Integrative redescription of Hypsibius pallidoides Pilato et al., 2011 (Eutardigrada: Hypsibioidea) with the erection of a new genus and discussion on the phylogeny of Hypsibiidae

An integrative redescription of Hypsibius pallidoides Pilato, Kiosya, Lisi, Inshina & Biserov, 2011 was undertaken following a reexamination of the type material and new material using highquality light microscopy, scanning electron microscopy and methods of molecular taxonomy. Detailed morphological investigations revealed a unique complex of characters that precluded the attribution of this species to the genus Hypsibius Ehrenberg, 1848. Furthermore, phylogenetic analyses indicated the affinity of this species within the subfamily Pilatobiinae (Hypsibiidae). Notahypsibius gen. nov. is erected for H. pallidoides and two putatively related species: H. scaber maucci, 1987 and Ramazzottius arcticus (murray, 1907). An emended diagnosis for the genus Pilatobius is given, while the subfamily Pilatobiinae lacks a cohesive morphological diagnosis despite representing, at the same time, a wellsupported molecular clade. obvious controversy between the results of the morphological and molecular analyses of the phylogeny of Hypsibioidea is discussed. The distribution of morphological characters such as the claw type, organization of the bucco-pharyngeal apparatus, and egg shell sculpture type within Eutardigrada is analyzed and their phylogenetic significance discussed.


Introduction
Phylum Tardigrada Doyère, 1840 is a group of widely distributed microscopical multicellular animals. Currently there are ca 1300 known species (Degma et al. 2019), but this is very likely an underrepresentation of the actual number of taxa, as the global diversity of tardigrades is considered poorly investigated (Bartels et al. 2016). Along with the description of new taxa, the redescription of known species using the integrative approach, i.e., combining a morphological analysis with methods of molecular taxonomy and phylogeny, is a promising way to improve our understanding of the real For scanning electron microscopy (SEm) specimens were thermally relaxed at 60ºC (morek et al. 2016), dehydrated in an ascending ethyl alcohol series (10%, 20%, 30%, 50%, 70%, 96%) and acetone, critical-point dried in Co 2 , mounted on stubs and coated with gold. The bucco-pharyngeal apparatus was prepared for SEM investigation following the protocol of Eibye-Jacobsen (2001) as modified by . A Tescan mIRA3 Lmu Scanning Electron microscope was used for observations (Centre for molecular and Cell Technologies, St Petersburg State university).

Morphometrics
The sample size for morphometrics was chosen following the recommendations of Stec et al. (2016a). All measurements are given in micrometres (μm). Structures were measured only if their orientations were suitable. Body length was measured from the anterior end of the body to the posterior end, excluding the hind legs. The bucco-pharyngeal tube was measured from the anterior margin of the stylet sheaths to the caudal end of the buccal tube, not including the buccal apophyses. Terminology for the structures within the bucco-pharyngeal apparatus and for the claws follows that of Pilato & Binda (2010). Elements of the buccal apparatus were measured according to Kaczmarek & michalczyk (2017). Claws were measured following Beasley et al. (2008), but the total length of the claws was also measured (according to Pilato et al. 2002) to maintain compatibility with the initial description. The pt index used is the percentage ratio between the length of a structure and the length of the buccal tube (Pilato 1981) and is presented here in italics. morphometric data were handled using ver. 1.6 of the "Parachela" template, which is available from the Tardigrada Register (michalczyk & Kaczmarek 2013) with total length of the claws added.
Genotyping DNA was extracted from 15 individual animals using QuickExtract Tm DNA Extraction Solution (Lucigen Corporation, uSA) using the following protocol (kindly provided by Torbjørn Ekrem, Norwegian university of Science and Technology). 1) Tardigrades were sorted in water and specimens were rinsed individually in ddH 2 o. 2) Each individual specimen was transferred by pipette into a PCR-tube containing 70 µl QuickExtract™.
3) PCR-tubes were vortexed well, spun down (5 min at 3500 RPm), then kept at room temperature (≈ 25ºC) for 2 hrs. 4) PCR-tubes were incubated at 65ºC for 15 min (in a PCR-machine), vortexed every 5 min and spun down. 5) PCR-tubes were incubated at 98ºC for 2 min. 6) 60 µl of the extract supernatant were transferred into a new, sterile PCR tube. The supernatant was collected in order to avoid the exoskeleton remaining at the bottom. The PCR-tubes containing extract were then stored at −20ºC for later use in PCR. 7) 70 µl ddH 2 o were added to the tube with the exuvium and mixed well with the pipette to wash the exoskeleton. 8) Water and exoskeleton were transferred to a glass staining block with ddH 2 o. The exoskeleton was collected and mounted on a microscope slide in Hoyer's medium and retained as the hologenophore (Pleijel et al. 2008).
Four genes were sequenced: a small ribosome subunit (18S rRNA) gene, a large ribosome subunit (28S rRNA) gene, internal transcribed spacer (ITS-2), and the cytochrome oxidase subunit I (CoI) gene. Both 18S rRNA and 28S rRNA are nuclear markers used in phylogenetic analyses to investigate high taxonomic levels (Jørgensen et al. 2010(Jørgensen et al. , 2011Guil & Giribet 2012;Bertolani et al. 2014;Guil et al. 2019;Gąsiorek et al. 2019bGąsiorek et al. , 2019c. CoI is a protein-coding mitochondrial marker that is widely used as a standard barcode gene of intermediate to high effective mutation rate (Bertolani et al. 2011b). ITS-2 is a non-coding nuclear fragment with high evolution rates used for both intra-specific comparisons and comparisons between closely related species Stec et al. 2018). A complete 18S rRNA gene was amplified in several overlapping fragments using primer pairs: SSU_F_04 and SSU_R_26, 18Sfw and rev960, fw390 and rev18S, 5F and 9R (for primer details see Table 1). These products were sequenced with PCR primers and the internal primers fw1230 and rev1460. A fragment of the 28S rRNA gene was amplified and sequenced using primers 28_F0001 and 28S_R1800. PCR reactions included 2 µl template DNA, 1 µl of each primer, 1 µl DNTP, 5 µl Taq Buffer (10X) (-mg), 4 µl 25 mm mgCl 2 and 0.2 µl Taq DNA Polymerase (Thermo Scientific™) in a final volume of 50 µl. The PCR cycling profile for the 18S and 28S genes was as follows: initial denaturation at 95ºC for 5 min, then 35 cycles of 95ºC for 1 min, 50ºC for 1 min, 72ºC for 2 min, and final elongation at 72ºC for 10 min. ITS-2 was amplified and sequenced using primers ITS2_Eutar_Ff and ITS2_Eutar_Rr . The PCR cycling profile for ITS-2 was as follows: initial denaturation at 95ºC for 3 min, followed by 30 cycles of denaturation at 95ºC for 1 min, annealing at 55ºC for 2 min, and elongation at 72ºC for 2 min, and a final elongation step at 72ºC lasting 10 min. A fragment of the COI mtDNA gene was amplified and sequenced using primers LCO1490 and HCO2198 (Folmer et al. 1994). The PCR cycling profile for the COI gene was as follows: initial denaturation at 94ºC for 5 min, followed by five cycles of denaturation 1 min at 94ºC, annealing at 42ºC for 1.5 min and amplification at 72ºC for 1.5 min; then 35 cycles of 94ºC for 1 min, 50ºC for 1.5 min, 72ºC for 1 min, and final elongation at 72ºC for 5 min. CoI sequences were translated to amino acids by using the invertebrate mitochondrial code implemented in mEGA7 (Kumar et al. 2016) in order to check for the presence of stop codons and therefore of pseudogenes.
PCR products were visualized in 1.5% agarose gel stained with Ethidium bromide. All amplicons were sequenced directly using ABI PRISm Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, uSA) on ABI Prism 310 Genetic Analyzer. Sequences were edited and assembled using ChromasPro software (Technelysium, uSA). maximum-likelihood (mL) topologies were constructed using the RaxmL 8.2.10 program (Stamatakis 2014) with GTR + γ + I model; the number of invariant sites, alpha parameter and tree topology were optimized by RAxmL, 1000 bootstrap pseudoreplicates were used. Bayesian analysis of the same datasets was performed using mrBayes ver. 3.2.6, GTR model with gamma correction for intersite rate variation (8 categories) and the covarion model (Ronquist & Huelsenbeck 2003). The analyses were run as two separate chains (default heating parameters) for 10 million generations, by which time they had ceased converging (final average standard deviation of the split frequencies was less than 0.01). The quality of chains was estimated using built-in mrBayes tools. jmodeltest, RaxmL and mrBayes programs were run at the Cipres ver. 3.3 web-site (miller et al. 2010). uncorrected pairwise distances were calculated using mEGA7 (Kumar et al. 2016) with gaps/missing data treatment set to "complete deletion".

Institutional acronyms
Specimens from the following institutions and collections were examined (curator in parenthesis).
Colour. Body uncoloured or whitish with green gut content. most specimens with eyespots, usually well-discernible after slide mounting ( Fig. 1A) but absent in some specimens. CephaliC sensory struCtures. Cephalic body portion with a pair of elliptical sensory organs developed in the form of flat porous areas, separated from the body surface with a oval cuticular groove. These structures are scarcely visible in Lm, but are well-discernible in SEm ( Fig. 3C-D, black arrowheads). Two indistinctly demarcated porous areas are also developed in the fronto-lateral region of the head, on the each side of the mouth opening (visible in SEm only; Fig. 4A-B, white arrowheads). Central concavity on the dorsal surface of the head (Fig. 3C, white arrowhead) seems to be similar to the structures present in some Isohypsibioidea (see Gąsiorek et al. (2019c: 91, fig. 4b, d) and in Cryobiotus roswithae (Dastych 2019).

Phylogenetic analysis
In the 18S DNA phylogenetic analysis, the order Hypsibioidea was highly supported and divided into two well-supported clades: clade I, embracing the family Ramazzottiidae, marley, mcInnes & Sands, 2011, and clade II, comprised of taxa currently attributed to the families Hypsibiidae Pilato, 1969, Calohypsibiidae Pilato, 1969and microhypsibiidae Pilato, 1998. Clade II was further divided into two subclades well-supported with Bayesian analysis, but weakly supported or unsupported with mL analysis. The first subclade included the families Microhypsibiidae s. str. (genus Microhypsibius Thulin, 1928) and Calohypsibiidae s.str. (genus Calohypsibius Thulin, 1928) (see Gąsiorek et al. 2019a for a discussion on the taxonomic composition of these two families), together with two genera of unclear taxonomic position, Acutuncus and Mixibius. The second subclade was divided into three subclades with unclearly resolved phylogenetic relationships. The first of these subclades included the species representing the subfamily Pilatobiinae Bertolani, Guidetti, marchioro, Altiero, Rebecchi & Cesari, 2014, Hypsbius pallidoides, and the species attributed to Hypsibius convergens (urbanowicz, 1925) by Guil & Giribet (2012). The second one included the species of the subfamily Itaquasconinae Rudescu, 1964 and the third one was comprised of two well-supported lineages, the subfamilies Diphasconinae Dastych, 1992 andHypsibiinae Pilato, 1969.
Analyses of the concatenated 18S + 28S sequences resulted in a phylogeny with the same tree configuration, but with slightly weaker support of the clades (see Supplementary file SM.02). This weakened support is possibly a consequence of the small number of sequences available for such analysis.

Comparison with the original description
morphometry of specimens from all analysed populations (including the type series) corresponds well with the data from the original description (Pilato et al. 2011). Small differences in the values of the stylet supports insertion point pt index (54.2-55.2 in the original description vs 56.9-63.3 in the material investigated) and the length of the first macroplacoid (3.8-4.2 µm (pt 15.5-17.0) vs 2.2-3.8 µm (11.2-15.9) respectively) should be considered as the result of some differences in the measuring process, taking into account that my own measurements of the type series specimens are concordant with those of the specimens from the other populations (see Table 2). It was stated in the original description (Pilato et al. 2011) that H. pallidoides had a smooth cuticle, but high quality Lm and SEm observations revealed the presence of a cuticular sculpture (Figs 2A-D, 3A-B). It is poorly visible in the type series specimens because of the intensive staining of soft tissues with acetocarmine during the slide preparation.
Contrary to the absence of lunules in H. pallidoides stated by Pilato et al. (2011), my investigation determined that scarcely visible lunules on the claws of legs I-III and well-developed wide lunules on the claws of legs IV are present (Fig. 6A, E). In the original description of the species, Pilato et al. (2011) indicated the absence of a cuticular bar between the claw bases of legs IV, but considered this as unconfirmed. My observations revealed the presence of a thickened zone of the posterior claw lunule, located between the anterior and the posterior claw bases (Fig. 6E-F). This thickening can give the impression of a cuticular bar in the case when the main part of the lunule is not discernible.
While Pilato et al. (2011) described the eggs of H. pallidoides as being smooth, further scrutiny ascertained the presence of a granular pattern formed by the system of internal pillars in the egg shell of this species (Fig. 7B-G).

New phenotypic differential diagnosis
Hypsibius pallidoides is similar to the species of the genera Ramazzottius Binda & Pilato, 1986 andCryoconicus Zawierucha, Stec, Lochowska-Cierlik, Takeuchi, Li & in having claws of the Ramazzottius type; AISm asymmetrical with respect to the frontal plane; cephalic elliptical sensory organs and in laying ornamented eggs. It clearly differs from all species of those genera by having wider primary branches of the external and posterior claws, with less pronounced differentiation between rigid distal and soft basal parts; the dorsal AISm raised and thickened in its anterior margin, and eggs laid in the exuvium without external processes, but with pillars inside the egg shell only.
Hypsibius pallidoides is similar to the species of the genus Mixibius in having AISm asymmetrical with respect to the frontal plane, where the ventral apophysis is similar, but not identical, to the "semilunar hook" of Hypsibius; dorsal apophysis more stumpy with a blunt and swollen caudal apex. Also a short median cuticular thickening caudal to both these apophyses is present (the ventral one slightly visible) (Pilato & Binda 2010). It clearly differs from all species of this genus by having: cephalic elliptical sensory organs and Ramazzottius-like claws (external claws with elongated primary branches and less developed secondary branches).
The type of egg shell sculpture of Hypsibius pallidoides is similar to that of Acutuncus antarcticus, from the Antarctic region (see Dastych 1991 for a review of the old records) in that the sculpture,   (Richters, 1911), egg shell (SPbU 203 (7)), PhC. Scale bars: A = 50 µm; B-H = 5 µm.
formed by the pillars within the egg shell, presents as a dot-like pattern when observed in Lm. Hypsibius pallidoides differs from A. antarcticus by having the Ramazzottius-type claws; AISm asymmetrical with respect to the frontal plane; a sculptured cuticle and a small dot-like septulum. The precise nature of the latter structure requires further investigation as its small size prevents it from being undoubtedly interpreted as microplacoid or septulum.
Two species of the genus Hypsibius are known as laying eggs with granulated chorion in exuvium, Hypsibius roanensis Nelson & mcGlothlin, 1993(Guidetti et al. 1999) and H. cf. scabropygus (Guidetti & Bertolani 2001). Hypsibius pallidoides clearly differs from both of these species by having a septulum, the Ramazzottius-like claws, and a different cuticular sculpture.

Genotypic differential diagnosis
The ranges of uncorrected genetic p-distances between the studied population of Hypsibius pallidoides and species of the order Hypsibioidea for which sequences are available from GenBank (see Supplementary file SM.01) are as follows: CoI: 20.9%-26.7% (mean 23.0%), with the most similar being Pilatobius recamieri (Richters, 1911 Sequences of the 18S and 28S rRNA genes, attributed to the species "Hypsibius convergens" by Guil & Giribet (2012) are nearly identical to those of Hypsibius pallidoides (p-distances 0.0% and 1.1% respectively).

Phylogeny of Hypsibioidea and phylogenetic position of Hypsibius pallidoides
The results of phylogenetic analysis presented herein correspond well with the molecular phylogenies of Tardigrada (2012)) were distinctly placed within the Pilatobiinae clade, and even more interestingly within the genus Pilatobiotus itself, forming a cluster with the species P. patanei (Binda & Pilato, 1971) / P. islandicus/ P recamieri, while the species P. ramazzottii (Robotti, 1970) and P. nodulosus (Ramazzotti, 1957) formed a separate paraphyletic group. Grouping of the species attributed to H. convergens with Pilatobius recamieri was obtained by Guil et al. (2019), but this result was not discussed by the authors. In an earlier publication (Guil & Giribet 2012), the taxon misidentified with H. convergens was joined with Astatumen trinacriae (Arcidiacono, 1962), but this result is likely an artefact because no species of Pilatobius were used in the analysis. In my opinion, extreme similarity of the 18S and 28S sequences of this species to the sequences of H. pallidoides (p-distances 0.0% and 1.1% respectively) should be considered as evidence of their identity on the genus level. Hypsibius pallidoides is morphologically similar to H. convergens and could be misidentified with this species, especially when temporary slides were used for the identification (Guil & Giribet 2012), because of the poor visibility of the cuticular sculpture and septulum in living specimens.
As a result, in the case of H. pallidoides we have a distinct contradiction between the morphological and molecular taxonomical approaches. Analysis of the morphological traits of this species reveals similarities with Ramazzottiidae (i.e., presence of the cephalic elliptical organs, the Ramazzottius-like claws, asymmetry of the AISm), but, according to the analysis of the gene sequences, this species should be attributed to the subfamily Pilatobiinae. Its position in the obtained phylogenetic tree also supports the presumably paraphyletic nature of the genus Pilatobius, also inferred by Gąsiorek et al. (2018). To my knowledge, this is the first occurrence of such a distinct controversy between morphological and molecular taxonomy within Tardigrada. Previously, genetic analyses have supported the erection of taxa recognized by traditional morphological analysis (e.g., genera Paramacrobiotus Guidetti, Schill, Bertolani, Dandekar & Wolf, 2009, Mesobiotus Vecchi, Cesari, Bertolani, Jönsson, Rebecchi & Guidetti, 2016, Acantechiniscus Vecchi, Cesari, Bertolani, Jönsson, Rebecchi & Guidetti, 2016 family Ramazzottiidae, order Isohypsibioidea) (Guidetti et al. 2009;Vecchi et al. 2016;Sands et al. 2008) or provided an opportunity to resolve the phylogeny of a group when morphological data were insufficient (e.g., the clarification of the phylogenetic position of the genera Apodibius Dastych, 1983 andHaplomacrobiotus may, 1948, revisions of Isohypsibioidea and Echiniscus C.A.S. Schultze, 1840) (Dabert et al. 2014;Cesari et al. 2016;Gąsiorek et al. 2019bGąsiorek et al. , 2019c. The presence of such controversy is a problem that has been acknowledged in current zoology since molecular methods began to be widely used (Hillis 1987;Osawa et al. 2004;Smirnov et al. 2005;Cohen 2018). Various authors who have analysed this problem (Hedges & Sibley 1994;Scotland et al. 2003;osawa et al. 2004;Wiens 2004;Smith & Turner 2005) came to the conclusion that the best (if not the only) way to align the conflicting morphological and molecular phylogenies is to improve the morphological data by involving new characters in the analysis and by re-evaluating some characters already in use.
Taking into account the unique combination of the morphological features and the phylogenetic position of Hypsibius pallidoides distant from the remaining species of Hypsibius, as demonstrated by the analysis of the molecular data, the erection of the new genus Notahypsibius gen. nov. for the species H. pallidoides is proposed.

Type species
Hypsibius pallidoides Pilato, Kiosya, Lisi, Inshina & Biserov, 2011 Diagnosis Hypsibiidae with Ramazzottius-like claws and completely rigid buccal tube. Apophyses for the insertion of the stylet muscles asymmetrical, dorsal AISm shorter and higher than ventral, with thickened anterior margin. Pharynx with two elongated macroplacoids and minute dot-like septulum. Cephalic elliptical organs present. Rugose cuticular sculpture. Eggs laid within the exuvium (or freely?), chorion with developed pillar-like internal structure visible in Lm.

Etymology
The name refers to the phylogenetic position of the new genus, the type species of which was originally described as belonging to genus Hypsibius, but according to the phylogenetic analysis, definitely is "not a Hypsibius".

Geographical distribution
This species was described from Kherson oblast, South ukraine (Pilato et al. 2011). Later it was recorded for the minsk oblast, Central Belarus (Pilato et al. 2012) and Sicily (Lisi 2015). my observations extend the distribution of this species to North-West Russia (St Petersburg and Karelia), Croatia and Austria (Carinthia). It should be noted that Dastych (1988) observed a configuration of the bucco-pharyngeal apparatus similar to N. pallidoides gen. et comb. nov. in some Polish specimens attributed by him to H. convergens (Dastych 1988: 147, pl. XXIa, c). Also, some of the microphotographs of the claws of Dastych's specimens of H. convergens show a similarity to those described for N. pallidoides gen. et comb. nov. (Dastych 1988: pl. XXIi). So, it is very likely that the latter species is present among the tardigrade fauna of Poland. The species attributed to as "H. convergens" by Guil & Giribet (2012) is nearly identical to N. pallidoides gen. et comb. nov. in 18S and 28S gene sequences (see Genotypic differential diagnosis). In my opinion, this is evidence for the presence of N. pallidoides gen. et comb. nov. in Spain, but it was recently shown (Guidetti et al. 2019a) that closely related species can share an identical 18S rRNA haplotype. Thus, without analyses of the more sensitive barcode genes (particularly CoI and ITS-2), and in the absence of morphological data, the possibility of the presence of another species similar to N. pallidoides gen. et comb. nov. in Spain cannot be excluded.

Notes
The species described as Hypsibius scaber maucci, 1987 (known from North America only; maucci 1987) is very similar to N. pallidoides gen. et comb. nov. in having cuticular sculpture consisting of irregular ridges (Fig. 9A-B, D-E), highly differentiated external and internal claws that closely resemble the Ramazzottius-type claws (Fig. 9H-I), and a similar bucco-pharyngeal apparatus with a thin buccal tube and minute dot-like septulum (Fig. 9C, F). In my opinion, it should be transferred to the genus Notahypsibius gen. nov. as Notahypsibius scaber gen. et comb. nov. It seems also that N. scaber gen. et comb. nov. has cephalic elliptical organs (Fig. 9G), but their presence requires further confirmation because of the difficulties in observing the dorsal surface of the type specimens (R. Guidetti, pers. com.). Notahypsibius pallidoides gen. et comb. nov. differs from N. scaber gen. et comb. nov. in having a less developed cuticular sculpture, especially on the ventral side of the body (Fig. 9D-E), and in having external claws with the common base thinner and longer in relation to the secondary branch. It should be noted that the comparison of N. pallidoides gen. et comb. nov. with N. scaber gen. et comb. nov. cannot be considered to have definitively resolved the possible synonymy of these species. The latter species description was based on only two specimens, both of which have most of their claws in positions that obstruct observation and correct measurement. Also, the eggs of N. scaber gen. et comb. nov. are unknown.  (murray, 1907) gen. et comb. nov. Murray, 1907b: 677, pl. 1, fig. 5a-f (description). Macrobiotus heinisi Richters, 1911: 15, fig. 15.

Notes
The species Hypsibius arcticus (Murray, 1907) was recently transferred by Gąsiorek et al. (2018) to the genus Ramazzottius on the basis of having Ramazzottius-like claws and freely laid eggs. In my opinion, the type of the chorion ornamentation in this species, consisting of the internal pillars, is definitely different from the external processes that are typical of the genus Ramazzottius (see Discussion). The combination of the Ramazzottius-like claws and eggs with developed internal pillars in the egg chorion makes this species more similar to N. pallidoides gen. et comb. nov. It should, therefore, be transferred to the new genus as Notahypsibius arcticus gen. et comb. nov. This species differs from N. pallidoides gen. et comb. nov. by having better developed pillars in the egg chorion and by laying free eggs. However, the latter trait requires confirmation as it is assumed upon the basis of a single observation (Murray 1907b), especially taking into consideration that Macrobiotus heinisi (Richters, 1911) -a similar species described from the same locality (Franz Joseph Land) and later synonymized with H. arcticus (marcus 1930) -has eggs with a similar chorion structure laid within the exuvium (Richters 1911). Also, Acutuncus antarcticus, which shares an eggshell structure of a similar appearance, is known to lay eggs both within the exuvium and freely (Dastych 1991;pers. obs.). other key characteristics, such as the presence of the minute septulum and cuticular sculpture, may have been overlooked by murray (1907b) in his original description, as visualisation of these structures requires the use of high quality optics unavailable at that time. The specimen and the egg from Scotland, which were described and figured by Murray (1907a: 658, pl. IV, fig. 27a-d) as "Macrobiotus oberhäuseri Doy. ?", could not be attributed to N. arcticus gen. et comb. nov. because of the evident differencies in the claw structure (claws similar to the Cryoconicus type), the egg chorion appearance (much shorter pillars), and the significant difference in the value of the pt index for the stylet support insertion point (57% in Scottish specimen vs 70% in N. arcticus gen. et comb. nov., measurements taken from the original Murray's drawings). This material possibly represents an undescribed species of the genus Cryoconicus.

Phylogenetic significance of some morphological characters
Ramazzottius-like claws morphology of the claws is one of the most important characters used in the taxonomy of Eutardigrada (Pilato 1969;Schuster et al. 1980;Pilato & Binda 2010;Gąsiorek et al. 2019c). Ramazzottius type claws were recognized as a separate morphotype (and denominated as "oberhaeuseri type" claws) by Binda & Pilato (1986) with the simultaneous erection of the genus Ramazzottius and were recently reanalysed by Guidetti et al. (2019b). Two other genera, Ramajendas Binda, 1990 andThalerius Dastych, 2009, were recognized as having a similar claw morphology (Pilato & Binda 1990;Dastych 2009). The phylogenetic and taxonomic position of these two genera is currently the subject of debate because of the evident controversy in their morphology and lack of DNA sequences. Being initially placed within Isohypsibiidae (marley et al. 2011;Guil et al. 2013), both genera were later attributed to the family Ramazzottiidae by Bertolani et al. (2014) on the basis of claw morphology. Zawierucha et al. (2018), taking into account the simple ridge-like form of the apophyses for the insertion of the stylet muscles and the deposition of smooth eggs in the exuvium known in the genus Ramajendas, proposed to place both of these genera back within Isohypsibiidae, suggesting the independent evolution of the Ramazzottius-like claws. Thus, for clarity, I use here the term "Ramazzottius-like claws" in order to distinguish the claws of Ramajendas, Thalerius and Notahypsibius gen. nov. from the Ramazzottiustype claws of Ramazzottiidae (see Guidetti et al. 2019b for a discussion). In a recent revision of the order Isohypsibioidea, Ramajendas and Thalerius are considered as incertae sedis pending molecular verification of their taxonomic positions (Gąsiorek et al. 2019c). In this situation, the obtained data showing the independent evolution of the Ramazzottius-like claws within the Pilatobiinae clade should be considered as an argument in favour of the hypothesis that Ramajendas and Thalerius are positioned phylogenetically distant from Ramazzottiidae.

Cephalic elliptical organs
Binda & Pilato (1986) pointed out the presence of these structures in the genus Ramazzottius and compared them with the papillae in the cephalic region of Calohypsibius ornatus (Richters, 1900). Since the genus Fractonotus Pilato, 1998, in which these structures are also known, was revised and transferred to the family Isohypsibiidae (Gąsiorek et al. 2019a(Gąsiorek et al. , 2019c the cephalic elliptical organs are currently known within two Eutardigrada orders, Hypsibioidea (Calohypsibius and Notahypsibius gen. nov.) and Isohypsibioidea (Fractonotus). These structures should not be confused with 'cephalic papillae' or 'frontal lobes' known in several Isohypsibiidae genera (Halobiotus Kristensen, 1982, Apodibius, Ursulinius Gąsiorek, Stec, Morek & Michalczyk, 2019and Paradiphascon Dastych, 1992 (Gąsiorek et al. 2019c) regarding the much more rostral position of the latter structures (within the anteriormost cephalic pseudosegment). Cephalic elliptical organs are located more caudally on the second pseudosegment following the cephalic one. In my opinion, only fronto-lateral porous areas of N. pallidoides (Fig. 4A-B) can be matched with the 'cephalic papillae' or 'frontal lobes' of Isohypsibiidae. Taking into account that cuticular sensory structures of Eutardigrada are very likely homologous to the cephalic sensory structures of Heterotardigrada (Zantke et al. 2008), the presence of such organs should be considered a plesiomorphic state and so supports the hypothesis of the basal phylogenetic position of Hypsibioidea and Isohypsibioidea (Marley et al. 2011;Gąsiorek et al. 2019c).

Chorion structure
Although egg shell structure has been considered a valuable taxonomic character within Eutardigrada from the early years of its investigation (marcus 1929, 1936Ramazzotti & maucci 1983), the phylogenetic significance of this trait was revealed considerably later (Bertolani et al. 1996). In their analysis, Bertolani and colleagues identified two types of the organization of the egg chorion -"smooth" and "ornamented" -and attributed A. antarcticus eggs to the ornamented type. In my opinion, the boundary between these two egg shell morphotypes within Tardigrada is not so obvious and the delimitation should be different.
Following the transfer of R. arcticus to the genus Notahypsibius gen. nov., there are only three genera known within Hypsibioidea with an egg chorion internal structure consisting of numerous pillars that connect the outer and inner layers of the shell: Acutuncus (Dastych 1991), Notahypsibius gen. nov. (present paper), and Pilatobius (present paper, see below). Additionally, two species of the genus Hypsibius (H. roanensis Nelson & mcGlothlin, 1993 and H. cf. scabropygus) have finely granulated egg shells (Guidetti et al. 1999;Guidetti & Bertolani 2001). The similarity to the visible external pattern of the egg shell in Acutuncus and Notahypsibius gen. nov. makes it possible to suppose the same structure of the egg chorion for these two species. ultrastructural investigations (Eibye-Jacobsen 1997; Poprawa 2011;Janelt et al. 2019) of egg development in Halobiotus crispae Kristensen, 1982, Grevenius granulifer (Thulin, 1928) and Thulinius ruffoi (Bertolani, 1982) revealed the presence of the distinct pillars connecting the inner and outer layers of the chorion in the species with the egg shell usually considered to be smooth. As a result, the only difference between the typical 'smooth' eggs of most of the Hypsibioidea and Isohypsibioidea and the eggs with visible pillars within the shell (Acutuncus, Notahypsibius gen. nov., H. roanensis and H. cf. scabropygus) is the degree of the pillars' development making them visible in LM. In my opinion, this trait could have often been omitted in older observations of the eggs of other species due to the insufficient quality of the optics and the prevailing opinion that eggs laid in the exuvium are always smooth. For example, a careful investigation of the eggs of Pilatobius recamieri revealed the presence of the same type structure of the egg chorion (Fig. 7H).
The same structure of the egg shell was also described in macrobiotidae Thulin, 1928(Poprawa et al. 2015. using the Transmission Electron microscopy, the presence of the pillars connecting the outer and inner layers of the egg shell was revealed in Macrobiotus polonicus Pilato, Kaczmarek, michalczyk & Lisi, 2003. This species has a continuous external layer of the egg shell, while in other species of the Macrobiotus hufelandi group this layer is modified to a mesh-like structure, supported by the pillars (Fig. 10A-B, white arrowheads). It also seems that such internal pillars form the dot-like pattern often visible in Lm between the egg processes in some species of macrobiotidae. For example, eggs of Tenuibiotus voronkovi (Tumanov, 2006) have a distinct dot-like pattern visible in Lm, between the processes bases (Fig. 10C, black arrowhead; not mentioned in the original description), while SEm shows the absence of any granulation on the egg surface (Fig. 10D).
In my opinion, the presence of egg shell pillars visible in Lm should be considered as a state, poorly delimited from the completely smooth egg shell (with pillars visible in Em only), and clearly different from the presence of true ornamentation consisting of the external processes. The development of the external processes does not exclude the presence of the internal pillars-like structures in the shell, but often these structures undergo progressive development, forming a mesh-like system of trabecules denoted as the "labyrinthine layer" (Węglarska 1982;Poprawa 2005).
It should also be noted that some tardigrade species have eggs with large, protruding pillar-like structures in the shell enclosed within the thin outer membrane. Egg shells of this type are known within both Heterotardigrada Marcus, 1927(Oreella murray, 1910 and Eutardigrada (macrobiotidae, murrayidae Guidetti, Rebecchi & Bertolani, 2000, Eohypsibiidae Bertolani & Kristensen, 1987 (Bertolani et al. 1996;Dastych et al. 1998). In my opinion, this type of egg shell structure could be derived from the primitive three-layered shell as a result of the progressive development of the internal pillars. This hypothesis can explain the emergence of similar-looking structures in several phylogenetically distant groups and possibly can partially resolve the known paradox in the tardigrade systematics formulated by Guidetti et al. (2006): "…there are closely related species, which share a very similar morphology of the animals but clearly differ in their egg morphology. Conversely, there are species belonging to different evolutionary lines that have similar eggs, but very different adult morphology". Surely, wide comparative TEm investigation of the egg chorion of different tardigrade species is needed to check this hypothesis.
It is interesting to note that, while within Eohypsibiidae the type of the egg shell ornamentation is genusspecific (pillar-like structures in Eohypsibius Kristensen, 1982 and external processes with labyrinthine layer in Bertolanius Özdikmen, 2008 and Austeruseus Trygvadóttir & Kristensen, 2011(Trygvadóttir & Kristensen 2011Hansen et al. 2017)), in macrobiotidae and murrayidae two genera include species with both types of the egg shell -Murrayon Pilato, 1988 andMinibiotus Schuster, 1980. The polyphyly of the genus Murrayon was previously demonstrated via molecular analyses (Bertolani et al. 2014;. It incorporates at least two clades -one with the pillar-like structures of the egg shell (M. dianeae (Kristensen, 1982) and related) and the other with the external conical processes (M. pullari (murray, 1907) and related). unfortunately, no molecular data are available on the species M. ovoglabellus (Biserov, 1988), which is known to lay smooth eggs. The genus Minibiotus is also suspected to be polyphyletic (Guidetti et al. 2007;Bertolani et al. 2014;Stec et al. 2015), but new molecular data are needed to test this hypothesis.

Bucco-pharyngeal apparatus of the Diphascon model
In the revision of the genus Diphascon, Pilato (1987) accepted the hypothesis of several independent origins of the long buccal tube with a flexible caudal part within Hypsibiidae as the most likely. He considered as a less likely alternative, the hypothesis of the presence of an independent monophyletic group within Hypsibiidae with a Diphascon-like buccal tube, where the shape of the apophyses for the insertion of the stylet muscles became identical to that of some species with a rigid buccal tube.  (Tumanov, 2006) with granulation-like pattern visible between the egg processes (SPbu 205(4)), PhC. D. Fragment of the egg shell of T. voronkovi in SEm, note the absence of the surface granulation. Scale bars: A-B = 2 µm; C-D = 5 µm.
The presence of the bucco-pharyngeal apparatus of the Diphascon model within macrobiotidae and Eohypsibiidae was considered as a confirmation of the possibility of the independent evolution of this trait. But after this work was published, a much wider distribution of the Diphascon-like buccal tube within Eutardigrada sensu lato (including the order Apochela Schuster, Nelson, Grigarick & Christenberry, 1980) has been revealed. Now, it is also known in three of four genera of milnesiidae Ramazzotti, 1962(Dastych 2011 and in the possibly closely-related Carphania Binda, 1978(Binda & Kristensen 1986); in all major clades of Hypsibioidea: Ramazzottiidae, Diphasconinae, Itaquasconinae, and Pilatobiinae (Bertolani et al. 2014); in several genera of macrobiotidae, possibly presenting different phylogenetic lines (Guidetti & Pilato 2003); in one of three known genera of Eohypsibiidae (Kristensen 1982) and, possibly, in Isohypsibioidea, if Paradiphascon is treated as belonging to that order (Gąsiorek et al. 2019c). Consequently, a new hypothesis of the presence of the flexible pharyngeal tube as a plesiomorphy of the whole Eutardigrada sensu lato, was suggested by Bertolani et al. (2014). In my opinion, the position of the genus Notahypsibius gen. nov. on the obtained phylogram within the morphological genus Pilatobius can be considered as evidence of a possible reduction of the caudal flexible part of the buccal tube within the taxon with an initially Diphascon-like buccal tube (such a reduction was also recently hypothesized by Gąsiorek & Michalczyk (2020) for the subfamilies Hypsibiinae and Itaquasconinae). As so, it strongly supports the hypothesis of the initially bipartite construction of the buccal tube within Eutardigrada sensu lato and independent reduction of the caudal flexible part in different phylogenetic lines.

Phylogeny and taxonomy of Pilatobiinae
The subfamily Pilatobiinae was established by Bertolani et al. (2014) when the phylogenetic analysis of 18S and 28S gene markers revealed that several species, previously attributed to the genus Diphascon (D. nodulosum, D. patanei and D. ramazzottii) form a separate clade within Hypsibiidae. These species, together with morphologically similar species of the genus Diphascon, were moved to a newly established genus Pilatobius. morphological diagnosis of the subfamily Pilatobiinae, given by Bertolani et al. (2014), was based on the characters of the genus Pilatobius as it was the only genus within this clade at that time.
The phylogenetic analysis shown herein involved two additional Pilatobius species (P. recamieri and P. islandicus) with recently obtained gene sequences (Gąsiorek et al. 2017;Buda et al. 2018). The presence of the separate clade forming the subfamily Pilatobiinae was confirmed, but the analysis showed that the genus Notahypsibius gen. nov. was positioned within this clade and, moreover, within the genus Pilatobius. In this situation, the subfamily Pilatobiinae is still valid in terms of being a wellsupported clade, but in lack of a suitable morphological diagnosis. No morphological characters can be pinpointed as an autapomorphy of this taxon. The only character possibly shared by Notahypsibius gen. nov. and Pilatobius is the presence of the pillars of the egg shell, visible in Lm (known only for one species of Pilatobius), but this trait should not be considered significant because it is known to be present in other species of the family Hypsibiidae, belonging to the different clades.
The genus Pilatobius appears to be paraphyletic, as it consists of the monophyletic clade P. islandicus / P. recamieri / P. patanei / N. pallidoides and two species (P. nodulosus and P. ramazzottii) being sister to this species complex. Taking into account the small number of species of the genus Pilatobius with known gene sequences available for inclusion in the phylogenetic analysis (5 of 26 species), I prefer not to change the taxonomical status of this genus, and instead leave it with the diagnosis given to Pilatobiinae by Bertolani et al. (2014) with the following redaction: "Genus Pilatobius Bertolani et al. 2014. Buccal tube followed by an annulated pharyngeal tube, with a drop-like thickening between them; pharyngeal bulb roundish or slightly oval, always containing 2 macroplacoids similar in length and in rows that look as parentheses, and a septulum. Claws of the Hypsibius type."