Deep ocean seascape and Pseudotanaidae (Crustacea: Tanaidacea) diversity at the Clarion-Clipperton Fracture Zone

Understanding the diversity and spatial distribution of benthic species is fundamental to properly assess the impact of deep sea mining. Tanaidacea provide an exceptional opportunity for assessing spatial patterns in the deep-sea, given their low mobility and limited dispersal potential. The diversity and distribution of pseudotanaid species is characterized here for the Clarion and Clipperton Fractures Zone (CCZ), which is the most extensive deposit field of metallic nodules. Samples were taken from the Belgian, German and French license areas, but also from the APEI 3 (Area of Particular Environmental Interest 3) of the Interoceanmetal consortium associates. The combination of morphological and genetic data uncovered one new pseudotanaid genus (Beksitanais n. gen.) and 14 new species of Pseudotanais (2 of them virtual taxa). Moreover, our results suggest that spatial structuring of pseudotanaid diversity is correlated with deep-sea features, particularly the presence of fractures and seamount chains crossing the CCZ. The presence of geographical barriers delimiting species distributions has important implications for the establishment of protected areas, and the APEI3 protected area contains only one third of the total pseudotanaid species in CCZ. The specimen collection studied here is extremely valuable and represents an important first step in characterizing the diversity and distribution of pseudotanaids within the Tropical Eastern Pacific.

The influence of habitat heterogeneity on species diversity has puzzled biologists for a long time and still raises many questions [1][2][3] . High habitat heterogeneity and spatial complexity provide shelter for many invertebrate taxa and might result in higher diversity of benthic organisms 4 . Competition and influence of predators are restricted in heterogeneous areas 5,6 while the number of potential ecological niches increases 7 . Studies concerning benthic marine fauna have traditionally focused on shallow-water areas, so that knowledge on deep-sea habitat heterogeneity and its influence at various spatial scales is still lacking 8 . The deep-sea ecosystem was considered as a rather homogeneous environment in the past, but the application of state-of-the-art technologies for habitat mapping has proven otherwise 1 . McClain and Barry (2010) 9 have shown that habitat heterogeneity is an important factor driving the structure of benthic assemblages and that significant species turnover can be observed at relatively small scales (<1 km) 8 . Abyssal hills increase habitat heterogeneity, benthic megafaunal biomass and diversity 10 . Furthermore, benthic meiofauna studies also show that deep sea nodule fields facilitate the coexistence of species with different modes of life, ranging from sediment dwelling to epifaunal 11 .
The Clarion and Clipperton Fractures Zone (CCZ) is a 6 million km 2 region located in international waters of the Tropical Eastern Pacific. Well-known to mining corporations, this is the most extensive deposit field of metallic nodules, rich in manganese, nickel, copper and cobalt 12,13 . The attraction for deep sea nodules has raised in the last few years because they host large quantities of other critical metals needed for high-tech, green-tech, and energy applications 14 . The exploration and exploitation of the CCZ is currently managed by the International Seabed Authority (ISA), an intergovernmental body that regulates mining and related activities in the seabed beyond national jurisdiction 15 . ISA has recently granted 15 mining licences in the CCZ area and selected 9 Areas Results phylogenetic analyses. Pseudotanaids were found in 87% (13 out of 15) of the stations surveyed, which confirms the generalized presence of these tanaids in the deep-sea benthos ( Table 1). The bathymetric range where pseudotanaids were captured was large, spanning from 4093 m to 4877 m depth. A total of 67 individuals were used for molecular analysis and gave positive DNA barcoding results (Table 2). A total of 16 different COI haplotypes were obtained (Fig. 1), representing one Beksitanais and 14 Pseudotanais species (two virtual taxa, without a voucher left for morphological analysis). The sequence alignment spanned 691 bp before trimming and was reduced to 611 bp after running Gblocks. The Hasegawa-Kishino-Yano (HKY + G + I) model showed the lowest BIC score (BIC = 9947.97) and it is considered to describe the substitution pattern the best. Non-uniformity of   Pseudotanais, the 'spicatus' group and either the 'affinis + longisetosus' (0.429 ± 0.051) or the 'abathagastor + denticulatus' (0.402 ± 0.055) clades show divergences almost twice as large as those observed between 'abathagastor + denticulatus' and 'affinis + longisetosus' (0.275 ± 0.037).
Spatial modelling and genetic gradients. The 3D-model based on mean sea level data reveal an extremely heterogeneous deep sea landscape at the CCZ, with the presence of several seamounts and knolls (Fig. 2). In fact, two underwater mountain chains cross the studied area: one rise running east-to-west around latitude 17°N and another running south-southwest around longitude 120°W. The first isolates the APEI3 area (located around 18°N) from the remaining sampling sites, and includes seamountains about 4000 m high, reaching to 250 m under the surface (see Discussion). The second runs over the IOM area and separates the BGR area (located around 117°W) from the rest. Plotting the distribution of the newly identified taxa on the 3D spatial model revealed several species (P. oloughlini, P. yenneferae, P.georgesandae and the sister species P. gaiae and P. uranos) to be restricted to the APEI3 area. Another group of species were only found in the BGR and/or IOM areas (P. romeo, P. mariae, B. apocalyptica, virtual Pseudotanais sp. B and P. chopini). The virtual Pseudotanais sp. A, P. julietae, P. geralti and P. kobro were found together in the GSR area, although P. kobro was also collected in the BGR and IOM areas, and P. geralti was also found in the IOM area. The Spearman rank coefficient revealed a significant correlation between geographical and genetic distances for the complete dataset (ρ = 0.046; p-value = 0.032), and this spatial correlation was even higher when each well-supported phylogenetic clade 'affinis + longisetosus' (ρ = 0.121; p-value = 0.009) or 'abathagastor + denticulatus' (ρ = 0.224; p-value ≤ 0.001) was analysed independently. The linear fitting of an isolation by distance model gave similar results, with the  Symbols indicate comparison between all taxa (O), between samples from the 'affinis + longisetosus' clade (X) or between samples from the 'abathagastor + denticulatus' clade (Δ).
Pseudotanais species described in the present study are grouped into previously erected morpho-groups by   31 and Jakiel et al. (2018) 32 . A list of characters that define each group are included before the species descriptions. An identification key is included at the end of the Results section as well to enable easier identification and clear separation of morpho-groups.   Pereopod-4 ( Fig. 11E) basis 6.2 L:W, 4.1x merus, with penicillate ventral seta; ischium with seta; merus 2.5 L:W, 0.6x carpus, with seta; carpus 3.6 L:W, with two short and one rod setae, and with blade-like spine 0.3x propodus; propodus 5 L:W, 2.3x dactylus and unguis combined length, with one simple and two serrate setae subdistally, and with serrate seta distally 0.8x propodus and microtrichiae on ventral margin; dactylus 2.7x unguis.
Pereopod-5 (Fig. 19F) basis 7.3 L:W, 7.3x merus; ischium naked; merus 1.1 L:W, 0.3x carpus, with seta; carpus 3.5 L:W, 1.2x propodus, with two simple setae, one rod seta 0.9x propodus, and with blade-like spine 0.5x propodus; propodus 6.0 L:W, 2.5x dactylus and unguis combined length, with two simple setae on ventral margin, one seta on dorsal margin, and microtrichia on ventral margin; dactylus 2.0x unguis. Distribution: P. romeo n. sp. is known from the Belgium licence area (GSR) of the Central Pacific. Remarks: Pseudotanais romeo n. sp. is morphologically and genetically most similar to P. julietae (Fig. 1) and it is distinguished from all other members of the 'affinis + longisetosus' group by the same character set as P. julietae (see remarks under P. julietae). P. romeo is distinguished from P. julietae by the number of setae on basis of pereopod 1-3: 5, 6, 3 and 6, 5, 5, respectively. P. romeo has naked maxillped endites whereas P. julietae has maxilliped endites ornamented with two tubercles (gustatory cusps) and one seta. The presence of two spines on cutting edge of the cheliped in P. romeo also allow to separate it from P. julietae with smooth cutting edge.     Remarks: P. geralti can be distinguished from the other species in this group by the same characters as listed in P. yenneferae. P. geralti is morphologically closer to P. yenneferae from which is distinguished by its relatively Remarks: The 'denticulatus + abathagasthor' group can be distinguished from the 'affinis + longisetosus' group by the presence of a long seta on merus pereopod-1 in the 'affinis + longisetosus' clade.  Remarks: Pseudotanais georgesandae n. sp. can be distinghuished from all the other members of the 'denticulatus + abathagastor' group by the wide mandible molar. The molar of P. georgesandae has two bifurcate long spines, which are absent in P. corollatus and P. denticulatus. The molar of Pseudotanais sp. C has one straight spine.  Mouthparts. Labrum (Fig. 29C) hood-shaped, naked. Left mandible (Fig. 29D) lacinia mobilis well developed and serrate distally, incisor distal margin gently serrate molar broken during dissection. Right mandible (Fig. 29E) incisor distal margin serrate, lacina mobilis merged to a small process. Maxillule (Fig. 29F,F') with eight distal spines and three subdistal setae, endite with two setae. Maxilla (Fig. 29G) semioval. Maxilliped (Fig. 29H,H') endites merged with groove in the mid-length, distal margin with two tubercles (gustatory cusps) and with seta; palp article-2 inner margin with three setae, outer margin with seta; article-3 with three setae, article-4 with six setae. Epignath (Fig. 29I) distally pointed. Cheliped (Fig. 30A) basis 1.6 L:W, with distoproximal seta; merus with seta; carpus 2.3 L:W, with two ventral setae, and with one dorsodistal and one dorsosubproximal setae; chela non-forcipate; palm 2.2 L:W, with row of six setae on inner side; fixed finger distal spine pointed, with three ventral setae; dactylus 6.7 L:W. Pereopod-2 (Fig. 30C) coxa with seta; basis 6.7 L:W, 3.9x merus; ischium with two ventral setae; merus 1.42 L:W, 0.8x carpus, with two setae; carpus 1.8 L:W, 0.9x propodus, with two setae, one spine and one blade-like spine 0.5x propodus; propodus 6.8 L:W, 1.5x dactylus and unguis combined length, with seta and microtrichia on ventral margin; dactylus 0.7x unguis.
Distribution: P. chopini n. sp. is known from the Belgium (GSR) and Interoceanmetal (IOM) licence areas of the Central Pacific.
Distribution: P. chaplini n. sp. is known from the IFREMER and IOM licence areas of the Central Pacific. Remarks: The exopod uropod being longer than endopod allows for distinguishing the new species from P. abathagastor, P. corollatus, P. denticulatus, P. georgesandae, P. chopini and Pseudotanais sp. C, as well as from all other species of the genus Pseudotanais. Diagnosis: Mandible molar acuminate with bifurcate distal tooth. Antennal articles 2-3 with spine. Pereopods 2 and carpus with long blade-like spine. Uropod exopod longer than endopod.
Etymology: The species is named in recognition of the great holothurian specialist and wonderful friend and colleague -Dr. Mark O'Loughlin.

Discussion
The present study uncovered a significant diversity of pseudotanaids within the CCZ. A total of 15 new species are described here combining morphological and molecular data. Pseudotanaidae had been reported only once before from CCZ and without including any description 36 . This is also the first time pseudotanaids are studied using a DNA barcoding approach, with the only entry available in GenBank for this family being the histone 3 sequence from a Pseudotanais sp. collected in Crawl Key, Panama 27 . Another study on Pseudotanaidae from the North Atlantic reported a complex of cryptic species in four ecologically-diverse basins around Iceland 31 , although the lack of genetic data prevented clear taxa delimitation. The wide geographic sampling carried out, combined with a reverse taxonomy approach, suggests that pseudotanaids might have comparatively narrow ranges (considering the entire study area), because most species were mainly limited to the closest stations.
Potentially narrow ranges could also be inferred from the extensive tanaid collection made in Amundsen and Scotia Seas 29 . Deep-sea species are generally rare and sparsely distributed, so it is not surprising that each species in our study was represented by just a few individuals. The mechanisms maintaining the immense diversity but low abundances in the deep sea are hardly understood 29 and the low number of properly preserved individuals obtained, despite immense logistic efforts, hampers morphological and molecular studies of the abyssal fauna 37,38 .
Resolving the presence of cryptic species is currently considered one of the main challenges for taxonomy [39][40][41] . Phenotypic plasticity and high sexual dimorphism may lead to misidentification of tanaidaceans 42,43 and lack of detailed morphological studies might obscure the real number of species and true diversity 44,45 . For example, dimorphic male and females of Beksitanais apocalyptica could be described for the first time here thanks to a DNA barcoding approach. Beksitanais apocalyptica is the only member of the genus described from the Pacific and the first for which molecular information is made available. The new genus is distinguished from the other Pseudotanaidae genera based on the following set of unique characters or character combination: Antennula article-3 with thickened rod seta; chela forcipate with serrate incisive margin, but propodus (palm) without small folds in distodorsal corner and pereopods 4-6 dactylus and unguis fused with a small hook on tip. Similarly, the separation of the known Pseudotanais species into the four groups proposed by Bird & Holdich 32 and Jakiel et al. namely, 'affinis' , 'denticulatus' , 'forcipatus' and 'longisetosus' was re-assessed here. Careful examination of the material from CCZ uncovered a close relationship between 'affinis' and 'longisetosus' and the presence of at least two more Pseudotanais species groups namely, 'abathagastor' and 'spicatus' . The recognition of these clades is supported by the setation pattern on pereopods 1, 5 and 6 and by the setal types on pereopods 2 and 3. The new 'spicatus' group can be characterized by very short blade-like spine in pereopod-2 and minute unguis in pereopods 5 and 6, whereas the 'abathagastor' group is distinguished by a combination of short setae on merus and carpus of pereopod 1, and by the presence of setae (not spines) on the antennal articles 2 and 3. The congruence observed for both morphological and molecular data suggests that Pseudotanais might in fact be formed by several complexes of cryptic species.
Discovering new taxa in a sample taken from any arbitrary chosen spot in the deep sea occurs quite frequently 46 . The deep-sea has traditionally been associated with a homogeneous environment, but state-of-the-art technologies proved that abyssal landscapes include different structures, such as seamounts, rises or fracture zones. This spatial heterogeneity is likely to impact the diversity and distribution of abyssal fauna, particularly for small epibenthic species 47 . The numerous asymmetric ridges, scarps, and elongate depressions at the Clarion facture zone can effectively limit dispersion and constitute geographical barriers, because none of the species collected from the APEI3 zone was found anywhere else. The Clarion Fracture Zone has been produced by seafloor spreading as the scar of transform faulting that began at least 80 million years ago and that is still continuing at present 48 . The patterns of magnetic intensity of the seafloor rocks in the studied area are displaced laterally, and rocks of the northern block are millions of years older than adjacent rocks south of the fracture zone 49 . Similarly, the elevated topography of the south-to-north ridge could be considered a remnant of an old east Pacific rise www.nature.com/scientificreports www.nature.com/scientificreports/ (EPR), a sea-floor spreading center that was active approximately 30 mya. Our results suggest that physical barriers restrict the distribution of Pseudotanaidae species, promoting genetic differentiation and allopatric speciation. The sessile lifestyle of pseudotanaid females, which are generally found in self-constructed tubes, makes them particularly sensitive to geographic barriers 44 .
Other environmental factors could explain the observed distribution of pseudotanaid taxa, and might be correlated with the CCZ deep sea landscape. There is mineralogical and chemical evidence for heterogenous sediment composition due to hydrothermal influence around the Clarion fracture zone between 113°W and 119°W. Similarly, nodules from pelagic clays found north of the Clarion fracture zone show higher Mn/Fe ratios 50 . Food availability might also affect the spatial distribution of diversity in the deep-sea 50 , because only a small part of the particulate organic carbon (POC) from the euphotic zone will ever reach the ocean bottom 16 . Megafauna studies suggest higher abundance and diversity in the eastern part of CCZ, where POC availability is larger 37 . For example, Polychaeta family richness was found to be higher in the eastern IOM area than in the more western IFREMER region 43 . Nevertheless, the northernmost area studied here (APEI3) showed similar Pseudotanaidae abundances and species richness as the southeastern areas despite a gradual increase in POC flux. Finally, other factors such as the calcite compensation depth (CCD), which in the Pacific Ocean is about 4200-4500 metres, could also have an impact on the carapace-bearing crustaceans 16 . Further sampling within the CCZ would be essential to properly evaluate the relative importance of these factors on the observed distribution of deep-sea pseudotanaids.
The Clarion-Clipperton Zone remains the focus of international mining companies and faces a real danger of industrial exploitation, so recognizing its biological diversity and how it is structured are primary and critical steps preceding any potential anthropogenic activity 51,52 . A marginal understanding of deep-sea ecosystems utterly prevents an adequate assessment of the potential impact of mining operations on the marine environment 53 . Deep-sea expeditions are generally deprived of an opportunity for repeated sampling, being highly costly and burdened with logistic difficulties, so the large collection of pseudotanaids studied here is extremely valuable. The correlation observed between spatial features and species distribution has important implications for the establishment of protected areas, and the APEI3 area studied here would only protect one third of the total pseudotanaid species found in CCZ. It is possible that some species might have wider ranges than suggested by our current sampling, but this study represents an important first step in characterizing the diversity and distribution of pseudotanaids from the Tropical Eastern Pacific.

Sampling. The European Joint Project Initiative -Oceans (JPI-O) 'Ecological Aspects of the Deep-Sea
Mining' is a long-term intergovernmental initiative to assess the potential impact of deep sea mining using ecological and genetic techniques 54,55 . The marine expedition 'EcoResponse 2015' was organized to assess the genetic connectivity between populations from different CCZ areas. The biological material included in the present study was collected during SO-239 cruise, conducted on RV Sonne, from 10 th March until 30 th April 2015. Tanaidacean samples were taken from the Belgian, German and French license areas, but also from the APEI3 and Interoceanmetal (i.e. the consortium associating Bulgaria, Cuba, Czech Republic, Poland, Russian Federation and Slovakia). Thus, the areas surveyed include APEI3 (Areas of Particular Environmental Interest 3); BGR (Bundesanstalt fur Geowissenschalfen und Rofstoffe, Germany); IOM (Interoceanometal Joint Organisation); GSR (Global Sea Mineral Resources NV, Belgium) and IFREMER (France) ( Table 1). An epibenthic sled (EBS) was used to collect material at each sampling site as in Brandt and Barthel 56 . Samples were sieved on board through a 300 µ mesh using cooled seawater and rapidly transferred to cold 96% EtOH. Fixed samples were stored at −20 °C until further processed. Detailed onboard and laboratory sample-processing procedures can be found in Rhiel 57 . phylogenetic analyses. A single cheliped was taken using sterile needles as starting material for DNA extraction using the Chelex (InstaGene Matrix, Bio-Rad) method as in Palero et al. 58 . The COI gene was amplified using a 25 μL volume reaction containing 22 μL H 2 O, 0.5 μL of each primer (10 pmol/μL) polyLCO and pol-yHCO 59,60 1U of Illustra PuReTaq Ready−To−Go PCR Beads (GE Healthcare) and 2 μL of DNA template. The PCR protocol was 94 °C for 3 min, 40 cycles of 94 °C for 40 s, 42 °C for 30 s, 72 °C for 1 min, and a final elongation step of 72 °C for 10 min. A 2 μL aliquot of the PCR products was visualized in Midori Green-stained (Nippon Genetics) 1.5% agarose gels to verify PCR product quality and length. PCR purification and sequencing using forward and reverse primers was carried out by MACROGEN (Amsterdam, Netherlands). Consensus sequences were built using Geneious version 9.1.3 (www.geneious.com) and compared with the GenBank database using BLAST 61 to discard contamination from non-arthropod sources. Sequences were aligned using alignment option (L-INSi) of MAFFT 62 as implemented in Geneious. To improve reliability, we extracted conserved (ungapped) blocks of sequence from the alignment by using Gblocks server with default settings 63,64 . Selection of the best nucleotide substitution model was performed according to the BIC criterion as implemented in MEGA v7 58,65 . The aligned sequences and selected evolutionary model were used to estimate genetic distances and the corresponding Maximum Likelihood phylogenetic tree in MEGA. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. Nodal support was assessed using 500 bootstrap replicates. Spatial modelling and genetic gradients. A 3D-model of the deep sea landscape of the CCZ was built using the GeoElevationData function as implemented in the Mathematica v11.0 software package (Wolfram Inc., USA). GeoElevationData returns the elevation with respect to the geoid (=mean sea level) of a specified location. An array including the bathymetry for 12,231 different latitude longitude coordinates was built by uniformly www.nature.com/scientificreports www.nature.com/scientificreports/ recording the mean sea level every 1/10 th of a decimal degree in the rectangular area spanning from 11°N 116°W to 19°N 131°W. A contour-plot representing the array of mean sea level values and the location of the sampling sites was generated using the ListPlot and ListContourPlot functions in Mathematica. Names for particular structures, including fractures, seamounts and knolls, are taken from the General Bathymetric Chart of the Oceans (GEBCO) undersea feature Gazetteer (https://www.ngdc.noaa.gov/gazetteer/). The degree of association between geographic and genetic distances was measured using the Spearman rank correlation. This non-parametric correlation test was selected because it does not carry any assumptions about the distribution of the data. A standard isolation by distance (IBD) analysis was also carried out in Mathematica to further analyze the presence of a linear correlation between geographic and genetic distances.
Morphological analyses and species descriptions. Specimens were dissected with chemically-sharpened tungsten needles, and the dissected appendages slide-mounted using glycerine. Drawings were prepared using a light microscope (Nikon Eclipse 50i) equipped with a camera lucida. Digital drawings were obtained using a graphic tablet following Coleman 66 . Total body length (BL) was measured along the main axis of symmetry, from the frontal margin to the end of the telson. Body width (BW) was measured at the widest point along the main axis of symmetry. To simplify species descriptions, the expression 'Nx' replaces 'N times as long as' and 'N L:W' replaces 'N times as long as wide' . The measurements were made with a camera connected to the microscope (Nikon Eclipse Ci-L) and NIS-Elements View software (www.nikoninstruments.com). The body width and the length of the carapace, pereonites, pleonites, and pleotelson were measured on whole specimens. The poor condition of individuals after DNA extraction or incompleteness even for well-preserved specimens, made the description of pereonite and pleonite setation not reliable. Therefore, this character was not included in the species description. The morphological terminology here follows Błażewicz-Paszkowycz et al. (2012) 67 . The unique blade-like spine of Pseudotanais, Mystriocentrus and Parapseudotanais species 67 , is recognized as 'long' when is at least 0.6x propodus, 'semilong' when it is 0.5x propodus and 'short' when it is at most 0.3x the propodus. The type of sensory seta present on carpus of pereopod 4-6 is defined as rod seta (slightly inflated distally and with a pore) following 68 and 69 . This seta is recognized as 'long' when is at least 0.8x propodus, 'semilong' when it is 0.5x propodus and 'short when it is at most 0.25x propodus. Beside simple setae (=without ornamentation), at least four setae types are recognized here: (1) serrate -with serration or denticulation, (2) plumose -with any type of plumose or delicate setulae tufts distributed along the main axis, (3) penicillate -with a tuft of setules located distally and with a small knob on which a seta is fixed to the tegument and, (4) sensory -specified above.
Among the studied individuals: manca, neuter, and male stages were recognized. Specifically, the term 'manca' describes juveniles with or without buds of pereopod-6, respectively; 'mature (swimming) male' 30 refers to individuals with completely developed sexual dimorphic characters. 'Neuter' is retained for the stage developed from manca that cannot be classified as either female or juvenile male. The examined material will be deposited in "Senckenberg Research Institute and Natural History Museum" (Hamburg, Germany). Taxonomic descriptions and the corresponding identification key were prepared using the DELTA software (DEscription Language for TAxonomy) 44,66,70 . Published: xx xx xxxx