Diversity under a magnifier lens: the case of Typhlotanaidae (Crustacea: Tanaidacea) in the N Atlantic

Research focusing on ‘stout-bodied’ typhlotanaids collected from several sites around Iceland and adjacent N Atlantic region has resulted in the description of 15 species new to science, as well as the creation of eight new genera. Typhlotanais eximius Hansen, 1913 is redescribed and transferred to a new genus, while Typhlotanais crassus and Peraeospinosus adipatus are transferred to the genus Larsenotanais. The morphological and the molecular data were combined to consolidate and confirm the validity of the results obtained from both approaches. The polyphyletic nature of the Typhlotanaidae and its serious of its taxonomic diversity are emphasized, although molecular analysis reveals that the ‘stout-bodied’ Typhlotanaidae form monophyletic clade. Depth and temperature are identified as the main environmental parameters determining the distribution of this group of Typhlotanaidae. Several species are clearly associated with the shelf and upper bathyal of Iceland. The Greenland-Iceland-Faroe Ridge is shown to be a distinct zoogeographical barrier for typhlotanaids inhabiting the deeper slope and abyssal regions around Iceland.

The oceanic floor below the continental shelf is the largest and scarcely known ecosystem and is inhabited by a high number of rare species many of which remain unknown to science 1,2 . The consequences of climate change, clearly visible and severe on land, also affect the fragile and unrecognized organisms living in the deepest part of the ocean [3][4][5] . Evolving under specific environmental conditions the fauna is potentially vulnerable to dynamic environmental changes that can disrupt their biological and physiological processes and reproductive cycles, leading to changes in population structure, shifts in ecosystem functioning or even extinction [6][7][8][9] . Besides, the deep sea is designated as a territory for large-scale economic operations that could inducing extreme ecosystem transformations that are difficult to assess for scale and direction 10 . For this reason, understanding the biodiversity, in the sites directly affected by human activity or consequences caused by climate change is a priority for current research.
The N Atlantic is an important sink for global ocean waters and the origin of the thermohaline circulation [11][12][13] . The dynamic warming observed in this part of the world (Atlantification) is of the highest concern due to its multifaceted threat to the climate-sensitive N Atlantic fauna 11,[14][15][16][17][18] and has become a natural observatory for climate change [19][20][21] . The knowledge about which factors structure or modify benthic communities in the N Atlantic is essential for assessing whether and how the changing ecosystem affects on its sensitive fauna. The region has many complexities, and the Greenland-Icelandic-Faroe Ridge aligned perpendicular to Mid-Atlantic Ridge, is a main topographical feature separating cold Arctic waters from warmer Atlantic waters 22,23 and acts as a natural biological barrier 12,13,23 . It also limits biological migrations from south to north in the N Atlantic and is an ideal natural experimental zone for observing the zoogeographic shifts of fauna caused by environmental factors. For large or commercially important species (e.g. fish), the diversity and zoogeographical ranges are considered well understood when compared to smaller animals of lesser, or as yet unknown, commercial potential. Nevertheless, these smallest species are proving to be a highly diverse component of deep-sea ecosystems whose correct identification allows for reliable biological analyses, and are important object of taxonomic, phylogenetic and zoogeographic research [24][25][26] .
Among these small benthic organisms is the Typhlotanaidae Sieg, 1984-a diverse and poorly known family of the peracarid Tanaidacea. The family is represented by small and specialized taxa with a three-article antennule and a 'clinging-type' of pereopods 4-6. These appendages have a robust basis, shorter dactylus-unguis (claw), and Phylogenetic approach: Morphological analysis. As a result of the phylogenetic analysis, three parsimony equal trees with a length of 326 steps were obtained. A strict consensus tree was calculated from the trees obtained ( Fig. 2A). The consistency index (CI) and retention index (RI) had values of 0.53 and 0.39, respectively.
In the obtained tree, the Typhlotanaidae species formed into two large clades. The first clade is composed of several 'stout-bodied' and 'slender-bodied' genera and spits into several smaller clades. In the first clade "Gudmundotanais" is most distinct taxon and makes a separate lineage supported by the following synapomorphies: antennule article-1 2.6 L:W, antenna article-2 without seta, pereopod-1 basis 0.6× other articles, pereopod-1 merus 2.1 L:W, pereopods 2-3 propodus with long dorsodistal seta (longer than dactylus) and uropod exopod 0.9× endopod. The more internal clade includes three species of the genus "Hansenotanais" (Bremer support 0.49) that share a long ventrodistal seta on the carpus (longer than half of the propodus) in pereopods 2-3. In three equal parsimony trees, these species swap places within this subclade. The most internal clade forms a cascade of branches composed of 'slender-bodied' typhlotanaids (Typhlamia sp., Pulcherella sp., Ty. proctagon, and Ty. greenwichensis) as well as two 'stout-bodied' species of the genus "Caesatanais" (Bremer 0.2). Their common feature is a long ventrodistal seta on the pereopod-1propodus.
• Maxilliped palp four-article; article-2 with three inner and one outer setae; article-3 with four inner setae; article-4 with five terminal and one subdistal setae. • Cheliped basis separated from pereonite by short gap.
• Cheliped fixed finger with two ventral setae and three near cutting margin.
These are excluded from the following descriptions apart from where an exception is observed.

Species included Larsenotanais amabilis
Remarks   27 established Larsenotanais for one species. As a result of the present study, the genus now comprises five species including the incompletely described Typhlotanais crassus (Dojiri & Sieg, 1997).
Larsenotanais is morphologically similar to the Brevitanais n. gen. group-1 and Caesatanais (see below) by having uniarticulate uropod rami, but the simple unguis on pereopods 4-6 distinguishes it from the other two groups ( Table 2) of similar uropodal form. Another useful character separating Larsenotanais from Brevitanais group-1 and Caesatanais is the short distodorsal propodal seta in pereopods 4-5 (long in Brevitanais and Caesatanais). It should be noted that at least several typhlotanaids 'stout-bodied' tanaids have short distodorsal propodal seta in 2.8 Absent Absent n/a n/a n/a n/a Rounded n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a Sarsotanais sp. A n/a n/a n/a n/a n/a n/a n/a n/a Remarks Sieg (1986) illustrated and described a species that is considered to be conspecific with Typhlotanais adipatus described by Tzareva (1982) 57 from the shelf of the Cosmonaut Sea in E Antarctica based on collections made on the shelf of the western Antarctic Peninsula. Also, he considered his species to be a representative of the genus Peraeospinosus. As a consequence, the species described by Tzareva (Ty. adipatus) has become member of Peraeospinosus. In light of current knowledge and with a better understanding of the morphology of typhlotanaids, it is clear that Sieg (1986) 56 and Tzareva (1982) 57 studied two distinct species belonging to two different genera. This difference was perceived by   27 who classified Tzareva's Ty. adipatus to the 'cornutus' group, but considered Sieg's Ty. adipatus a synonym of Typhlotanais greenwichensis Shiino, 1970. Unfortunately, Tzareva incompletely illustrated her specimen, but in her figure of the pereopod-4 (Tzareva 1982: Fig. 9 57 ) has a propodus distodorsal seta clearly longer than in the same appendage illustrated by Sieg (Sieg 1986: Fig. 55 56 ). Besides, the two taxa have a quite different body habitus, where short pereonites and a short cephalothorax are clear in Tzareva's species and more elongated pereonites and a slender cephalothorax are present in Sieg's species. The set of these features is sufficient to assess that the species described by Tzareva is closely related to Typhlotanais cornutus G.O. Sars, 1879, in contrast to the species described by Sieg. Since it has all the characteristics that define genus Larsenotanais, we have made the decision to transfer it to this genus, and name it as Larsenotanais siegi n. sp. Larsenotanais amabilis and L. siegi are two Antarctic Larsenotanais congeners which could be distinguished by the length of pereonite-1 (clearly shorter than pereonite-2 in L. siegi, and almost as long as pereonite-2 in L. amabilis), and the setation of antenna article-2 (one seta in L. amabilis, and two in L. siegi). Finally, the distodorsal seta on the propodus of pereopods 4-5 is long in L. siegi (reaching the end of dactylus) and short in L. amabilis (0.5× dactylus).

Larsenotanais tillardi Bamber, 2014
Larsenotanais tillardi-Bamber (2014) 58  Remarks As L. tillardi, three other Larsenotanais species have an uropod exopod reaching 0.8× of endopod length. Nevertheless L. tillardi differs from L. amabilis and L. siegi by the lack of mesial setation on the uropod endopod, where L. amabilis has a PSS and L. siegi has four simple and two PSS. Also, L. tillardi has a naked pereopod-1 basis, where L. amabilis has a middorsal seta and L. siegi has two middorsal setae and proximoventral seta.       , and noted that these species have biarticulate uropod rami. A further detailed analysis of nine species (four listed species and five new species studied during the current study), which morphologically fit well into the group, support establishing a new genus-Brevitanais.
The new genus is defined by several morphological characters that allows separation from the other 'stoutbodied' typhlotanaids ( Table 2). A long dorsodistal seta on the pereopods 4-6 propodus is characteristic of Brevitanais and Caesatanais, although the former as genus has a short antennule article-1 (long in Caesatanais) and rounded pereonite margins (straight in Caesatanais). Because of the variety in uropod articulation, the genus is provisionally divided into three groups based on the number of articles in both uropod rami: group-1 with uniarticulate uropodal rami; group-2 with uniarticulate uropodal exopod and biarticulate uropodal endopod, and group-3 with biarticulate uropodal rami.
Distribution Known from one location on the shelf of the Reykjanes Ridge (Fig. 9), from 209.4-218.4 m depth (this study).

Stuttotanais frenchae Gellert & Błażewicz
Etymology This species is dedicated to Kate French -British Olympic champion in the modern pentathlon, women's individual. Antennule (Fig. 49A) 1.5× cephalothorax; article-1 3.9 L:W, with three setae on inner margin and seta and a PSS on outer margin, and seta and three PSS distally; article-2 2.4 L:W, 0.3× article-1, with dorsal simple seta on inner margin; article-3 5.6 L:W, 1.3× article-2, with five setae.   (Table 3) for typhlotanaid species (Canberra similarity, transformed data, complete method); the mean environmental parameters defining the regions are in Table 2

Discussion
Incorporation of a variety of scientific methods for species delimitation, biodiversity and zoogeographical studies is extremely useful in the study of deep-sea fauna represented by a large number of rare and unknown species 12,13,24,[69][70][71][72] . In our research genetic and morphological approaches were combined to examine a rich and diverse collection of historical and more recent deep-sea tanaidaceans collected in the N Atlantic. Despite these efforts, we were not able to obtain sequences from both markers for all targeted individuals and, as a result, could not construct a concatenated tree. Nevertheless, even with our limited results and relying on a conservative molecular marker such as 18S, it was possible to assess the diversity of 'stout bodied' typhlotanaids. In order www.nature.com/scientificreports/ to define a new genus and species we used integrative approach confronting 18S phylogeny and morphological diversity. Because 18S is very conservative marker, we assumed that each clade represented a taxonomical unit that we understand as a genus. The integrative approach was used for the first time for investigation of Tanaidacea and offered a solution through which deficiencies in our genetic or morphological data can be mutually supplemented and confirmed. In our research we focused on 'stout-bodied' typhlotanaids recovered in 185 samples of over 1465 benthic samples from the N Atlantic over the last five decades (Table S1), supporting the view that the deep-sea fauna of the N Atlantic is relatively well-recognized. By discovering 15 new typhlotanaids species to science and establishing eight new genera and talking into account that our material represents only a small portion form a much larger collection of tanaids, we conclude that knowledge on the diversity of tanaidaceans in the N Atlantic is still far from being completed. In addition, the result of our study elevates the total number of known typhlotanaids species by 13%, and total number typhlotanaids genera by 35% compared to already published taxa (WoRMS accessed 6 April. 2023).
Before our research, three 'stout-bodied' typhlotanaids were recorded from the N Atlantic (Typhlotanais cornutus, Typhlotanais eximius and Typhlotanais inermis) and none of them was recovered by us. Eight taxa targeted by us taxa were delineated with a conservative DNA fragment (18S) and calculation of intra-and infra-species genetic variability allowed us to infer reliable species delimitation (Table S2). Except for a pair of closely related species (Stuttotanais frenchae and Su. carringtonae), we could not obtain COI sequences for all putative species, however calculated genetic distances prompted us to seek for morphological differences segregating both species and further morphological search to delineated morphologically similar species (Table 2).
Morphology. The morphological characteristics that define the clades distinguished using the 18S marker contribute to defining the diagnoses for the genera we have distinguished, but also provide a better understanding of the features (or "characters"), many of which have been previously either underestimated or not even perceived 28,67,73 . Among the important and easy to observe diagnostic features is the articulation and character of the uropods (length relative to the body, and proportions of individual members). Direct observation of deepsea tanaids is virtually impossible, but it is assumed that the uropod exposed at the posterior opening of the tube may acquire the environmental cues in posterior part of the tanaid analogous to the antennae exposed to its anterior part 60,74 , thus could be a subject to evolution by natural selection. Usually, all stout typhlotanaids have rather short uropods in relation to the body, in which the exopod usually reaches 0.8-0.9 of endopod length. The shortest exopod is in the genus Hansenotanais (0.6 × endopod), while the most elongate uropods are observed in Egregiella, where both uropod rami are almost equal.
It should be noted that, with some exceptions, none of the taxa we have distinguished can be diagnosed on the basis of a single apomorphy, but rather by a unique combination of several features ( Table 2). For example, the setation of the pereopods and character of clinging apparatus in pereopods 4-6 was proved to be a diagnostic character 27,28 . The modern imaging techniques 50 applied in our research allowed us to demonstrate that the presence of prickly tubercles is not invariant in the typhlotanaids. Although many of the 'stout-bodied' typhlotanaids have them well developed, they are absent in genus Hansenotanais (Fig. 35D) (also absent in Typhlamia). A bifurcate unguis in the pereopods 4-6 is present in several genera, Brevitanais, Hansenotanais, Caesatanais, Jurundurella, Sarsotanais and Stuttotanais, but only Brevitanais and Caesatanais have a long propodus dorsodistal seta in the pereopods 4-6. Furthermore, calcified microtrichia in pereopods 2 and 3 are unique for Stuttotanais and Sarsotanais (Figs. 44,47,50), although Sarsotanais has them in pereopod-2 merus and carpus, while Stuttotanais only on the pereopod-2 merus. Finally, Jurundurella has distinct spines on the basis of the pereopods 2-5 ( Fig. 40) which are unique feature. Similar spines however are observed also in two other species: Typhlotanais spinibasis (although spines on basis of pereopods 4-6) and Typhlotanais spinipes (spines present in pereopods 2-5). However, Ty. spinibasis is a slender-bodied typhlotanaid (8.5 L:W) and because of long setae in merus, carpus and propodus in pereopods 1-3 was classified to 'ptrispinosus' group , while Ty. spinipes has biarticulate uropod rami (unriarticulate in Jurundurella), long and slender antennule (short in Jurundurella), and ventrodistal seta in propodus 2-3 (spine in Jurundurella) 59,75 . Intriguingly, Ty spinipes, as Jurundurella, has long distal setae in pereopods 2-3 (Fig. 40), thus further study with more material and possible genetic data are needed to test if those taxa are congeneric or at least are phylogenetically closely related taxa.
Phylogeny. The Typhlotanaidae is recognized to be a highly diverse tanaidomorphan family whose persived polyphyletic nature has been frequently underlined 27,28,39,76,77 . Our phylogenetic results, which were based on two markers, should be considered only as a primary step towards complex phylogenetic studies of 'stoutbodied' forms of this family. Nevertheless, the examination of reference barcodes 18S and H3 is given for five new genera (Brevitanais, Caesatanais, Gudmundotanais, Hansenotanais and Stuttotanais) and these data are a baseline for future assignment of new (and difficult-to-identify) 'stout-bodied' typhlotanaids species. The low support of some clades in the 18S tree can be explained by the insufficient number of genera included in the analysis, and it is expected that a more complete analysis with more genera will improve the statistical parameters of the tree. The unrelated clades group together by with low support. Baratheonus groups with the 'stoutbodied' clade in the 18S tree and several 'stout-bodied' genera (Caesatanais and Gudmundotanais) group with the 'slender-bodied' forms in the morphological tree. The gene H3 has an erratic pattern of substitution and is not resolved. Nevertheless, it is quite apparent that the 'stout-bodied' forms do constitute a monophyletic group (Brevitanais, Egregiella, Gudmundotanais, Jurundurella, Sarsotanais and Stuttotanais). The Hansenotanais that grouped with Typhlamia share the cusps in the pereopods 4-6 where the other 'stout-bodied' typhlotanaids have prickly tubercles. Moreover, the morphological tree is congruent with the genetic tree and consists of two large clades with 'stout-bodied' forms and 'slender-bodied' typhlotanaids. The latter clade is supported by the www.nature.com/scientificreports/ pereopod-1 propodus with a long ventrodistal seta. The appearance of two 'stout-bodied' forms (Caesatanais and Gudmundotanais) in this group may be coincidental. The exception is the genus Hansenotanais, which in molecular trees (18S and H3) is always grouped with Typhlamia. Morphological affinities are considered insufficient for testing evolutionary hypotheses because of interference from homoplasy and convergence 78 . The scarcity of genetic data in our research, which could serve for haplotype and private allele inquiry, hamper inferences concerning the radiation of taxa grouped in specific clades. Nevertheless, grouping of all Icelandic Brevitanais in one clade in the morphological tree, is certainly worth noting, especially that Ty. cornutus (Kval Island, N Norway) and Ty. adipatus (W Antarctic) group in a paraphyletic node. Although, the lack of genetic, biogeographical data does not provide a way to go much beyond conjecture, we can hypothesize that high diversity of Icelandic Brevitanais might be result of panmictic populations that genetically diverged after fragmentation from the Quaternary Last Glacial Maxima (LGM). The grounded ice cover that growing on the Arctic shelf forced fauna to migrate to ice-free locations obtainable at greater depths i.e. lower shelf or bathyal or southerly waters 22,[79][80][81][82] . Brevitanais taxa that reveal clear preference for lower shelf depths seems to be naturally predestined to survive glacial maxima on the Icelandic slope. The bathyal glacial shelters are not particularly thoroughly explored in the Arctic, where the research focused on the shallow littoral fauna 80,83 , but the Antarctic fauna could only survive a glaciation period in deeper glacial refuges due to its long and profound topographic isolation 24,84-86 . Diversity. The typhlotanaids are seen to be a diverse family of deep-sea communities with contribution of 11-25% to all tanaidaceans recovered from the NE Pacific, W Australian Slope or E Antarctic 55,87-90 . Regardless of the high diversity, the typhlotanaids were often recorded as unique taxa, represented by few individuals 91,92 . The low frequency and abundance of small deep-sea fauna, as tanaidaceans, are constricted by trophic and physio-chemical conditions of the ecosystem 93,94 , is more often low a consequence of sampling efforts resulted from high cost and logistical challenges of deep-sea surveys 9,95 . In our research, we analysed a well sampled area of the N Atlantic (Table 3). With four of eleven large regions particularly so (W European Seas, Rockall Trough, Iceland Basin, and Norwegian Sea) with 75% of the collected samples. Much less sampled, but also smaller, regions, were the Reykjanes and Feni Ridge (16 and 4 samples, respectively). Regardless of the sampling efforts, 'stout-bodied' typhlotanaids were present in 10.6-33.6% of samples, with the exception of best sampled WES region where the 'stout-bodied' typhlotanaids occurred in only 2.2% of all samples. With current knowledge, it is difficult to conclude why 'stout-bodied' typhlotanaids are relatively lacking in the WES region. Because average temperatures (6.1 ± 1.6 °C) and mean depth (1274.3 ± 489.3 m) of WES are similar to those in Irminger Basin where 'stout-bodied' typhlotanaids were recovered in 13% samples, it is rather doubtful that such distribution is the effect of environmental selection. Certainly, this result deserves further and more detailed research, including modelling based on extensive data covering the other groups of Typhlotanaidae, the other families of Tanaidacea, but also other groups of Peracarida.

Distribution.
In the distribution of the six species with a preference for greater depths, a clear split is drawn between the fauna of the north and south (Figs. 5, 9 and 31). The abundant and frequent species i.e. C. igae, G. gudmundssoni, Sa. georgi, J. bioice. and L. martini, are present south of Iceland, and only H. inermis was located in northern sites. The seas surrounding Iceland are separated by the well-defined latitudinally-oriented topographic structure Greenland-Iceland-Faroe (GIF) with the deepest sill depth at 840 m between the Faeroe Islands and Scotland 22,96 . This ridge is known as a prominent topographic barrier in the N Atlantic that hampers distribution of water masses between the Arctic and the deep basins of the Nordic seas. The complex system of ocean currents and oceanic ridges around Iceland separates neighbouring areas, where temperature, salinity, availability of organic matter and oxygen can significantly vary 12,22,[97][98][99][100] , and thus shape habitat mosaic affects the species compositions of benthic invertebrates 13,23,[101][102][103] . Moreover, the GIF is known to hamper geneflow and hybridization effect for genetic population structure of isopods 104 . As a consequence, this distinct topographic structure becomes a distinct zoogeographical barrier shaping a region extremely sensitive to climate change 12,13,23,105 .
The low dispersal mobility and high taxonomic diversity of the tanaidacean family Typhlotanaidae make it an excellent model in biodiversity studies and an ideal indicator in assessing the effects of anthropogenic activities. The climate change in oceanic zones will cause numerous shifts in hydrological conditions, although the scale is difficult to assess due to the marginal understanding of the biodiversity of deep-sea ecosystems. The character and the scale of disturbances caused in the ecosystems by anthropogenic activity is unpredictable, but it is assessed that the changes will be irreversible and will last for decades. Knowledge of biodiversity by integrative approaches along with the geographic distribution of newly discovered species and their environmental preferences can help protect exposed warming sites in the marine ecosystem. We are strongly convinced that the data presented here is an important piece of knowledge in the study of climate change affecting the N Atlantic.

Materials and methods
Sampling. The 4043 typhlotanaids specimens for the research were collected during 18 expeditions, with the distribution of the stations given in Table S1. Apart from the samples taken during the IceAGE 1 and 2 expeditions, all samples were fixed (and possibly stored) with formalin 106,107 . Distribution maps were prepared using the QGIS 3.28 software 108 . The environmental variables were collected from the cruise reports 106 , publications 109,110 and complemented from World Ocean Atlas 2020 (https:// www. ncei. noaa. gov/ produ cts/ world-ocean-atlas). The variables for the defined region have been averaged. Correlation matrix of environmental variables reveals high correlation between depth, nitrogen, phosphorus, oxygen and (r > 0.88, p < 0.02) thus for the species distribution only temperature and depth were considered (Fig. S2).