having a single male specimen. The mature male paratype (USNM 1615732) has calceoli present on both antenna 1 and antenna 2. Both antennae are shorter than the holotype with antenna 1 accessory flagellum being 10-articulate, antenna 1 25-articulate, and antenna 2 54-articulate. Additionally, the maxilliped inner plate of the male paratype has three apical protruding nodular setae, specifically lacking the third setae present on the holotype (Fig. 5F). There were differences present in the juvenile paratype (USNM 1615730) that included typical cohort differences among Eurythenes, such as fewer setae on pereopods and uropods and reduced articulation on antennae (antenna 1 accessory flagellum 7-articulate, antenna 1 15-articulate, and antenna 2 38-articulate). In addition, the juvenile paratype had more pronounced and raised dorsal carination than on the adults (Fig. 7E). This difference was present among all the juvenile specimens observed.
Differential Diagnosis. As highlighted in d'Udekem d' Acoz & Havermans (2015), the morphological characteristics that separate and define the species within the gryllus-complex are hard to observe and should be used with caution. Eurythenes plasticus sp. nov. is a member of the gryllus-complex morphologically and genetically. Nevertheless, there is a combination of characters that are unique to E. plasticus sp. nov. and allow it to be distinguished from the morphologically similar species E. andhakarae, E. magellanicus, and E. aequilatus. The most distinctive characteristics are the robust, spine-like setae on rami of uropod 1 and 2 (Fig. 8A, B) and the lobes of pereopod 5 coxa (Fig. 7A), here being unequal, which is novel within Eurythenes. Eurythenes plasticus sp. nov. can be differentiated from E. andhakarae with article 2 of the mandible palp being narrow (instead of expanded), four protruding nodular spines on the inner plate of the maxilliped (versus three non-protruding), and straight ventral border of coxa 4 (opposed to curved). Eurythenes plasticus sp. nov. can be separated from E. magellanicus with a long gnathopod 1 palm (instead of short), a straight ventral border of coxa 4 (opposed to curved), a subquadrate posteroventral corner in epimeron 3 (instead of bearing a small tooth), and the rami of uropod 1 and 2 being subequal (opposed to uropod 2 outer ramus being shorter than inner ramus and uropod 1 outer ramus being longer than inner ramus). Eurythenes plasticus sp. nov. can also be distinguished from E. aequilatus by its eyes with a variable width (opposed to constant width), the outer plate of maxilla 1 with 8/3 crown arrangement (instead of 9/3 arrangement), and a long gnathopod 1 palm (instead of short).

Discussion
The salient finding of this study is the paired molecular and morphological identification approaches provided congruent evidence that E. plasticus sp. nov. represents an undescribed species within Eurythenes. Further, as a scavenger at upper hadal depths (6010 -6949 m) in the Mariana Trench, E. plasticus sp. nov. is not exempt from ingesting microplastics that are bioavailable within the hadal zone.
In comparison to described Eurythenes species, E. plasticus sp. nov. was placed as part of the gryllus-complex and most closely related to the abyssal E. magellanicus (Fig. 2). The bPTP analysis of COI and both K2P analyses delineated E. plasticus sp. nov. to be a distinctive lineage, and these methods aligned with previous studies that detected cryptic speciation within the gryllus-complex (Havermans et al. 2013;Eustace et al. 2016;Narahara-Nakano et al. 2017). The 16S phylogeny specifically showed E. plasticus sp. nov. to be nearly identical to Eg7 ( Fig. 2A; France & Kocher 1996;Havermans et al. 2013). This Eurythenes sp. was a singleton recovered from abyssal depths at the Horizon Guyot seamount, Pacific Ocean, and it was collected along with another Eurythenes sp. from the divergent Eg9 clade (Havermans et al. 2013). Confidence in the identification of Eg7 would be further strengthened with additional genetic or morphological data.
The morphological variation seen in E. plasticus sp. nov., such as an uneven coxa 5 lobe and lack of a tooth on the posteroventral corner of epimeron 3, supported the phylogenetic evidence as an undescribed lineage. Consistent with previous studies, these morphological characteristics should be used with caution, as some are difficult to discern objectively. Additional specimens, like from the Eg7 clade, may reveal phenotypic plasticity in the characteristics observed in this morphological study (d'Udekem d'Acoz & Havermans 2015). Continued application of a combined molecular and morphological approaches in future studies is likely to reveal further species diversity within the gryllus-complex.
The discovery of E. plasticus sp. nov. continues to align with the pattern Eurythenes that the geographic and bathymetric species distributions are complex (Havermans 2016). With the Eg7 singleton, the geographic range of E. plasticus sp. nov. thus far appears to be restricted to the Central Pacific Ocean. Across that ocean basin, E. plasticus sp. nov. has broad bathymetric range, ~3000 m. While it is common among Eurythenes to be found only in a single ocean basin and have a wide vertical distribution (Eustace et al. 2016;Havermans 2016), it is less common to span across the abyssal and hadal zones. Although, this is not unique, as it has been documented in other amphipods, such as A. gigantea . A species needs to be able to cope at the cellular, reproductive, and physiological levels in both the stable abyssal (Smith et al. 2008) and the dynamic hadal environments Downing et al. 2018). Yet, it was curious that during the present study, E. plasticus sp. nov. was only collected from upper hadal depths, despite amphipods being captured at shallower and deeper depths (43 additional deployments 4506 to 10545 m; data unpublished). This highlights that the distribution of E. plasticus sp. nov. is a patchwork. Further work and sampling will be required to understand the conditions that support the presence of this species.
The finding of a microplastic fibre in the hindgut of a juvenile was not unexpected. Deep-sea scavenging amphipods, as an adaption to their food limited environment, indiscriminately consume carrion (Blankenship & Levin 2007) and are known to inadvertently ingest microfibres present in the carrion and sediment . The detection of a microplastic adds to the number of hadal scavenging amphipods, including adult specimens of H. gigas from the Mariana Trench and Eurythenes sp. 'hadal' the Peru-Chile Trench , which have been found to have consumed plastic microfibers. Microplastic consumption by a juvenile indicates that scavenging amphipods are potentially ingesting microplastics throughout their life, which could pose acute and chronic health effects. While the ecotoxicological impacts of microplastic exposure has yet to be investigated on deep-sea amphipods, early work on other Malacostraca indicates that the ingestion of polypropylene fibres by the sand crab, Emerita analoga, increases adult mortality and decreases in retention of egg clutches (Horn et al. 2019).
This study adds to the growing body of literature on marine organisms ingesting plastic and microfibers (Besseling et al. 2015;Lusher et al. 2015;Bellas et al. 2016;Alomar & Deudero 2017). The microplastic found in the hindgut of E. plasticus sp. nov. was most similar to PET, which is one of the top five most prevalent synthetic plastic polymers produced and discarded globally (Geyer et al. 2017). Without substantial global changes to the life cycle of plastic, from reducing the rate of plastic production to improving waste management (Forrest et al. 2019), plastics and microfibres will continue to be transported to the deep sea and be ubiquitous in the hadal food chain for the foreseeable future.