A huge undescribed diversity of the subgenus Hystricochaetonotus (Gastrotricha, Chaetonotidae, Chaetonotus) in Central Europe

Abstract. The subgenus Hystricochaetonotus Schwank, 1990 is one of the most species-rich subgenera of Chaetonotus Ehrenberg, 1830. It has a worldwide distribution and encompasses 37 species predominantly living in the benthos and periphyton of limnetic habitats. We have discovered further nine new species in running and stagnant waters in Slovakia (Central Europe): Ch. (H.) arcanus sp. nov., Ch. (H.) avarus sp. nov., Ch. (H.) gulosus sp. nov., Ch. (H.) iratus sp. nov., Ch. (H.) luxus sp. nov., Ch. (H.) mirabilis sp. nov., Ch. (H.) optabilis sp. nov., Ch. (H.) slavicus sp. nov., and Ch. (H.) superbus sp. nov. Their morphology was studied using differential interference contrast microscopy and subsequent morphometric analyses were carried out. In addition, the primary and secondary structures of their 18S, ITS2, and 28S rRNA molecules as well as their barcoding mitochondrial gene encoding for cytochrome c oxidase (COI) were analyzed. Species boundaries were tested also using the compensatory base change analysis. The new species could be well separated both morphologically and molecularly. The present barcoding analyses revealed that the nuclear ITS2 sequences represent a powerful DNA barcode in addition to the mitochondrial COI gene. According to the multi-gene phylogenetic analyses, the lineage leading to the last common ancestor of the ‘Hystricochaetonotus’ clade is the longest internal branch within the family Chaetonotidae Gosse, 1864. Since members of the subgenus Hystricochaetonotus are morphologically highly heterogeneous, parallel evolution of Chaetonotus-like and/or Hystricochaetonotus-like characters of scales and spines occurred during its radiation.

Dispersal abilities have also a significant impact on the diversification rate and alfa diversity detected in individual study plots. Dispersion of meiofaunal organisms is quite limited due to their small size, absence of a larval phase, and low movement capacity (Giere 2009;Cerca et al. 2018;Magpali et al. 2021). 'Cosmopolitan' species could comprise cryptic, near-cryptic (subtle, easily overlooked differences, or only statistically supported differences; Lücking et al. 2021), or pseudo-cryptic species (overlooked or briefly mentioned and not effectively characterized morphological characters; Magpali et al. 2021) that could significantly affect the diversity estimates. To properly address the distribution patterns and species identities of gastrotrichs, broad sampling and sound taxonomy are indispensable.
Morphology-based identification of gastrotrichs is, however, hampered by (1) the lack of standardized methods for species descriptions in the past, (2) a great disparity of species descriptions among authors, (3) the briefness and generality of 'old' descriptions, (4) the over-schematization and simplification of illustrations, and (5) the necessity of detailed examination of living specimens (Kisielewski 1991;Balsamo et al. 2008;Garraffoni & Melchior 2015;Kieneke & Nikoukar 2017;Magpali et al. 2021 and references cited therein). Nowadays, an integrative approach combining detailed morphological observations (e.g., differential interference contrast optics, confocal laser scanning microscopy, SEM) with multi-gene data is the preferred way for the description of new gastrotrich species (e.g., Kånneby 2011;Todaro et al. 2012;Garraffoni et al. 2017Garraffoni et al. , 2019aKolicka et al. 2016Kolicka et al. , 2018Kolicka 2019aKolicka , 2019bBosco et al. 2020;Magpali et al. 2021). Molecular data are, indeed, crucial for the recognition of cryptic and near-cryptic species, as it is the very definition of cryptic species that it is difficult, if not impossible, to find morphological characters that distinguish among them.
Using the integrative morpho-molecular approach, we have noticed a huge undescribed diversity of gastrotrichs belonging to the family Chaetonotidae Gosse, 1864 in stagnant as well as running waters in Slovakia. Besides morphological and phylogenetical analyses, we employed also the information contained in the secondary structure of the nuclear rRNAs and ITS2 molecules to further test species boundaries of morphologically delimited species. Especially, compensatory base changes (CBC), which force pairing in a helix to remain after a mutation occurred, are taxonomically important (Müller et al. 2007;Coleman 2009;Wolf et al. 2013). Just a single CBC in helix III of the ITS2 molecules is suggestive of the existence of two biological species incapable of mating.

Secondary structure predictions
The secondary structures of the 18S and 5.8S-28S rRNA molecules were predicted with R2DT (Sweeney et al. 2021). The R2DT package analyses RNA sequences using covariance models, taking into account a library of over 3000 templates representing the majority of known structured RNAs. The generated 2D layouts are consistent with 3D ribosomal structures and hence account also for the presence of noncanonical base pairs along with Watson-Crick pairings (Petrov et al. 2014). Since helix 25 of the 28S rRNA molecule is highly divergent, its secondary structure was studied also using the thermodynamic and homology modeling on the Mfold web server ver. 3.0 (http://www.unafold.org/) (Zuker 2003). The helix number system of rRNA molecules followed Petrov et al. (2014).
The common folding pattern of ITS2 molecules was found by running the multilign and TurboFold algorithms on the RNAstructure webserver (http://rna.urmc.rochester.edu/RNAstructureWeb) (Mathews 2004). Consequently, ITS2 molecules were folded on the Mfold web server, using the free-energy minimization approach and homology modeling. The 50%-majority rule consensus ITS2 secondary structure model was built in 4SALE based on the 2D-based alignment. The online program WebLogo ver. 2.8.2 (https://weblogo.berkeley.edu/) (Crooks et al. 2004) served to prepare the visualization of the nucleotide frequencies at individual positions of ITS2 helices with the relative entropy method. The total length, GC content, length of helices, unpaired bases in bulges and loops, number of bulges, and GU pairings were counted and statistically evaluated. Gibbs free energy ΔG of the ITS2 molecules was calculated with the webserver program RNAeval ver. 2.4.13 (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAeval.cgi) (Lorenz et al. 2011(Lorenz et al. , 2016. The tertiary structure of ITS2 molecules was predicted from the secondary structure with RNAComposer ver. 1.0 (http://rnacomposer.cs.put.poznan.pl/) (Popenda et al. 2012).

Phylogenetic analyses
Taxon sampling mostly followed Kånneby et al. (2012) and Kolicka et al. (2020), but only chaetonotids having sequences from 18S, 28S, and COI were included in phylogenetic analyses as specified in our previous study (Križanová & Vďačný 2021). All sequences, except for the newly obtained ones, were retrieved from GenBank (https://www.ncbi.nlm.nih.gov/genbank/) and their accession numbers are provided in Supp. file 1: Table S1. The nuclear rRNA gene sequences were aligned on the MAFFT ver. 7 server (https://mafft.cbrc.jp/alignment/server/) (Katoh et al. 2019), using the G-INS-i strategy, the 200PAM/κ=2 scoring matrix, and the gap opening penalty at 1.53. The mitochondrial protein-coding COI sequences were aligned with MEGA X (Kumar et al. 2018), using the invertebrate mitochondrial genetic code and the Muscle codon algorithm.
Two probabilistic methods, maximum likelihood and Bayesian inference, were utilized to build phylogenetic trees. Maximum likelihood (ML) analyses were carried out in IQ-TREE ver. 1.6.10 (Nguyen et al. 2015) on the webserver (http://iqtree.cibiv.univie.ac.at/) (Trifinopoulos et al. 2016), while Bayesian inferences (BI) were performed in the program MrBayes ver. 3.2.7 (Ronquist et al. 2012). Settings in ML analyses were as follows: (1) the best substitution model, as selected under the Bayesian information criterion by the in-built program, was assigned to each molecular marker, (2) the edge-unlinked partition model that accounts for heterotachy (rate variation across sites and lineages) and allows each partition to have its own set of branch lengths, (3) thousand ultrafast bootstrap pseudoreplicates, (4) the bnni algorithm that reduces overestimating bootstrap support (Hoang et al. 2018), and (5) all other parameters were left default. Settings of Bayesian analyses were as follows: (1) prior parameters of evolutionary models as estimated with IQ-TREE were implemented with the 'prset' European Journal of Taxonomy 840: 1-93 (2022) 8 command, (2) all model parameters were unlinked across partitions, (3) five million Markov chain Monte Carlo (MCMC) simulations, (4) a sampling frequency of trees and parameters at one hundred, and (5) a relative burn-in fraction of 25%. Convergence of the MCMC analyses was confirmed with the in-built diagnostics of the program MrBayes. All trees were computed as unrooted and were rooted in FigTree ver. 1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/). ) were detected at two localities, while all other species were found only at a single spot each. Interestingly, two or three species co-occurred at four out of the six localities studied (Table 1).

Molecular identification
In total, 72 new sequences were obtained from the nine new Chaetonotus (Hystricochaetonotus) species (Table 2). They can be unambiguously distinguished by the primary structure of ITS2, 28S, and COI sequences. According to the present barcoding analyses, 28S can serve as a pre-barcode and ITS2 and COI as barcodes ( Fig. 4B-D). Our two-step barcoding approach thus comprises a preliminary identification using a universal eukaryotic barcode (28S), called the pre-barcode, followed by a more precise species-level assignment using group-specific barcodes (ITS2 and COI). Based on the present barcoding analysis, a combination of ITS2 and COI is preferred but not required for the determination of species within the subgenus Hystricochaetonotus. These two barcodes thus can be used for species identification also separately. Table 1. Occurrence of nine new species of the subgenus Hystricochaetonotus Schwank, 1990 at six collection sites. a For abbreviations of collection sites, see Material and methods. Plus sign (+) means presence, minus sign (-) absence.

Species
Collection  In addition, two heterozygotes were detected in the former species, which had cytosine and guanosine at position 1729 in helix 44. This position is apparently polymorphic in Ch. (H.) luxus sp. nov. Both nucleotide states are involved in a noncanonical interaction that retains the RNA helical structure. Interspecies p-distances typically spanned a range of 0.05-4.39%, whereby no difference was found only between Ch. (H.) luxus sp. nov. and Ch. (H.) iratus sp. nov. Nevertheless, due to some overlap between intra-and interspecies distances, 18S cannot be used as a reliable DNA barcode for this group of gastrotrichs (Fig. 4A). The 18S secondary structure models of the nine new gastrotrichs are shown in Figure 5 and Supp. file 1: Figs S1-S8. Diagnostic autapomorphies are marked by arrows. The distinctness of the majority of species is also strengthened by 1-16 CBCs (Table 3).

Fig. 4.
Histograms showing intra-and interspecies p-distances of four molecular markers studied in nine new gastrotrich species. A. Barcoding analysis of 18S rRNA gene sequences. The black arrow marks the overlap between intra-and interspecies p-distances. B. Barcoding analysis of ITS2 molecules. Sequences are identical within species (red arrow), while interspecies p-distances range from 5.88% to as much as 34.76%. C. Barcoding analysis of the first two domains of 28S rRNA gene. Although there is no overlap between intra-and interspecies p-distances, no distinct barcoding gap is recognizable. D. Barcoding analysis of COI showing the distinct gap between intra-and interspecies p-distances (red arrow).

Fig. 5.
Secondary structure of the 18S rRNA molecule of Chaetonotus (Hystricochaetonotus) mirabilis sp. nov. The single diagnostic molecular autapomorphy (red arrow) is situated in the terminal loop of helix 21es6c in the V4 region of the C domain. Note the long-range tertiary contacts between the V2 and V4 regions (highlighted in pink). The reference 18S secondary structure map of Saccharomyces cerevisiae Meyen ex E.C.Hansen (inset) is from http://apollo.chemistry.gatech.edu/RibosomeGallery (Petrov et al. 2014).  Schwank, 1990. A. Putative consensus secondary structure. The central loop radiates four unequally long helices. The first two helices are highly conserved, while the two following helices are much less conserved. For localization of helix 10, which is made by the interaction of 3'-end of 5.8S and 5'-end of 28S, see Fig. 7. B. Two different views on the tertiary structure, showing that helices II and IV run in parallel. C. Structure logos of helices I-IV. The height of a base is proportional to its frequency in the multiple sequence alignment.
The ITS2 transcripts were comparatively short, ranging from 150 to 164 ribonucleotides with a GC content of 30.52-35.95% (Table 4). All gastrotrichs shared a loop model consisting of a 17-37 ntlong central loop and four helices (Fig. 6A). Helix I is the shortest and most conservative, displaying a motif 5'-UAU vs AUA-3' in the stem and a terminal tetraloop. Helix II is also rather conservative, carrying a motif 5'-UAACAG vs CUGUUG-3' at its base (Fig. 6C). This helix is typically composed of eight rarely ten (Ch. (H.) superbus sp. nov., Supp. file 1: Fig. S9B) or eleven (Ch. (H.) mirabilis sp. nov., Supp. file 1: Fig. S9A) pairs. In addition, there is a pyrimidine mismatch in Ch. (H.) superbus sp. nov. (Supp. file 1: Fig. S9B, opposed arrows). The terminal loop of helix II consists of six to eight ribonucleotides. Helix III is less conservative and contains 11-13 pairs, a single bulge, and 4 or 6 nucleotides in the terminal loop. Helix IV is the longest having as many as 46-72 ribonucleotides. It is the most variable helix concerning both the primary and the secondary structure. Only the terminal part of this helix is somewhat conserved, displaying a motif 5'-RAU vs AUY-3' in front of the terminal loop. Interestingly, there are no variable positions in the terminal loop of helix IV (Fig. 6A, C). ITS2 sequences were consistently identical within a species. Interspecies p-distances ranged from 5.88% to as much as 34.76%, making the ITS2 molecule an optimal barcode for species identification (Fig. 4B). Moreover, both the primary and secondary structures unambiguously define each species (Supp. file 1: Figs S9-S12) and up to four CBCs have been recognized between species (Table 3).
The amplified region of the 28S rRNA gene covers a portion of domain 0 (helix 25a), the whole domain I (helices 2-25es7), and a part of domain II (helices 27-35a). Sequences within species were identical except for Ch. (H.) slavicus sp. nov. in which the sequence identities ranged from 99.91% to 100%. Interspecies p-distances varied from 0.26% to 9.78%. Since the gap between intra-and interspecies distances is very small (i.e., the maximum intraspecies distance is 0.09% while the minimum interspecies distance is 0.26%), 28S is suitable only as a DNA pre-barcode (Fig. 4C). Despite this, each species could be unambiguously distinguished by both the primary and the secondary structure of the first two domains of the 28S rRNA molecule. Species-specific mutations tend to accumulate in helices 25es7-25es7c, which are hence taxonomically most important. Secondary structure models of the nine new gastrotrichs are shown in Figure 7 and Supp. file 1: Figs S13-S20. Diagnostic autapomorphies are marked by arrows. The distinctness of the majority of species is also strengthened by 1-18 CBCs (Table 3). and interspecies distances (more than one order of magnitude) makes COI to be a very suitable DNA barcode for the subgenus Hystricochaetonotus (Fig. 4D).

Phylogenetic analyses
The multi-gene phylogenetic analyses resulted in well-resolved trees (Figs 8-9  . This clade is collapsed in the three, highlighted in the inset phylogram, and presented in detail in Fig. 9. ML bootstrap values and posterior probabilities were mapped onto the best ML tree. ( Fig. 9). This species-rich clade was separated from other chaetonotids included in phylogenetic analyses by a very distinct and comparatively long branch (Fig. 8,arrow). This split was well recognizable also when 18S, 28S, and COI were analyzed separately (data not shown). Thus, both nuclear and mitochondrial genes strongly corroborate the long independent evolution of this diverse clade.
This clade was consistently divided into two statistically well-supported lineages.

Morphological diagnosis
Body slender and about 83-107 µm long. Head wider than neck, separated from trunk by a distinct neck constriction. Cephalion clearly demarcated, epipleurae and hypopleurae only inconspicuously marked in head outline. Trunk widest at ca U62, gradually tapers towards furca base (U85). Mouth ventral, with two cuticular teeth. Pharynx with anterior and posterior dilatations. Intestine straight, with a marked anterior section. Scales spined, three-lobed, not overlapping, distributed in 10-12 columns, 17 scales per column. Dorsal surface covered from anterior end of hypopleurae (ca U6) to neck constriction with fairly small scales bearing short spines. Neck caries broader scales with longer spines. Anterior trunk region bears similarly shaped, slightly bigger scales with posterior lobes closer together. Dorsolateral and lateral scales with elongated base and posterior lobes more divergent, spine with an inconspicuous denticle. Mid-trunk to terminal trunk region covered by (i) five horizontal rows of big, anteriorly tapered scales carrying very elongated, massive spines with a denticle and membrane and (ii) two horizontal rows of smaller, anteriorly rounded scales carrying significantly shorter spines with a denticle. Furca base short, lateral margins vaulted, furcal indentation U-shaped, adhesive tubes well-developed, diverging posteriorly. Furca base and branches covered with three-lobed, spined scales and oblong, keeled scales.

Type material
A DNA sample of the holotype specimen BZs 02 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427888).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen BZs 02 have been deposited in GenBank under the following accession numbers: OM421704, OM421680, and OM424059, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) mirabilis sp. nov. is about 83-107 µm long and has a slender body that is tenpin-shaped, with a clearly defined head region, a narrowing neck, and a slightly bulbous trunk (Figs 10A-B, G, 12A-C). Body width is 10-19 µm at U10, 11.2-11.6 µm at U50, and 17.5-18.5 µm at U60. The head is relatively wide, with a plate-like, rounded cephalion. The neck (ca U12-U27) is clearly demarcated and smoothly continues to the trunk region, i.e., a distinct neck constriction is formed. The trunk is nearly as wide as the head, gradually dilating from about U39 to U62 where it reaches the maximum width. Then it gradually tapers towards U84 where vaulted margins of the furca branches start to form. Dorsal sensory bristles (setolae) arise from the cuticle in two pairs at U25 and U75 (Figs 14A, 16C). The furcal indentation is broadly U-shaped. The furca branches are set apart and diverge posteriorly. Well-developed adhesive tubes are 9.4-10.5 µm long, they are straight and narrow (Fig. 17G).
Head. The head is roughly five-lobed. The cephalion (U1-U2) is rounded, 0.4-0.9 µm wide, clearly demarcated in the body outline, appears as a lens in the ventral view, and has a free posterior (dorsal) edge (Figs 10A-B, F, 12B, 13A-C). The epipleurae are approximately at U3-U5 and the hypopleurae are at ca U6-U9. Notches separating the epipleurae from the hypopleurae are very shallow, causing that they are only inconspicuously marked in the head outline (Fig. 13A, E). Two pairs of cephalic ciliary tufts emerge laterally between the cephalion and the epipleurae edge (ca U3) as well as between the epiand the hypopleurae edge (U5) (Figs 10A, F-G, 12A-C, 13A, E). The hypostomium is absent and the under-mouth area carries only ciliary patches. Each field is composed of several irregular groups of basal   bodies (kinetosomes). The mouth ring is oval, 5.0-5.5 µm in the largest diameter, located subterminally at U2-U4. There are strong but short, rod-like reinforcements lining the walls of the mouth ring as well as an inner pair of cuticular teeth located in the center of the mouth opening (Figs 10C, 13D).
internal morpHology. The pharynx extends from ca U5 to U30, is 26-30 µm long and 5.3-7.8 µm wide, sinuous, and has marked anterior and posterior dilatations (Figs 10F, 13A, C, white arrowheads). The posterior dilatation (ca U20-U30) is larger than the anterior one (ca U5-U8) (Supp. file 1: Table S2). The pharynx is connected to the straight intestine through a pharyngeal-intestinal junction (Fig. 13B). The intestine runs from U31 to U84 and has a separate, well-differentiated anterior section (U31-U33) (Figs 10F, 13A). There are highly refractive and regularly arranged structures well recognizable in the lateral view of the intestine (Figs 13F, 14G, double white arrowheads). Transversal bands connected to the base of dorsal scales are recognizable from ca U10 (Figs 10F, 13A). The adhesive gland (ca U85-U91) is placed right behind the terminal part of the intestine, it is broadly pyriform forming a short dichotomy at the subtle furca base (Figs 10F, 13G).
scales. Almost the entire body is covered by not overlapping three-lobed scales that adhere to the basal cuticle layer along their whole perimeter. Scales are distributed in a minimum of 10-12 longitudinal rows, with 17 scales in the central row. Central dorsal and dorsolateral longitudinal rows of scales begin at the level of the anterior edge of the hypopleurae (ca U6), lateral rows start at their posterior end (ca U9). Ventral and ventrolateral rows are hardly visible due to the highly pronounced dorsal spines (for further explanation, see below). Head scales (ca U6-U25) are fairly small, i.e., 3.3-4.1 × 1.9-3.9 µm in size. Two types could be recognized: (i) lateral boomerang-shaped scales with a subtle anterior lobe, the transition between anterior and posterior lobes is indistinct and continuous, α = 170-177°, and β = 70-85° ( Fig. 11A) and (ii) dorsal/dorsolateral scales with a distinctly elongated anterior lobe and a smaller angle α ranging from 146 to 169°, angle β spans a wide range of 68-108°, and the transition between anterior and posterior lobes is marked (Figs 11B,14B). Neck dorsal scales are 3.2-5.1 × 2.0-2.7 µm in size. They are anteriorly more broadly rounded, their angle α is slightly smaller (153-163°), β is closer to the right angle (82-91°) than in head scales, and the transition between anterior and posterior lobes is marked (Figs 11C, 14A). This type of scale terminates right at the anterior border of the trunk region (ca U26-U39). The trunk region is covered by four types of scales: (i) lateral scales (U37-U50) with a tongue-shaped anterior lobe and comparatively narrow posterior lobes, α = 149-150°, and β = 76-82°, the transition between anterior and posterior lobes is indistinct and continuous (Figs 11D, 14F); (ii) dorsolateral scales (U50-U81) with a tongue-shaped anterior lobe and wider posterior lobes being closer together, α = 147-158°, β = 68-76°, the transition between anterior and posterior lobes is indistinct (Figs 11E, 14A, C); (iii) dorsal big scales (U50-U78) with a tapered, triangular anterior lobe and narrowly rounded posterior lobes, α = 175-179°, β = 57-68°, the transition between anterior and posterior lobes is also indistinct (Figs 11F, 14A); and (iv) dorsal smaller scales (U80-U90) with an anteriorly rounded anterior lobe, the transition between anterior and posterior lobes is marked (Figs 11G, 14D).
spines. All spines bear a lateral denticle and gradually narrow towards the distal end. Three main types could be distinguished. The most common type of spines emerges from the head (2.7-3.7 µm long), neck (4.5-7.1 µm long), trunk dorsolateral (4.0-6.8 µm long), and lateral scales (6.8-8.8 µm long). The lateral denticle is minute and its tip is distant only 0.7-0.9 µm from the spine apex, which corresponds to a d-ratio of 25.3-31.2%. Lateral spines are, however, slightly thinner (0.32-0.35 µm vs 0.30-0.43 µm) and their subterminal denticle is rather inconspicuous and much closer to the spine apex (d-ratio 8-10%) in comparison with head, neck, and dorsolateral spines. The second type is represented by the prominent, elongated spines carrying a conspicuous denticle (2.6-5.6 µm long) associated with a membrane (Figs 10D, 11F, 12B-C, 13F, 14A, G). The denticle is 6.4-7.9 µm distant from the spine apex, which corresponds to a d-ratio of 18.2-22.6%. Type 2 spines are 20.7-35.9 µm long and comparatively wide (1.8-1.9 µm) at the base. Altogether only five horizontal rows of dorsal scales are equipped with this type of spines (U50-U78). The last type could be found only on the two terminal rows of dorsal scales on the furca base and branches. These spines are significantly shorter (13.0-18.4 µm vs 20.7-35.9 µm) and thinner (0.7-0.8 µm vs 1.1-1.6 µm) than the previous type. They are slightly curved and hair-like tapered distally. Although their denticle is well recognizable, it is not associated with a distinct membrane. The d-value is 2.9-4.4 µm and the d-ratio is 22.0-25.7% (Figs 10E, 11G, 14D).
Ventral ciliary bands and Ventral interciliary field. Unfortunately, the ventral side could not be observed in detail due to the very long and strong type 2 spines that precluded turning over and squeezing the worms. Despite that, the following observations could be done. The longitudinal ciliary bands begin at ca U7 and run backward to ca U87. They are somewhat wider on the neck than on the trunk where they narrow slightly from ca U78. The ciliary bands are accompanied by a ventrolateral row of small (2.4-3.0 × 1.0-1.5 µm in size), oblong, and keeled scales that start at U7. The upper furcal region (U84-U90) carries two types of scales: (i) three-lobed, spined scales with a broadly rounded anterior lobe, narrowly rounded posterior lobes, α = 137-171°, and β = 59-72° (Figs 10G, 11H) and (ii) oblong and keeled scales being 1.5-3.9 × 0.7-1.5 µm in size (Figs 10G, 11I).

Morphological diagnosis
Body stocky and about 107 µm long. Head wider than neck, separated from trunk by an inconspicuous neck constriction. Cephalion, epipleurae, and hypopleurae clearly demarcated. Trunk gradually widens from ca U37 to U60 and then gradually tapers towards furca base at U80. Mouth ventral, with delicate protruding structures, no cuticular teeth. Pharynx without dilatations. Intestine straight, with a marked anterior section. Scales spined, three-lobed, not overlapping, distributed in 10-12 columns, 15 scales per column. Spines with a short lateral denticle. Scales and spines increase gradually in size in a posterior direction. Dorsal surface covered by scales from posterior end of cephalion (ca U3) to furca base (ca U93). Furca base short, lateral margins of furca branches more or less straight, furcal indentation deeply U-shaped, adhesive tubes well-developed, almost parallel. Furca base and branches covered with oval, keeled scales. Reference molecules are shown in Supp. file 1: Figs S1, S9B, S13. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.

Molecular diagnosis
The p-distance from species described in the present study is 0.

Type material
A DNA sample of the holotype specimen ZPvs 55 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427126).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen ZPvs 55 have been deposited in GenBank under the following accession numbers: OM421708, OM421684, and OM424063, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) superbus sp. nov. is about 107 µm long and has a stocky body that is tenpin-shaped, with a clearly defined head region, a narrowing neck, and a rather bulbous trunk (Figs 15A, H, 16A). Body height in lateral view is 14.0-15.0 µm at U10, 15.0-15.4 µm at U50, and 16.1-17.0 µm at U60. The head is relatively wide, with a plate-like, rounded cephalion. The neck (ca U12-U27) smoothly continues to the trunk, which is significantly wider than the head, gradually dilates from about U37 to U60 where it reaches the maximum width. Then it gradually tapers towards U80 where curved margins of the furca branches begin to emerge. Dorsal sensory bristles arise from the cuticle in two pairs at U12 and U79 (Fig. 19A). The furcal indentation is deeply U-shaped. The furca branches are set apart and diverge posteriorly. Well-developed adhesive tubes are approximately 9 µm long, more or less straight, and run almost in parallel (Figs 15A, H, J, 16A).
Head. The head is five-lobed. The cephalion (U1-U2) is rounded, clearly demarcated in the body outline, appears as a lens in the ventral view, and has a free posterior (dorsal) edge (Figs 15A, H, J-K, 16B, 17C). The epipleurae are formed approximately at U3-U5 while the hypopleurae at ca U6-U13. The latter structures are well recognizable in the head outline (Figs 15A-B, 16A, black stars). Two pairs   of cephalic ciliary tufts emerge laterally between the cephalion and the epipleurae edge (ca U3) as well as between the epi-and the hypopleurae edge (U6) (Figs 15A, H, 16A). The hypostomium (ca U4-U6) is free of structures, lined only with ciliary patches. The mouth ring is oval, 4.4-4.9 µm in the largest diameter, located subterminally at U3-U6. There are strong, rod-like reinforcements lining the mouth walls and delicate structures protruding from the mouth ring (Figs 15H, K, 17C). Inner cuticular teeth are not present.
internal morpHology. The pharynx extends from ca U5 to U34, is 28-32 µm long and 5.9-8.3 µm wide, sinuous, and has no dilatations (Figs 15J, 16B, 17C). The cerebral ganglion appears as a mass surrounding the pharynx along its whole length (Fig. 15K). The salivary glands are recognizable as small hyaline balls situated at the dorsal side of the pharynx around at U33 (Figs 15J, 16B, 17A). The intestine runs from U35 to U82 and has a separate, well-differentiated anterior section (U35-U38) (Figs 15J, 16B). Transversal bands connected to the base of dorsal scales are well recognizable under DIC and secondary magnification of 2500 × (Figs 15J, 16B, 17A-B). The adhesive gland (ca U85-U91) is placed right behind the terminal part of the intestine, forming a short dichotomy at the subtle furca base.
scales. Almost the entire body is covered by not overlapping three-lobed scales that adhere to the basal cuticle layer along their whole perimeter. Scales are distributed in a minimum of 12 longitudinal rows, with 10 scales in the central row. Their size increases gradually in a posterior direction. Central dorsal and dorsolateral longitudinal rows of scales begin at the level of the anterior edge of the epipleurae (ca U5), while lateral rows start at the posterior end of the hypopleurae (ca U13). Ventral rows are hardly visible due to the thick, elongated dorsal spines (for further explanation, see below) and long locomotory cilia. Three main types of scales could be recognized with respect to the shape of the anterior lobe of scales: (i) head and upper-neck scales (U5-U28) with a small, rounded anterior lobe and elongated posterior lobes, α = 160-170°, and β = 89-98° ( spines. All spines bear a distinct lateral denticle and gradually narrow towards their distal end. Keels start near the anterior margin of scales. Spines are not straight but distinctly curved (Figs 15A, G-J, 16A-B, 17D-G). They do not differentiate into various types, only their length changes in a posterior direction (Supp. file 1: Table S3). The most pronounced change occurs at the beginning of the trunk (Figs 15H, 16A), where dorsal spines increase significantly from 6.7-10.7 µm to 10.8-13.0 µm. The lateral denticle is comparatively distant from the spine apex, i.e., d-value ranges from 1.7-5.1 μm, which corresponds to a d ratio of 21.4-34.7%.
Ventral ciliary bands and Ventral interciliary field. Unfortunately, the ventral side could not be observed in detail due to the thickness of the dorsal spines that precluded turning over and squeezing the worms. Moreover, the ventral locomotory cilia also hampered detailed observations of the ventral side. Despite these problems, the following observations could be done. The longitudinal ciliary bands begin at ca U10 and run backward to ca U85. The ciliary bands are accompanied by two ventrolateral rows of small (2.0-2.9 × 1.9-2.2 µm in size), three-lobed scales that start at U13. The interciliary field bears small (1.7-2.6 × 0.9-1.2 µm in size), oval scales without spines or keels (Fig. 15E). The upper furcal region (U84-U90) carries two types of scales: (i) lateral three-lobed, spined scales with a broadly rounded anterior lobe, narrowly rounded posterior lobes, α = ~140°, and β = ~73° and (ii) oblong and keeled scales being 0.84-1.2 × 1.47-1.72 µm in size (Fig. 15F).

Morphological diagnosis
Body elongated and about 108-143 µm long. Head slightly wider than neck, separated from trunk by an inconspicuous neck constriction. Cephalion clearly demarcated, epipleurae and hypopleurae not recognizable. Trunk comparatively narrow, not broader than head, widest at ca U63, narrowest at ca U88. Mouth ventral, no cuticular teeth. Hypostomium bears a rectangular cuticular structure accompanied by two posterior lateral lamellae. Pharynx with anterior and posterior dilatations. Intestine straight, with a marked anterior section. Scales slightly overlapping, distributed in 16-18 columns, 40-45 scales per column. Dorsal surface covered from posterior end of cephalion (ca U1) to furca base (U94) with (i) spined, three-lobed scales with rounded anterior end and (ii) spined, three-lobed scales with truncated anterior end. Spines short, narrowed posteriorly, without denticles. Ventral side carries two broad ciliary bands and two pairs of ciliary fields around mouth and hypostomium. Interciliary filed bears (i) V-shaped scales without keels, (ii) bowl-shaped scales without keels, (iii) obtriangular scales with very minute keels, and (iv) narrowly obovate, keeled scales. Furca base longer than adhesive tubes, lateral margins more or less straight, furcal indentation deeply V-shaped, adhesive tubes diverging posteriorly. Dorsal side of furca base and branches covered with three-lobed scales, ventral side with oblong, spined scales and oval to broadly fusiform, keeled scales.  Reference molecules are shown in Supp. file 1: Figs S2 and S14. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.

Etymology
The

Type material
A DNA sample of the holotype specimen DB 34 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427607).

Gene sequences
The nuclear 18S and 28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen DB 34 have been deposited in GenBank under the following accession numbers: OM421710, OM421686, and OM424065, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) optabilis sp. nov. is about 108-143 µm long and has a slender elongated body, with a head region slightly broader than the inconspicuous neck and a more or less bulbous trunk (Figs 18A,19A,E,20A). Body width is 18.8-21.7 µm at U10, 22.5-23.7 µm at U50, and 26.1-30.7 µm at U60. The head is relatively wide (11-15 µm at U6), with a plate-like, rounded cephalion. Epi-and hypopleurae are not recognizable. The neck (ca U12-U25) is only inconspicuously marked and smoothly continues to the trunk region. The trunk is about as narrow as the head, gradually dilating from approximately U38 to U63, where it reaches the maximum width. Then it gradually tapers towards U88, where the curved margins of the furca branches start to form. Dorsal sensory bristles were not observed. The furcal indentation is deeply V-shaped. The furca branches are set apart and diverge posteriorly. Well-developed adhesive tubes are approximately 9 µm long, they are straight and almost in parallel (Figs 18A, 19A, D-E, 21D, 22E).
Head. The cephalion (U1) is rounded, clearly demarcated in the body outline, appears as a lens in the ventral view (Figs 19A,E,21A). Pairs of cephalic ciliary tufts emerge laterally at U4 and U6 (Figs 19A,E,21A). The mouth ring is oval, 5.2-6.2 µm in the largest diameter, and is located subterminally at U2-U5. There are strong but short, rod-like reinforcements lining the walls of the mouth ring and inner delicate structures directed towards the center of the mouth ring. Inner cuticular teeth are not present (Figs 20A,21B). The hypostomium (ca U5-U9) has a complex morphology, i.e., bears a rectangular     plate (5.0 × 2.9 µm) with a small central protuberance and two lamellae situated posterior to the plate. The lateral sides of the hypostomium are lined by a relatively wide pair of sensoric ciliary patches from U3 to U8 (Figs 19A,20A,21B).
internal morpHology. The pharynx extends from ca U5 to U29, is 28.0-33.4 µm long and 7.3-7.8 µm wide, sinuous, and has marked anterior and posterior dilatations (Fig. 19E). The posterior dilatation (ca U20-U25) is larger than the anterior one (ca U7-U9) (Supp. file 1: Table S4). The intestine runs from U29 to U85 and has a separate, well-differentiated anterior section (U29-U33). Transversal bands connected to the base of dorsal scales are well recognizable (Fig. 19E). The adhesive gland (ca U85-U87) is placed right behind the terminal part of the intestine forming a short dichotomy at the subtle furca base. The organ X has a lemniscate shape and is situated ahead of adhesive glands (Figs 19E,21D).
scales. Almost the entire body is covered by slightly overlapping three-lobed scales that adhere to the basal cuticle layer along either all or most of their perimeter. Scales are distributed in 16-18 longitudinal rows, with 40-45 scales in the central row. Their size increases slightly in a posterior direction. Central dorsal and dorsolateral longitudinal rows of scales begin at the level of the anterior edge of the cephalion (ca U1), while lateral rows start at ca U6. They run almost along the whole body length (till U94). Ventrolateral rows commence at U23 due to the highly developed and wide anterior part of ciliary rows and terminate in the furcal region (at ca U91). Ventral rows start at U10, are staggered, and exhibit a horizontal zonation pattern. Given the shape of the anterior scale lobe, two main types were recognized: (i) head and neck scales (U1-U30), with a rounded anterior lobe and comparatively short posterior lobes, α = 180-196°, β = 47-68° (Figs 18B, D, 20B, 22C) and (ii) trunk and dorsal furca scales (U31-U93), with a truncated anterior lobe and longer posterior lobes, α = 162-188°, β = 58-85° (Figs 18C, E, 22B). Both types share an indistinct and continuous transition between the anterior and posterior lobes. Rounded structures (1.0-2.4 µm across) irregularly interspersed within type 1 and 2 scales were observed along the whole dorsal side (Figs 18A-C, 20B-C). As they were clearly recognizable only in a single specimen, we cannot exclude that these structures are artifacts. The V-shaped furcal indentation carries the fourth type of dorsal scales that are three-lobed and have a tongue-shaped anterior lobe, relatively narrow posterior lobes, α = 156-176°, β = 56-80°, and a marked transition between the anterior and posterior lobes (Fig. 18G).
spines. Dorsal, dorsolateral, lateral, and ventrolateral scales bear simple, slightly curved spines that gradually narrow towards the distal end (Figs 18D-G, 20A). The spine emerges rather close to the anterior margin of scales. Spines do not differentiate into various types, only their length slightly increases in a posterior direction (Supp. file 1: Table S4). A lateral denticle is not formed. Ventral scales are spineless.

Morphological diagnosis
Body elongated and about 130 µm long. Head slightly wider than neck, separated from trunk by an inconspicuous neck constriction. Cephalion, epipleurae, and hypopleurae clearly demarcated. Trunk comparatively narrow, not broader than head, widest at ca U62, narrowest at ca U87. Mouth ventral, three cuticular teeth. Hypostomium bears a pentagonal cuticular structure with two protuberances. Pharynx without dilatations. Intestine straight, with a marked anterior section. Scales slightly overlapping, distributed in 14 columns, 30-32 scales per column. Dorsal surface covered with: (i) head scales small, with a subtle, well-delimited anterior lobe and comparatively long and diverging posterior lobes; (ii) neck scales boomerang-like, with a broadly rounded anterior lobe, elongated posterior lobes; (iii) dorsolateral neck scales with a well-delimited, broadly rounded anterior lobe and distinct transition between anterior and posterior lobes; (iv) main trunk scales with a rounded anterior end, posterior lobes comparatively long, narrowly rounded distally, and diverging; (v) posterior trunk scales with a conspicuously prolonged, tongue-shaped anterior lobe and relatively short posterior lobes tapering distally; and (vi) furca base scales with a narrower and more elongated anterior lobe and short posterior lobes very narrowly rounded or almost acute distally. Spines slightly increasing in length from head to posterior trunk region, gradually tapering to become hair-like terminally, without denticles. Ventrolateral spines shorter than dorsal ones. Ventral side carries two broad, anteriorly connected (U9-U27) ciliary bands. Interciliary field covered by (i) minute, tongue-like, keeled scales; (ii) small, roughly rectangular, keeled scales; (iii) oblong scales with a short spine; and (iv) a pair of big, elongated oval, keeled scales. Furca branches slightly longer than adhesive tubes, lateral margins more or less straight, furcal indentation deeply V-shaped, adhesive tubes short. Dorsal side of furca branches covered with elongated oval, keeled scales and ventral side of furca branches covered with (i) two pairs of three-lobed, spined scales, having an elongated, tongue-shaped anterior lobe and comparatively short, narrowly rounded to acute posterior lobes and (ii) cordiform, short-spined scales.

Type material
A DNA sample of the holotype specimen VP 32 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427566).

Gene sequences
The nuclear 18S and 28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen VP 32 have been deposited in GenBank under the following accession numbers: OM421713, OM421689, and OM424068, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) avarus sp. nov. is about 130 µm long and has a slender elongated body, with a head region slightly broader than the neck and trunk (Figs 23A, K, 24A, 25A-C). Body width is 22.0-23.0 µm at U10, 18.0-19.0 µm at U50, and 20.0-21.0 µm at U60. The head is relatively wide (11-15 µm at U6), with a plate-like cephalion. Epi-and hypopleurae are clearly demarcated in the head outline (Figs 23A, K, 24A, 25A-C, 26A). The neck (ca U14-U27) is only inconspicuously marked and smoothly continues to the trunk. In comparison with the head, the trunk is comparatively slender, gradually dilatating from about U35 to U55, where it reaches the maximum width. Then, the trunk gradually narrows towards U87, where more or less straight margins of the furca branches start to form. Dorsal sensory bristles arise from the cuticle in two pairs at U28 and U78 (Fig. 23A, L). The furcal indentation is deeply V-shaped. Well-developed adhesive tubes are approximately 6-12 µm long, they are straight and short (Figs 23A, K, 24A, H, 27E).
Head. The cephalion (U1) is clearly demarcated in the body outline, surrounds the mouth ventrally like a bib, and is distinctly flattened and lenticular in the dorsal view (Figs 23A, K, 24B, 25A-C, 26A, C). Two pairs of cephalic ciliary tufts emerge at U3 and U7. In addition, inverted V-shaped streaks of basal bodies extend from the level of the mouth to the posterior level of the hypostomium (Figs 24A, 26C). The mouth ring is oval, approximately 5.5 µm in the largest diameter, and located subapically at U2-U5. There are strong but short, rod-like reinforcements lining the walls of the mouth ring and inner delicate structures directed towards the center of the mouth ring (Fig. 26C). Three inner cuticular teeth are clearly visible, two are located laterally and one apically (Fig. 25A). The hypostomium (ca U5-U9) is rich in structures, i.e., it is composed of a pentagonal cuticular plate bearing two protuberances (Figs 24B, 26C).  internal morpHology. The pharynx extends from ca U5 to U29, is almost 35 µm long and 7.4-9.1 µm wide, sinuous, and without dilatations (Figs 25A, 26A). The intestine runs from U29 to U85 and has a separate, well-differentiated anterior section (U29-U33). Transversal bands connected to the base of dorsal scales are well recognizable. The adhesive gland is placed right behind the terminal part of the intestine (ca U85-U87), forming a short dichotomy at the subtle furca base (Figs 23K-L, 25C).
scales. Almost the entire body is covered by slightly overlapping, mostly three-lobed scales that adhere to the basal cuticle layer along either all or most of their perimeter. Scales are distributed in 14 longitudinal rows, with 30-32 scales in the central row. Their size increases slightly in a posterior direction. Central dorsal and dorsolateral longitudinal rows of scales begin at the level of the anterior edge of the cephalion (ca U1), while lateral rows start at ca U14 at the posterior edge of the hypopleurae. They run almost along the whole body length (till U92) (Figs 23A, 27B). Ventrolateral rows commence at U13 due to the highly developed hypopleurae and terminate in the furcal region at ca U85. Ventral rows commence rather far away from the anterior body end at U29, because of the strongly developed and apically completely fused bands of locomotory cilia (Figs 24A, 25C). Seven main types of dorsal scales were recognized.  23F). (iii) Dorsolateral neck scales possess a well-delimited, broadly rounded anterior lobe, i.e., the transition between anterior and posterior lobes is distinct, α = ~171°, β = ~70° (Fig. 23D). (iv) The main trunk region bears distinctly bigger scales (5.3-9.3 × 3.1-6.1 µm) whose anterior end is rounded, posterior lobes are comparatively long, narrowly rounded, and diverging, the transition between anterior and posterior lobes is well marked, α = 146-176°, β = 86-102° (Fig. 23G).
spines. Dorsal, dorsolateral, lateral, and ventrolateral scales bear simple, slightly curved spines that gradually narrow towards the distal end (Figs 23A, E, K-L, 25A-B). The spine emerges rather close to the anterior margin of scales. Spines do not differentiate into various types, only their length slightly increases from 3.5 µm to 10.9 µm in a posterior direction (Supp. file 1: Table S4). A lateral denticle is not developed.
Ventral ciliary bands and Ventral interciliary field. Ventral ciliary bands commence almost right behind the hypostomium. Their anterior region is conspicuously broad, causing both bands to be completely fused from U9 to U27 (Figs 24A, 26B-C). However, the middle distal part of the merged ciliary bands more or less follows the course of the posterior margin of the hypostomium, i.e., it is concave from ca U10 to U13. Basal bodies are densely packed and do not form regular rows in the anterior fused region of the ciliary bands. The anterior field of basal bodies splits into two distinct ventral bands about at U27. Locomotory cilia then become regularly arranged in more or less equidistantly spaced horizontal rows. Each row typically consists of seven narrowly spaced basal bodies from ca U28 to U83. Rows of basal bodies are lined up with horizontal rows of ventral scales. This regular pattern changes about at U84 where distances between horizontal ciliary rows gradually decrease. Also, the number of basal bodies per row gradually decreases and ciliary bands terminate at U88 (Figs 24A, 25C). The ciliary bands are accompanied by two ventrolateral rows of three-lobed scales, with an indistinct and continuous  transition between the anterior and posterior lobes (Fig. 24A). The inner row starts at U22 and consists of smaller scales (2.4-4.3 × 1.2-2.8 µm) (Fig. 24E), while the outer row begins at ca U40 and is built up from bigger scales (4.8-5.3 × 2.2-3.6 µm) (Fig. 24D). The ventral interciliary field bears four types of horizontally distributed scales: (i) minute (1.5-1.8 × 1.3-1.8 µm), tongue-like, double-edged scales with a minute keel (U28-U32); (ii) small (1.9-2.3 × 1.6-2.1 µm), roughly rectangular, double-edged scales with a keel (U33-U83) (Figs 24C, 27D); (iii) slightly bigger (3.9-4.4 × 1.0-1.5 µm), oblong, doubleedged scales with a short spine (U83-U87); and (iv) a pair of big (6.1-7.8 × 2.0-3.2 µm), elongated oval scales with a keel (Figs 24H, 27E). All ventral interciliary field scales are anteriorly merged into the cuticle. The furca branches carry two types of scales (U88-U95): (i) two pairs of three-lobed, spined scales (3.2-5.0 × 1.1-2.0 µm), with an elongated, tongue-shaped anterior lobe and comparatively short, narrowly rounded to acute posterior lobes (Fig. 24F)   Reference molecules are shown in Supp. file 1: Figs S4, S10, S16. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.
The p-distance from species described in the present study is 0.

Etymology
The Latin noun 'luxus' [m] ('extravagance, luxury') refers to the 'extravagant' appearance of the new species. The species group name is treated as a noun in the nominative singular standing in apposition to the generic name (Article 11.9.1.2 of the ICZN 1999).
Photomicrographs of paratype specimens are available at the Department of Zoology, Comenius University in Bratislava at https://fns.uniba.sk/en/gastrotricha/. Paratypes are shown in Fig. 30.

Type material
A DNA sample of the holotype specimen ZPvs 20 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427593).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen ZPvs 20 have been deposited in GenBank under the following accession numbers: OM421714, OM421690, and OM424069, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) luxus sp. nov. is about 87-112 µm long and has a stocky body that is more or less tenpin-shaped, with a clearly defined head region, a narrowing neck, and a rather bulbous trunk (Figs 28A, 29A). Body height in lateral view is 13.5-14.0 µm at U10, 16.5-16.8 µm at U50, and 18.0-18.2 µm at U60. The head is relatively wide, with a plate-like cephalion. The neck (ca U15-U28) smoothly continues to the trunk, which is significantly wider than the head, gradually dilates from about U37 to U60 where it reaches the maximum width. Then it gradually tapers towards U81, where curved margins of the furca branches begin to emerge. Dorsal sensory bristles were not observed. The furcal indentation is deeply U-shaped. The furca branches are set apart. Well-developed adhesive tubes are approximately 11-13 µm long, slightly curved in lateral view, and run almost in parallel (Figs 28A, H, 29A, D-E, 30A).
Head. The head is five-lobed. The cephalion (U1-U2) is rounded, clearly demarcated in the body outline, and has a free posterior (dorsal) edge. The epipleurae are formed approximately at U3-U5 while the hypopleurae at U6-U13. The latter structures are clearly demarcated in the head outline (Fig. 28A,  H). Two pairs of cephalic ciliary tufts emerge laterally between the cephalion and the epipleurae edge (ca U3) as well as between the epi-and the hypopleurae edge (U6). The hypostomium (ca U4-U6) is free of structures. The mouth ring is oval, 3.2-6.0 µm in the largest diameter, located subterminally at U2-U6. There are strong but short rod-like reinforcements lining the walls of the mouth ring and inner   internal morpHology. The pharynx extends from ca U5 to U34, is 28-32 µm long and 5.9-8.3 µm wide, sinuous, and has no dilatations (Figs 29D, 30A-B). The intestine runs from U35 to U82 and has a separate, well-differentiated anterior section (U35-U38). A pair of protonephridia runs from ca U45 to U60. Transversal bands connected to the base of dorsal scales are well recognizable (Figs 29D, 30A). The adhesive gland (ca U80-U91) is placed right behind the terminal part of the intestine, forming a short dichotomy at the subtle furca base.
scales. Almost the entire body is covered by not overlapping three-lobed scales that adhere to the basal cuticle layer along their whole perimeter. Scales are distributed in a minimum of 10 longitudinal rows, with 15 scales in the central row. Their size increases gradually in a posterior direction. Central dorsal and dorsolateral longitudinal rows of scales begin at the level of the anterior edge of the epipleurae (ca U5), while lateral rows start at the posterior end of the hypopleurae (ca U13). Ventral rows are hardly visible due to the thick, elongated dorsal spines (for further explanation, see below) and long locomotory cilia. Five main types of scales could be recognized with respect to the shape of the anterior lobe. spines. Spines do not differentiate into various types, only their length increases from 3.4 µm to 13.5 µm in a posterior direction (Supp. file 1: Table S6). However, the width of individual spines decreases from 0.7 µm at the base to 0.2 µm at the tip and, hence, spines do not become hair-like terminally. The spine base is situated near the anterior margin of scales. All spines are distinctly curved and bear an inconspicuous lateral denticle (Figs 28A, E, H, 29A, D, 30D-E). The lateral denticle emerges comparatively near the spine apex, i.e., the d-value ranges only from 0.9-1.5 μm, which corresponds to a d ratio of 5.8-7.4%.

Morphological diagnosis
Body elongated and about 124 µm long. Head slightly wider than neck, separated from trunk by an inconspicuous neck constriction. Cephalion, epipleurae, and hypopleurae clearly demarcated. Trunk widest at ca U61, gradually tapers towards furca base (U82). Mouth ventral, one central cuticular tooth. Hypostomium bears two parallel, horizontally arranged lamellae accompanied by tear-shaped protuberances. Pharynx without dilatations. Intestine straight, with a marked anterior section. Scales spined, three-lobed, slightly overlapping, distributed in about 12 columns, 22 scales per column. Scales and spines increase gradually in size in a posterior direction. Dorsal surface covered by scales from posterior end of cephalion (ca U4) to furca branches (ca U83). Furca branches slightly shorter than adhesive tubes, lateral margins more or less straight, furcal indentation V-shaped, adhesive tubes comparatively short. Reference molecules are shown in Supp. file 1: Figs S5, S11A, S17. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.

Molecular diagnosis
The p-distance from species described in the present study is 0.  Fig. 32.

Type material
A DNA sample of the holotype specimen STV 65 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427609).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen STV 65 have been deposited in GenBank under the following accession numbers: OM421720, OM421696, and OM424075, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) iratus sp. nov. is about 124 µm long and has a slender elongated body, with a head region slightly broader than the neck and trunk (Figs 31A, J, 32A). Body width is about 21.5 µm at U10, 20.0 µm at U50, and 22.5 µm at U60. The head is relatively wide (20.2 µm at U6), with a plate-like cephalion. Epi-and hypopleurae are clearly demarcated in the head outline (Figs 31A, J, 32A). The neck (ca U13-U34) is only inconspicuously marked and smoothly continues to the trunk. In comparison with the head, the trunk is comparatively slender, gradually dilatating from about U35 to U61, where it reaches the maximum width. Then, the trunk gradually narrows towards U82, where more or less straight margins of the furca branches start to form. Dorsal sensory bristles were not observed. The furcal indentation is deeply V-shaped and approximately 19.4 µm long. Welldeveloped adhesive tubes are straight and approximately 10.7 µm long (Fig. 31A, J).
Head. The cephalion (U1) is clearly demarcated in the body outline, distinctly flattened, and surrounds the mouth ventrally like a bib. The epipleurae are formed approximately at U3-U7 while the hypopleurae at U8-U14. The latter structures are well recognizable in the head outline (Figs 31A, J, 32A). Two pairs of cephalic ciliary tufts (6.9-18.8 µm) emerge laterally between the cephalion and the epipleurae edge (ca U3) as well as between the epi-and the hypopleurae edge (U7). The mouth ring is oval, approximately 6.9 µm in the largest diameter, and located subapically at U2-U5. There are strong but short, rod-like reinforcements lining the walls of the mouth ring and inner delicate structures directed towards the center of the mouth ring (Figs 31I-J, 32C, F). One cuticular tooth is clearly visible in the center of the mouth ring (Figs 31I, 32F). The hypostomium (ca U5-U9) is composed of two more or less parallel, horizontally arranged lamellae whose lateral sides are accompanied by tear-shaped protuberances. Moreover, the lateral sides of the hypostomium are lined from U3 to U8 by relatively wide patches of irregularly arranged basal bodies (Figs 31I, 32F). internal morpHology. The pharynx extends from ca U5 to U31, is 28 µm long and 6.2-8.8 µm wide, sinuous, and without dilatations (Figs 31J, 32A-B). The intestine runs from U22 to U87 and has a separate, well-differentiated anterior section (U32-U35). Transversal bands connected to the base of dorsal scales are well recognizable. The adhesive gland is placed right behind the terminal part of the intestine (ca U85-U87), forming a short dichotomy at the subtle furca base. scales and spines. Almost the entire body is covered by slightly overlapping, mostly three-lobed scales that adhere to the basal cuticle layer along either all or most of their perimeter. Scales are very densely packed, forming a minimum of 12 longitudinal rows on the dorsal side, with 22 scales in the central row. They have a rounded anterior lobe and elongated posterior lobes narrowly rounded distally. The transition between the anterior and posterior lobes is marked except for the ventro-and dorsolateral scales in which the transition is indistinct (Fig. 31D-G). The size of scales increases from 2.3 × 2.1 µm to 7.4 × 3.1 µm in a posterior direction. Dorsal furca branches scales are 4.3 × 3.4 µm in size, threelobed and spined but their anterior lobe is more elongated, their posterior lobes are slightly shorter and narrower, the transition between the anterior and posterior lobes is continuous and indistinct (Fig. 31H).  Spines do not differentiate into various types, only their length slightly increases from 2.3 µm to 6.5 µm in a posterior direction (Figs 31A-H, J, 32A, D, Supp. file 1: Table S7). Spines are slightly narrower posteriorly but do not become hair-like terminally. A lateral denticle is not developed (Figs 31B-C, 32A, D).
Ventral ciliary bands and Ventral interciliary field. Ventral ciliary bands commence almost right behind the hypostomium (U10) and terminate at ca U87. Their anterior region is broadened as typical of most species described herein. The ciliary bands are accompanied by two ventrolateral rows of threelobed scales that start at U13. They are 3.6-5.3 × 2.5-3.1 µm in size and have a similar morphology as the dorsal and dorsolateral scales but the transition between the anterior and posterior lobes is continuous and hence indistinct (Fig. 31G). Unfortunately, no further features of the ventral side were observed. Hypostomium bears two parallel cuticular lamellae. Pharynx without dilatations. Intestine straight, with a marked anterior section. Scales spined, three-lobed, not overlapping, distributed in 14 columns, 12 scales per column. Scales and spines increase gradually in size in a posterior direction. Dorsal surface covered with: (i) small head and anterior neck scales with a small, broadly rounded anterior lobe, equally long posterior lobes, and marked transition between anterior and posterior lobes; (ii) small posterior neck scales with a slightly elongated anterior lobe and an indistinct transition between anterior and posterior lobes; and (iii) comparatively big and triangular trunk scales. Interciliary field covered by (i) small, oval to oblong scales without keels; (ii) a pair of small, oval scales with a keel; (iii) a single distinctly bigger, oblong, and keeled scale; and (iv) a pair of big, very narrowly ovate, and spined scales. Furca branches about as long as adhesive tubes, with lateral margins more or less straight, furcal indentation shallowly U-shaped, adhesive tubes well-developed. Ventral side of furca branches covered with minute, oval to oblong, and keeled scales. Reference molecules are shown in Supp. file 1: Figs S6, S11B, S18. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.

Molecular diagnosis
The p-distance from species described in the present study is 0.

Type material
A DNA sample of the holotype specimen VP 18 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427574).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen VP 18 have been deposited in GenBank under the following accession numbers: OM421721, OM421697, and OM424076, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) gulosus sp. nov. is about 107 µm long and has a stocky body that is tenpin-shaped, with a clearly defined head region, a narrowing neck, and a rather bulbous trunk ( Fig. 33A-B). Body width is ca 18 µm at U10, ca 26 µm at U50, and ca 29 µm at U60. The head is relatively wide, with a plate-like, slightly narrower cephalion. The neck (ca U13-U27) smoothly continues to the trunk, which is distinctly wider than the head, gradually dilates from about U37 to U50 where it reaches the maximum width. Then it gradually tapers towards U80 where the curved margins of the furca branches begin to emerge. Dorsal sensory bristles arise from the cuticle in two pairs at U23 and U60 (Fig. 33A). The furcal indentation is deeply U-shaped. The furca branches are set apart and diverge posteriorly. Well-developed adhesive tubes are approximately 9 µm long and more or less straight (Figs 33A-B, 35C, 36C).
Head. The head is roughly five-lobed. The cephalion (U1-U2) is rounded, clearly demarcated in the body outline (Figs 33A-B, 34A). The epipleurae are formed approximately at U3-U6, the hypopleurae are only inconspicuously marked. Two pairs of cephalic ciliary tufts emerge laterally between the cephalion and the anterior edge of the epipleurae (ca U3) as well as close to the posterior edge of the epipleurae (U6). The mouth ring is oval, ca 4.6 µm in the largest diameter, located subterminally at U2-U4. There are strong, rod-like reinforcements lining the walls of the mouth ring and inner delicate structures directed towards the center of the mouth ring (Figs 33B, 35B). Inner cuticular teeth are not present. The hypostomium (ca U6-U9) is in a form of two parallel horizontal cuticular lamellae laterally lined with a few ciliary patches (Fig. 33B). Each field is composed of several irregular groups of basal bodies (Figs 33B, 35B). internal morpHology. The pharynx extends from ca U5 to U27, is about 28 µm long and 6.5-8.9 µm wide, sinuous, and has no dilatations. The cerebral ganglion appears as a mass surrounding the pharynx along its   whole length. The intestine runs from U27 to U81 and has a separate, well-differentiated anterior section (U28-U31). A pair of protonephridia runs from ca U32 to U40 (Fig. 35C). Transversal bands connected to the base of dorsal scales are well recognizable. The adhesive gland (ca U85-U91) is placed right behind the terminal part of the intestine, forming a short dichotomy at the subtle furca base (Fig. 35C).
scales. Almost the entire body is covered by not overlapping three-lobed scales that adhere to the basal cuticle layer along their whole perimeter. Scales are distributed in 14 longitudinal rows, with 13 scales in the central dorsal row. Their size increases in a posterior direction. Central dorsal and dorsolateral longitudinal rows of scales begin at ca U5 (behind the posterior edge of the cephalion), while lateral rows start at the posterior end of the epipleurae (ca U13). Ventral rows start at U10 and exhibit a horizontal zonation pattern. Four main types of scales could be recognized with respect to the shape of the anterior lobe. (i) The head and anterior neck scales (U5-U28) are 3.1-3.5 × 2.0-2.3 µm in size, have a small, broadly rounded anterior lobe and equally long posterior lobes, the transition between the anterior and posterior lobes is marked, α = ~156°, and β = ~123° (Fig. 34A) (Fig. 34D). (iv) The trunk lateral scales are 5.5-5.9 × 3.7-3.8 µm in size, distinctly three-lobed, all lobes are narrowly rounded to more or less tapered, and the transition between the anterior and posterior lobes is marked (Fig. 34E).
spines. All spines bear a distinct lateral denticle and gradually narrow towards their distal end (Figs 33A-B, 34A-F). Keels start close to the anterior margin of scales. Spines are not straight but distinctly curved (Fig. 35C-D). They do not differentiate into various types, only their length increases from 3.1 μm to 14.0 μm in a posterior direction (Supp. file 1: Table S8). The lateral denticle is comparatively distant from the spine apex, i.e., the d-value ranges from 1.0-3.1 μm, which corresponds to a d ratio of 7.5-25%.

Molecular diagnosis
The p-distance from species described in the present study is 0.  Fig. 38.

Type material
A DNA sample of the holotype specimen STV 67 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427571).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen STV 67 have been deposited in GenBank under the following accession numbers: OM421723, OM421699, and OM424078, respectively.

Description
Habitus. Chaetonotus (Hystricochaetonotus) arcanus sp. nov. is about 100 µm long and has a stocky, tenpin-shaped body, with a clearly defined head region, a narrowing neck, and a rather bulbous trunk  ( Figs 37A, H, 38B). Body width is ca 19.8 µm at U10, ca 15.6 µm at U50, and ca 17.8 µm at U60. The head is relatively wide (19.8 µm at U10), with a plate-like, rounded cephalion. The neck (ca U17-U27) is rather inconspicuously marked and smoothly continues to the trunk. The trunk gradually dilatates from about U35 to U61, where it reaches the maximum width that is only slightly narrower than the maximum width of the head. Then, the trunk gradually narrows towards U82, where curved margins of the furca branches begin to emerge. Dorsal sensory bristles were not observed. The furcal indentation is deeply U-shaped. The furca branches are set apart and diverge posteriorly. Well-developed adhesive tubes are approximately 9.4 µm long and straight (Figs 37A, H, 38D).
Head. The cephalion (U1) is rounded, clearly demarcated in the body outline (Fig. 37A). The epipleurae (U3-U6) and hypopleurae (U6-U10) are only inconspicuously marked. Pairs of cephalic ciliary tufts emerge laterally at U3 and U5. The mouth ring is oval, ca 5.2 µm in the largest diameter, and located subterminally at U2-U5. There are strong but short, rod-like reinforcements lining the walls of the mouth ring and inner delicate structures directed towards the center of the mouth ring. Inner cuticular teeth are not present. The hypostomium (ca U5-U9) carries a small cuticular boomerang-like structure. The lateral sides of the hypostomium are lined by a relatively wide pair of basal body patches (from U3 to U8) (Figs 37G, 38E).
internal morpHology. The pharynx extends from ca U5 to U34, is 28-29 µm long and 4.6-7.2 µm wide, sinuous, with inconspicuously marked anterior and posterior dilatations (Figs 37H, 38B). The cerebral ganglion appears as a mass surrounding the pharynx along its whole length. The intestine runs from U35 to U82. The adhesive gland (ca U82-U91) is placed right behind the terminal part of the intestine, forming a short dichotomy at the subtle furca base.
scales. Almost the entire body is covered by not overlapping three-lobed scales (U3-U73) that adhere to the basal cuticle layer along their whole perimeter. Scales are distributed in about 12 longitudinal rows, with 14 scales in the central row. Their size increases gradually from 3.0-4.5 × 1.9-2.2 µm to 4.9-5.6 × 4.1-4.3 µm in a posterior direction (Supp. file 1: Table S9). Central dorsal and dorsolateral longitudinal rows of scales begin at ca U3, while lateral rows start at ca U13. Two main types of scales could be recognized concerning the shape of the anterior lobe. (i) The head, neck, and upper-trunk scales (U3-U37) have a narrowly rounded anterior lobe and elongated posterior lobes, α = 156-163°, and β = 54-85°. The transition between the anterior and posterior lobes is indistinct and continuous, providing the scales with an A-shaped appearance (Figs 37B, D, 38C). (ii) The posterior trunk scales (U50-73) exhibit a broadly rounded anterior lobe and elongated posterior lobes, α = 153-167°, and β = 75-96°. The transition between the anterior and posterior lobes is marked unlike in the first scale type (Fig. 37F).
spines. All spines bear a distinct lateral denticle and gradually narrow towards their distal end. Keels start comparatively close to the anterior margin of scales. Spines are not straight but slightly curved. They do not differentiate into various types, only their length increases in a posterior direction (Figs 37A, C, E, H, 38A-B, Supp. file 1: Table S9). More specifically, the length increase is rather inconspicuous and gradual from the head (3.1 µm) to the posterior neck region (5.8 µm). The most pronounced length change occurs at the beginning of the trunk (ca U50), where dorsal spines increase significantly from 5.8 µm to 15.2 µm (Figs 37A, H, 38B). The lateral denticle is comparatively distant from the spine apex, i.e., d-value ranges from 1.2-3.0 μm, which corresponds to a d ratio of 16.7-23.3%.
Ventral ciliary bands and Ventral interciliary field. Unfortunately, the ventral side was not observed in detail. Reference molecules are shown in Supp. file 1: Figs S8, S12B, S20. All diagnostic molecular autapomorphies are marked by arrows. Reference alignments with corresponding nucleotide positions are in Supp. file 1: Alignments 1-4.

Etymology
The Latin adjective 'slavic·us, -a, -um' [m, f, n] ('Slavic') refers to the type locality (Devín castle) of the new species, which is an important place in Slovak history.

Paratypes
SLOVAKIA • 3 adults (the specimens were destroyed during DNA extraction); same collection data as for holotype.
Photomicrographs of the holotype are available at the Department of Zoology, Comenius University in Bratislava at https://fns.uniba.sk/en/gastrotricha/.

Type material
A DNA sample of the holotype specimen DB 40 has been deposited in the Natural History Museum, Vajanského nábrežie 2, 810 06 Bratislava, Slovakia (ID Collection Code 01427570).

Gene sequences
The nuclear 18S and ITS1-5.8S-ITS2-28S rDNA sequences as well as the mitochondrial COI sequence of the holotype specimen DB 40 have been deposited in GenBank under the following accession numbers: OM421724, OM421700, and OM424079, respectively.

Remarks
Unfortunately, all our attempts for thorough morphological investigations of this species failed. Without a name, Ch. (H.) slavicus sp. nov. would be nothing but a label and nucleotide sequences in the GenBank database. However, this entity can be clearly separated from other species by the combination of 18S, ITS region, 28S, and COI sequences. Moreover, it represents a distinct lineage in multi-gene phylogenies (Fig. 9). As ICZN (1999) allows that any part of an animal is eligible to be a name-bearing type (Article 72.5.1), we interpret the isolated DNA as a part of an animal and use it as type material. This strategy is also used in protists whose names are governed by the Zoological Code (e.g., Lynn et al. 2018;Pecina & Vďačný 2022). It is important to mention that the principle of priority applies even if any part of an animal is named before the whole animal (Article 23.3.2.1).

Key to species of the subgenus Hystricochaetonotus Schwank, 1990
The following key contains all species originally assigned to the subgenus Hystricochaetonotus by Schwank (1990) and, later on, by Balsamo (1990), Balsamo & Todaro (1995), Kisielewski (1997a), Kolicka (2016Kolicka ( , 2019a, Kolicka et al. (2018), and Todaro (2022). The key uses body shape and size as well as the scale and spine patterns as the main species discriminators following Schwank (1990). The subgeneric assignment of species without associated molecular information needs to be tested (i.e., confirmed or rejected) with phylogenetic analyses in the future. Identification of species within the subgenus Hystricochaetonotus will be most reliable for specimens from which COI and/or ITS2 barcode data are available for comparison of nucleotide sequences with those from type specimens. To avoid any doubts in species identifications, the original descriptions and authoritative redescriptions need to be considered.     Greuter, 1917

Differential diagnoses of new species
The new species could be well separated from each other by (i) body shape, (ii) the organization of the cephalic pleurae and hypostomium, (iii) the presence/absence of cuticular teeth and pharyngeal dilatations, (iv) the number of longitudinal rows of scales, scales in a central dorsal row and scale types as well as by (v) the spine morphology. Their differential diagnoses are summarized in Table 5.
Chaetonotus (H.) gulosus sp. nov. is outstanding among members of the subgenus Hystricochaetonotus in having comparatively big, triangular trunk scales. In this aspect, it most resembles Ch. (H.) aemilianus Balsamo, 1978. However, Ch. (H.) gulosus sp. nov. can be differentiated from it by (i) the body size (107 µm vs 80-100 µm), (ii) the arrangement of triangular scales (close to each other but not overlapping and forming about 7 horizontal rows vs loosely spaced and restricted to the central dorsal trunk region forming 3-4 horizontal rows), (iii) the length of spines emerging from the triangular scales (up to 14 µm vs 18-26 µm), and (iv) the minimum length of lateral spines (3.1 µm vs 1.8 µm).
Although Ch. (H.) avarus sp. nov., Ch. (H.) optabilis sp. nov. and Ch. (H.) iratus sp. nov. do not display a lateral denticle on spines, they fall within the maximally supported Hystricochaetonotus clade (Figs 8-9) similarly as does Ch. (Ch.) bombardus (Kolicka et al. 2018). This indicates that the presence/ absence of a lateral denticle is a homoplastic feature. Indeed, a lateral denticle is present on the spines of Lepidochaetus Kisielewski, 1991(Križanova & Vďačný 2021 and references cited therein) and some members of the subgenus Chaetonotus (for a review, see Schwank 1990). However, as other members of the subgenus Hystricochaetonotus, these four species also display distinctly three-lobed scales and well-developed, gradually elongating spines towards the trunk rear.
Chaetonotus (H.) avarus sp. nov. matches well Ch. (Ch.) bombardus in having a five-lobed head with clearly demarcated pleurae and a pharynx without anterior and posterior dilatations. However, both species could be distinguished by (i) the body shape (slender vs stocky), (ii) the number of scale rows (12 vs 29-31), (iii) the number of scales per central dorsal row (30-32 vs 21-25), (iv) the morphology of posterior lobes of dorsal scales (comparatively broadly rounded vs narrow and tapering), (v) the presence vs absence of cuticular teeth in the mouth, (vi) the morphology of the hypostomium (a pentagonal plate with two protuberances vs ship-shaped with a deep, semicircular anterior notch), (vii) the anterior
Chaetonotus (H.) optabilis sp. nov. resembles Ch. (Ch.) bombardus in having (i) rather short spines, (ii) unmerged anterior regions of the ventral ciliary bands, (iii) non-overlapping scales in the ventral interciliary field, and in lacking (iv) cuticular teeth in the mouth center. However, they can be separated by (i) the body shape (slender vs stocky), (ii) the cephalic pleurae (not recognizable vs clearly demarcated), (iii) the presence vs absence of pharyngeal dilatations, (iv) the number of scale rows (16-18 vs 29-31), (v) the number of scales per central dorsal row (40-45 vs 21-25), (vi) the morphology of posterior lobes of dorsal scales (comparatively broadly rounded vs narrow and tapering), (vii) the presence vs absence of rounded scales scattered throughout the dorsal side, (viii) the morphology of the hypostomium (a rectangular plate with small central protuberance and accompanied by two lamellae vs ship-shaped with a deep, semicircular anterior notch), (ix) the shape of scales in the interciliary field (bowl-shaped and obtriangular vs tear-shaped), and (x) the morphology of ventral furcal scales (oval to broadly fusiform and keeled vs leaf-like and spined).
Chaetonotus (H.) iratus sp. nov. and Ch. (Ch.) bombardus share (i) a similar body shape, (ii) clearly demarcated pleurae, (iii) a pharynx without anterior and posterior dilatations, and (iv) a similar number of scales per central dorsal row (22 vs 21-25). However, both species could be distinguished by (i) the morphology of posterior lobes of dorsal scales (comparatively broadly rounded vs narrow and tapering), (ii) the presence vs absence of a cuticular tooth in the mouth center, and (iii) the structure of the hypostomium (bears two horizontal lamellae accompanied by tear-shaped protuberances vs ship-shaped with a deep, semicircular anterior notch).

Subgeneric affiliation of the new species
As mentioned above, the genus Chaetonotus and its subgenera Chaetonotus and Hystricochaetonotus, as defined by characters of the external morphology by Schwank (1990) and Kisielewski (1997a), are non-monophyletic (e.g., Kånneby et al. 2013;Kolicka et al. 2018Kolicka et al. , 2020present study). Due to the great intraspecific morphological variability and the existence of species with intermediate features, Balsamo et al. (2009) synonymized the subgenus Hystricochaetonotus with the subgenus Chaetonotus. They also stated that there is a continuum of morphological characters between both subgenera, which might thus form a single natural group. However, the synonymization of both subgenera did not erase the nonmonophyly problem of the genus Chaetonotus, which embraces even members of other chaetonotid genera in molecular phylogenies (Fig. 8). Kånneby & Hochberg (2015: 214), therefore, suggested that the re-classification of the suborder Paucitubulatina needs to be based on other characters than cuticular structures. We agree and propose that molecular diagnostic characters, further splitting of the genus Chaetonotus, and elevation of its subgenera to the generic level might help to reduce polyphyly of Chaetonotus. Phylogenetic positions of type species (i.e., name-bearing types) of chaetonotid genera and subgenera will be crucial to solving the non-monophyly problem, as name-bearing types define nominal taxa and are objective standards of reference for the application of their scientific names (Article 61.1 of the ICZN 1999).  (Fig. 9). According to the present multi-gene phylogenetic analyses, the lineage leading to the last common ancestor of this clade is the longest internal branch within the family Chaetonotidae (Fig. 8). We, therefore, argue that the subgenus Hystricochaetonotus is valid and molecularly well-delimited. In the future, only species that will statistically robustly cluster with the type species Ch. (H.) hystrix should be assigned to the subgenus Hystricochaetonotus. Indeed, the existence of this clade is corroborated by both nuclear and mitochondrial markers independently. However, only some members of this clade display all the morphological features (i.e., a cuticular covering made of three-lobed scales, spines thick and with a lateral denticle, and a varying number of spines considerably elongated to form a dorsal group) used to diagnose the subgenus Hystrichochaetonotus by Schwank (1990) and Kisielewski (1997a). Already Kolicka et al. (2018) Kolicka et al. (2018) proposed that adding further species to the phylogenetic trees might either resolve or revise the subgeneric affiliation of Ch. (Ch.) bombardus. However, after the inclusion of nine further species, the situation has become even much more intricate (Fig. 9). We, therefore, decided to assign all the nine new species to the subgenus Hystricochaetonotus. Our decision is based on the core principle of phylogenetic systematics-the grouping of taxa based on common ancestry (e.g., Wägele 2005;Wiley & Lieberman 2011). Since members of the subgenus Hystricochaetonotus are morphologically highly heterogeneous, parallel evolution of Chaetonotus-like and/or Hystricochaetonotus-like characters of scales and spines very likely occurred during its radiation. Possibly, with the growing number of species and detailed morphological data, the 'Hystricochaetonotus' clade could be split into multiple subgenera or its diagnosis could be broadened to better reconcile morphology and molecules.

Morphology and molecules in systematics and taxonomy
The phenetic approach determines taxonomic relationships based on similarity. On the other hand, morphology-based phylogeny is based on assumed homologies. These might be, however, sometimes very difficult to recognize not only in micro-but also in macroorganisms. Erroneous assumptions about homologies and not recognized homoplasies could lead to improper taxonomic frameworks (e.g., Wägele 2005;Wiley & Lieberman 2011). Moreover, there is no law in nature that would require that the diversity and evolutionary relationships be also reflected in the morphological disparity or similarity. A textbook example of how morphology can disguise true evolutionary relationships and classification is that of aquatic cetaceans (whales and dolphins), which evolved from even-toed ungulates, with hippopotami as their closest living relatives. Despite their huge size, detailed anatomical data, and a body of fossil record, this fact was recognized only after the application of molecular methods and resulted in paraphyly of the traditional order Artiodactyla Owen, 1848 (Geisler & Theodor 2009). Such sort of non-monophyly problems becomes even more true in the microscopic world. For instance, characters of cuticular structures, which are traditionally used in the chaetonotid taxonomy (Kisielewski 1997a), are distributed homoplastic along phylogenetic trees inferred from several thousands of nucleotide characters (Kieneke & Schmidt-Rhaese 2015: 89), causing polyphyly/paraphyly of Chaetonotus and its subgenera Chaetonotus, Hystricochaetonotus, and Primochaetus Kisielewski, 1997(e.g., Kånneby et al. 2013Kolicka et al. 2020;Križanová & Vďačný 2021;Figs 12-13). To reconcile these problems, which became apparent already in the first morphology-based cladistic analyses (Kieneke et al. 2008), molecular data need to be called and fully integrated into the gastrotrich systematics. It is also important to mention that molecular data are necessary to avoid tautology that might affect morphology-based frameworks since chaetonotid genera and subgenera have been defined only morphologically and rather broadly. Besides the circular reasoning, limitations of cladistic analyses include also the lack of explicit assumptions about character evolution and phylogenetic informativeness of the characters selected as well as the non-detection of multiple changes in the course of evolution. Although molecular markers provide much more independent and neutrally evolving characters than morphology, the reliability of molecular phylogenetic analyses significantly depends on the sequence quality, data curation (e.g., alignment and masking strategies), taxon sampling, and accuracy of species identification (for a review, see, for instance, Lücking et al. 2021). Electropherograms need to be, therefore, always very carefully inspected during sequence assembly and sequences containing PCR or sequencing errors should be excluded from analyses. It is also recommended to test various alignment strategies and tree-building methods to ensure the robustness of phylogenetic trees. Hologenophores (i.e., specimens that first served for morphological and then for molecular analyses) constitute an essential link between sequence data and their taxonomic origin, providing a means to verify the taxonomic identity of the specimens sequenced (Pleijel et al. 2008;Degma 2018). Finally, the taxon sampling should be comprehensive and balanced, which applies also to morphological analyses. Already in the late 70's of the past century, Woese & Fox (1977) noted that an organism's genome seems to be the ultimate record of its evolutionary history and molecules can be used to estimate the branching order of speciation events. Essential components of ribosomes were among the first molecules utilized in the reconstruction of the evolutionary history of all cellular domains of life. Ribosomal RNA genes have hypervariable regions separated by highly conserved stretches of DNA, which is transcribed into structurally constrained RNAs (Petrov et al. 2014). Hypervariable regions diverge with increasing evolutionary distance while conserved regions essential for the secondary and tertiary structure remain unchanged, causing rRNA genes to be subject to both neutral and purifying selection (Chakravorty et al. 2007). Similar rules hold also for the internal transcribed spacers (ITS1 and ITS2) of rRNA genes, though their molecular evolution is much faster and more neutral as they represent non-coding DNA (Coleman 2003(Coleman , 2007(Coleman , 2009. At the turn of the millennium, the mitochondrial gene encoding for cytochrome c oxidase subunit I was added to rRNA genes to improve the delimitation of very closely related species. The taxonomic power of COI comes from its accelerated molecular evolution, which stems from the absence of an excision repair system in animal mitochondria (e.g., Hebert et al. 2003aHebert et al. , 2003b and references cited therein).
Ribosomal RNA genes were considered to be also a reliable molecular clock, though in its relaxed form that accounts for the rate variation across species, different genes, and different sites along a single gene. The conventional estimate for evolutionary rates of prokaryotic 16S rDNA is 1.8 × 10 -10 substitutions per site per year, i.e., 1.8% sequence divergence per 100 Ma (Ochman & Wilson 1987). In microbial eukaryotes, the homologous 18S rDNA evolves at a rate of 1.24-3.96 × 10 -10 substitutions per site per year, i.e., about 2.6% sequence divergence per 100 Ma (Wright & Lynn 1997;Vďačný 2015;Vďačný et al. 2019). In multicellular organisms, about 0.8% sequence divergence per 100 Ma is considered to reflect the rate of the whole 18S rDNA whereas the 2% difference might reflect the rapid evolution of hypervariable regions of the 18S rRNA molecule. ITS2 molecules evolve about 23-55 times faster than 18S rDNA (Bargues et al. 2000 and references cited therein). Nevertheless, the coding part of the rDNA cistron evolves comparatively slowly and is subject to concerted evolution, a process that converts copies of a gene in a multigene family into the same copy. Every single mutation fixed in the rDNA cistron, therefore, matters. To reflect this very important fact, we included molecular autapomorphies and 2D structures of molecules into species diagnoses in the present study.
We have recently assessed the utility of molecules in the identification of chaetonotid species (Križanová & Vďačný 2021). The present barcoding analyses (Fig. 4A-D) corroborate very well our previous findings. More specifically, intra-and interspecies p-distances of 18S sequences might overlap, i.e., two distinct species might share identical 18S rDNA sequences. This phenomenon is associated with the high conservativeness of the 18S rRNA gene (see above) and/or a rather recent divergence of species. However, a few nucleotide positions might be rarely polymorphic, leading to intraspecies distances of up to 0.11% (present study). Due to the combination of high conservativeness and rare polymorphism, 18S is not suitable to become a DNA barcode or even a pre-barcode in the chaetonotid molecular taxonomy. On the other hand, 28S can be used as a DNA pre-barcode, whereby helices 25es7-25es7c are taxonomically most important since species-specific mutations tend to accumulate therein. And, eventually, COI is an ideal mitochondrial DNA barcode enabling unambiguous identification of chaetonotids. In the present study, we have added ITS2 as a further reliable nuclear DNA barcode (Fig. 4B). Each studied species of the subgenus Hystricochaetonotus has a unique primary and secondary structure of the ITS2 molecule (Supp. file 1: Figs S9-S12). Furthermore, the interspecies p-distances range from 5.88% to as much as 34.76% and there are up to four CBCs between species (Table 3). It is important to note that no intraspecies variability has been detected in ITS2, which might be associated with its shortness. According to Müller et al. (2007), a single CBC in a helix can differentiate two species with a probability of 0.93. Interestingly, we have detected at least a single CBC within the rDNA cistron between almost all species pairs (Table 3). If there are no CBCs, species could be well separated by the primary structure of 18S, ITS2, and/or 28S and by the primary structure of COI. These findings document that rDNA cistron along with COI are powerful tools for the delimitation and identification of gastrotrich species.

Conclusions
Extant gastrotrich species are the outcome of millions of years of evolution that has been written not only in their morphology but also in their molecules. Hitherto, this fact has not been reflected in species diagnoses that traditionally contain only morphological features. We, therefore, argue here for the full integration of molecules in gastrotrich taxonomic and classification frameworks and for their place in species recognition, identification, and diagnosis. Molecules should also become reliable diagnostic features of cryptic or near-cryptic species, as they contain the key characters that distinguish among them. Using the integrative morpho-molecular taxonomic approach, we have discovered and characterized nine new gastrotrich species collected from the continental waters of Central Europe. Table S1. List of taxa with GenBank accession numbers of 18S, 28S and COI sequences included in phylogenetic analyses. Table S2. Morphometric characterization of Chaetonotus (Hystricochaetonotus) mirabilis sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. N = number of specimens analyzed. Table S3. Morphometric characterization of Chaetonotus (Hystricochaetonotus) superbus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. N = number of specimens analyzed. Table S4. Morphometric characterization of Chaetonotus (Hystricochaetonotus) optabilis sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. N = number of specimens analyzed. Table S5. Morphometric characterization of Chaetonotus (Hystricochaetonotus) avarus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. N = number of specimens analyzed. Table S6. Morphometric characterization of Chaetonotus (Hystricochaetonotus) luxus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. N = number of specimens analyzed. Table S7. Morphometric characterization of Chaetonotus (Hystricochaetonotus) iratus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. Table S8. Morphometric characterization of Chaetonotus (Hystricochaetonotus) gulosus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement. Table S9. Morphometric characterization of Chaetonotus (Hystricochaetonotus) arcanus sp. nov. All measurements are given in μm. Ranges include the smallest and the largest structure measurement.              S13. Secondary structure of the first two domains of the 28S rRNA molecule of Chaetonotus (Hystricochaetonotus) superbus sp. nov. Diagnostic molecular autapomorphies are marked by red arrows. The reference 28S secondary structure map of Saccharomyces cerevisiae Meyen ex E.C.Hansen (inset) is from http://apollo.chemistry.gatech.edu/RibosomeGallery (Petrov et al. 2014). Fig. S14. Secondary structure of the first two domains of the 28S rRNA molecule of Chaetonotus (Hystricochaetonotus) optabilis sp. nov. Diagnostic molecular autapomorphies are marked by red arrows. The reference 28S secondary structure map of Saccharomyces cerevisiae Meyen ex E.C.Hansen (inset) is from http://apollo.chemistry.gatech.edu/RibosomeGallery (Petrov et al. 2014).