Integrative taxonomy resolves species identities within the Macrobiotus pallarii complex (Eutardigrada: Macrobiotidae)

The taxonomy of many groups of meiofauna is challenging due to their low number of diagnostic morphological characters and their small body size. Therefore, with the advent of molecular techniques that provide a new source of traits, many cryptic species have started to be discovered. Tardigrades are not an exception, and many once thought to be cosmopolitan taxa are being found to be complexes of phenotypically similar species. Macrobiotus pallarii Maucci, 1954 was originally described in South Italy and has been subsequently recorded in Europe, America, and Asia. This allegedly wide geographic range suggests that multiple species may be hidden under this name. Moreover, recently, genetic evidence to support this was put forward, and the Macrobiotus pallarii complex has been proposed to accommodate putative species related to M. pallarii. Here, we describe three new pseudocryptic species based on populations that would have been all classified as Macrobiotus pallarii if molecular methods were not employed. Using an integrative taxonomy approach, we analyzed animals and eggs from the topotypic population of Macrobiotus pallarii, together with four other populations of the complex. We recovered four distinct phylogenetic lineages that, despite the overlap of morphometric traits, can be separated phenotypically by subtle but discrete morphological characters. One lineage corresponds to Macrobiotus pallarii, whereas the other three are newly described as Macrobiotus margoae Stec, Vecchi & Bartels, sp. nov. from the USA, Macrobiotus ripperi Stec, Vecchi & Michalczyk, sp. nov. from Poland and Finland, and Macrobiotus pseudopallarii Stec, Vecchi & Michalczyk, sp. nov. from Montenegro. To facilitate species identification, we provide a dichotomous key for species of the M. pallarii complex. Delimitation of these pseudocryptic taxa highlights the need for an integrative approach to uncover the phylum’s diversity in full.


Background
Tardigrades constitute a phylum of microinvertebrates inhabiting both terrestrial and aquatic (freshwater and marine) habitats worldwide [1], with approximately 1300 species known so far [2][3][4]. Tardigrade taxonomy is challenging due to their small size, limited number of taxonomically informative traits, and many outdated species descriptions that do not comply with modern taxonomic standards. Advances have been made possible with the recent advent of integrative taxonomy (e.g., [5][6][7][8][9][10]) that includes the use of DNA sequencing with detailed morphological techniques such as scanning electron microscopy (SEM) and sophisticated analysis of morphometric data. Many studies using an integrative taxonomy approach have recently revealed that various tardigrade species once thought to be widespread are actually complexes of cryptic species with more localized distributions [5,6,11,12]. In addition to their distribution, tardigrade cryptic species have also been shown to diverge based on reproductive mode [13], ploidy [5], and anhydrobiotic survival [14]. Three classes have been recognized in the phylum Tardigrada: Heterotardigrada, Mesotardigrada, and Eutardigrada [15]. The class Heterotardigrada encompasses both primarily marine tardigrades (Arthrotardigrada) and mostly limnoterrestrial armoured tardigrades (Echiniscoidea). The class Mesotardigrada has been found only once from a thermal spring in Japan and is considered a classis dubium [16]. The class Eutardigrada comprises two mostly limnoterrestrial orders, Apochela and Parachela. The order Parachela, which is the most common, widespread, and speciose group with a wide range of dietary preferences [17], comprises the superfamily Macrobiotoidea. The nominal genus for the superfamily Macrobiotus CAS Schultze, 1834 [18] is characterized by the presence of symmetrical diploclaws, a rigid buccal tube with a straight ventral bar lacking a ventral hook, two macroplacoids and one microplacoid in the pharynx, 10 peribuccal lamellae, pores in the cuticle, and freely laid ornamented eggs [19]. Macrobiotus is one of the most species-rich and widespread genera in the phylum, and it was also the first formally described tardigrade genus [20]. Animals of the Macrobiotus pallarii complex have the very typical morphology of Macrobiotus. However, this group is characterized by egg ornamentation composed of large conical processes separated by a single row of areolation (such eggs are also known in other macrobiotid genera, such as Paramacrobiotus Guidetti et al., 2009 [21] but have not been found in other Macrobiotus species). To date, three described species belong to this complex, with the type species being Macrobiotus pallarii Maucci, 1954 [22], described from South Italy and a senior synonym of Macrobiotus aviglianae Robotti, 1970 [23,24], which was described from North Italy. Macrobiotus pallarii has been found in various localities in Europe, Asia and North America [25]; however, none of the records outside the type locality have been genetically verified; thus, the true geographic range of this species is unknown. The second species, Macrobiotus ragonesei Binda et al., 2001 [26], was described and found only in Central Africa [26,27], whereas the third species, Macrobiotus caymanensis Meyer, 2011 [28], was found only in the Cayman Islands (the Caribbean). Additionally, three phylogenetically distinct but undescribed and morphologically very similar species of this group have been recently detected in several localities in Europe and North America [19]. Given that the existence of these species and morphological characters that allow for their phenotypic identification became apparent only after the molecular analysis had been performed, they can be considered pseudocryptic species [29,30]. To elucidate the taxonomy of this species complex, we analyzed (and resequenced) the populations reported by Stec et al. [19], as well as individuals and eggs from the type locality of M. pallarii, from phylogenetic (multilocus phylogeny and species delimitation), morphometric (principal component analysis; PCA), and morphological perspectives. Morphometric traits alone do not allow for a separation of M. pallarii and the three new pseudocryptic species, as almost all their biometric values overlap considerably. However, our analysis identified some qualitative phenotypic characters that could be used to separate and formally describe three new species within this morphologically uniform species complex. Last, to facilitate species identification, we provide a dichotomous key for species of the M. pallarii complex.

Samples and specimens
Five populations of the Macrobiotus pallarii complex were isolated from moss samples collected from five localities in Europe and North America (see Table 1 for details). The samples were examined for tardigrades using the protocol by [31], with modifications described in detail in [32]. The live animals and eggs were placed into culture. Specimens were reared in plastic Petri dishes according to the protocol by [32]. To perform the taxonomic analysis of these species/populations, animals and eggs were taken from the cultures and split into several groups for specific analyses: (i) morphological and morphometric analyses with light contrast microscopy (LCM), (ii) morphological analysis with SEM, and (iii) DNA sequencing (for details see sections "Material examined" provided below for each description). Genotyping DNA was extracted from individual animals following a Chelex® 100 resin (BioRad) extraction method by [33] with modifications described in detail in [34]. Individuals from the same populations that were sequenced by Stec et al. [19] (populations FI.066-PL.015-ME.007-US.057) were resequenced in the present study, as [19] provided sequences at the haplotype level, whereas species delimitation methods require sequences at the level of individuals. Each specimen was mounted in water and examined under LCM prior to DNA extraction. We sequenced four DNA fragments, three nuclear fragments (18S rRNA, 28S rRNA, ITS-2) and one mitochondrial fragment (COI). All fragments were amplified and sequenced according to the protocols described in [34]; primers with original references are listed in Table 2. Sequencing products were read with an ABI 3130xl sequencer at the Molecular Ecology Lab, Institute of Environmental Sciences of Jagiellonian University, Kraków,  Table 2 Primers with their original references used for amplification of the four DNA fragments sequenced in the study. The primer set LCO1490-JJ + HCO2198-JJ was used for COI amplification in four populations (FI. 066, IT. 337, PL. 015, US. 057), whereas LCO1490 + HCOoutout was used for one population (ME.007) Poland. Sequences were processed in BioEdit ver. 7.2.5 [42] and submitted to NCBI GenBank [43]. For accession numbers see Table 3.

Phylogenetic analysis
To reconstruct the phylogeny, we used sequences representing four markers (18S rRNA, 28S rRNA, ITS-2, and COI) from five different populations of the Macrobiotus pallarii complex. Sequences were downloaded from GenBank or produced de novo (Table 3). Type sequences of Macrobiotus caelestis Coughlan, Michalczyk & Stec, 2019 [44] were used as the outgroup. 18S rRNA, 28S rRNA, and ITS-2 sequences were aligned with MAFFT ver. 7 [45,46] with the G-INS-i method (thread=4, threadtb=5, threadit=0, reorder, adjustdirection, anysymbol, maxiterate=1000, retree 1, globalpair input). COI sequences were aligned according to their amino acid sequences (translated with the invertebrate mitochondrial code) with the MUSCLE algorithm [47] in MEGA ver. 7.0.26 [48] with default settings (all gap penalties=0, max iter-ations=8, clustering method=UPGMB, lambda=24) and then translated back to nucleotide sequences. Alignments were visually inspected and trimmed in MEGA ver. 7.0.26 [48]. The absence of saturation in the COI and ITS-2 alignments was confirmed with transition/transversion and saturation plots (SM.01) produced with the R package "ape ver. 5.0" [49]. Aligned sequences were concatenated with an in-house R script (written by MV, available upon request to the author). Model selection and phylogenetic reconstructions were performed on the CIPRES Science Gateway [50]. Model selection was performed for each alignment partition (6 in total: 18S, 28S, ITS-2, and three COI codons) with Partition-Finder ver. 2 [51]. Bayesian inference (BI) phylogenetic reconstruction was performed with MrBayes ver. 3.2.6 [52] without BEAGLE. Four runs with one cold chain and three heated chains were run for 20 million generations with a 10% burn-in, sampling a tree every 10000 generations. Posterior distribution was checked with Tracer ver. 1.7 [53]. The MrBayes input file is available as Supplementary Material (SM.02). The phylogenetic tree was visualized with FigTree ver. 1.4.4 [54], and the image was edited with Inkscape ver. 0.92.3 [55].

Species delimitation
Following the suggestion of [56], species delimitation was performed with both a tree-based (bPTP) and a distance-based (ABGD) method. Tree-based species delimitation was performed with bPTP software [57] on both ITS-2 and COI trees. Single gene alignments and BI trees were produced as described above (see Phylogenetic Analysis section). One thousand trees were sampled randomly from the posterior tree distribution after discarding the burn in (250 from each chain) and used as input for bPTP with 100,000 generations, a thinning of 100 generations and 10% burn in. Distancebased species delineation was performed with the ABGD online server [58] on both ITS-2 and COI alignments obtained for the phylogenetic analysis as described above. For both markers, simple distance was used, with 10 steps, a relative gap width of 1.5, and 20 bins for distance distribution. For COI, Pmin and Pmax were 0.01 and 0.1, respectively, while for ITS-2, they were 0.0001 and 0.1, respectively, to accommodate the difference in divergence between these loci.

The p-distances
As the species of the M. pallarii complex are phylogenetically and morphologically distinct [19], the p-distances for the genetic differential diagnosis were calculated between species of the M. pallarii complex for the four sequenced markers (18S, 28S, ITS2, and COI) using the alignment used for the phylogenetic analysis. Pairwise distances were calculated with the software MEGA ver. 7.0.26 [48] using pairwise deletion for the Gap/Missing Data Treatment option. Detailed p-distance tables are provided in SM.03.

Microscopy and imaging
Specimens for LCM were mounted on microscope slides in a small drop of Hoyer's medium and secured with a cover slip, following the protocol by [59]. Slides were examined under an Olympus BX53 light microscope with phase (PCM) and Nomarski differential interference contrast (NCM) associated with an Olympus DP74 digital camera. To obtain clean and extended specimens for SEM, tardigrades were processed according to the protocol by [32].

Morphometrics and morphological nomenclature
All measurements are given in micrometers (μm). Sample size was adjusted following recommendations by [60]. Structures were measured only if their orientation was suitable. Body length was measured from the anterior extremity to the end of the body, excluding the hind legs. The terminology used to describe oral cavity armature and eggshell morphology follows [61,62].
Macroplacoid length sequence is given according to [63]. Buccal tube length and the level of the stylet support insertion point were measured according to [64]. The pt index is the ratio of the length of a given structure to the length of the buccal tube expressed as a ratio [64]. Measurements of buccal tube widths, heights of claws and eggs follow [62]. Morphometric data were handled using the "Parachela" ver. 1.7 template available from the Tardigrada Register [65]. Tardigrade taxonomy followed [19,66]. Morphometric data for eggs and animals were analyzed with PCA. All analyses were performed with the base R software package [67]. For eggs, absolute values (raw measurements in μm) were used for the analysis, whereas for the animals, relative (pt) values were analyzed. Missing data in the animal data set were imputed with the PCA imputation technique with the "imputePCA" function of the R package "missMDA ver. 1.17" [68]. The number of components used to impute the missing values was determined by cross-validation (function "estim_ncpPCA"). PCAs were performed on the scaled data with the PCA function of the package "FactoMineR ver. 2.3" [69]. PCAs were visualized with the packages "ggplot2 ver. 3.3.2", "plyr ver. 1.8.6" and "gridExtra ver. 2.3" [70,71]. The presence of a structure in the PCA data was tested with a randomization procedure according to [72] on the eigenvalues and with the statistics ψ and ϕ using an in-house R script (written by MV, in Supplementary SM.04a-4b).
PERMANOVA was performed on the PCs using the species hypothesis obtained by phylogenetic methods as the independent variable with the R package "vegan ver. 2.5.6" and "pairwiseAdonis ver. 0.3" [73]. The R script of all the morphometric analyses is available as Supplementary SM.4a-4b. Additionally, morphometric data from the type series of Macrobiotus caymanensis were included in this analysis (population name: Cayman). The use of Thorpe normalization proposed by [74] for the comparison of morphometric traits between different species was not used due to the low sample size for M. pallarii and M. caymanensis. The α-level for multiple post hoc comparisons was adjusted separately for adult and egg traits using the Benjamini-Hochberg correction [75].

Availability of data and materials
The datasets supporting the conclusions of this article are available in the GenBank repository (https://www.ncbi. nlm.nih.gov/genbank; Accession Numbers in Table 3) and in the Tardigrada

Phylogenetic analysis
The BI phylogenetic reconstructions yielded a topology ( Fig. 1) with four well-supported clades: the first clade comprised all individuals from the US population; the second clade contained all individuals from the Polish (PL) and Finnish (FI) populations; the third clade contained individuals from the Montenegrin (ME) population; and the fourth clade contained individuals from the IT population.

Species delimitation
The bPTP analysis of COI and ITS-2 markers gave slightly different results (Fig. 1). The species identified based on the COI dataset are in agreement with the four clades found with the phylogenetic analysis. The ITS-2 dataset species delimitation, however, found only three species, as the IT and ME populations were delineated as a single species. The ABGD results were congruent with the bPTP results ( Fig. 1). For COI, only for the smallest prior intraspecific divergence (0.01) were six species recovered (the PL and FI populations were delimited as two species); however, we did not consider this result to be valid, as it was not congruent with all the other nine evaluated prior intraspecific divergence values or with the bPTP result for the same marker.

Morphometric analysis
The randomization test in the PCA revealed that for both animal and egg datasets, only the first two PCs explained more variation than expected by the null model (no data structure) (Supplementary SM.10); therefore, only the first two PCs were retained and used for further analysis and interpretation. Additionally, the ψ and ϕ statistics of the PCA were significantly different from their expectations under the null model (animals: ψ= 132.75 p<0.001, ϕ=0.45 p<0.001; eggs: ψ=6.29 p<0.001, ϕ=0.39 p<0.001). Principal component analysis (PCA) of the pt indices of animals ( Fig. 2a) described 56% of the total variance with the first two components (46.0% for PC1 and 9.9% for PC2). PERMANOVA showed an overall significant effect of species identity on the PCs (p< 0.001, Table 4). The majority of the post hoc pairwise Horizontal colored boxes highlight the identified species described in the Taxonomic Account. A schematic representation of the dorso-caudal granulation patterns for each described species is presented (for a detailed description, see the Taxonomic Account) PERMANOVA comparisons were significant (Table 4); however, the species could not be separated by any of the analyzed traits (Fig. 2a), which was also indicated by low R 2 values (Table 4), thus making morphometric indices impractical for traditional species identification. The only exception was represented by two groups of populations (M. margoae + M. caymanensis vs M. pallarii + M. pseudopallarii + M. ripperi) that showed some separation in the first and second PCs (Fig. 2a). According to the loading plot of PC1 and PC2 (Fig. 2a), the separation between these two groups was driven mainly by the pt indices related to the buccal apparatus structures. Principal component analysis (PCA) of egg measurements (Fig. 2b) described 60% of the total   Table 5). All of the post hoc pairwise PERMANOVA comparisons, except some concerning M. caymanensis (probably due to the very low sample size of this species), were significant (Table 5). However, similar to animal traits, egg measurements also did not separate the analyzed species (Fig. 2b).

Taxonomic accounts
Phylum: Tardigrada  Integrative description of the topotypic population of the species Animals (measurements and statistics in Table 6): In live animals, body almost transparent in smaller specimens and whitish in larger animals; transparent after fixation in Hoyer's medium (Fig. 3). Eyes present in live animals and after fixation in Hoyer's medium. Small round and oval cuticular pores (0.5-1.5 μm in diameter) visible under both LCM and SEM scattered randomly throughout the entire body (Figs. 4a-e, 5a-d). Patches of fine granulation on the external surface of legs I-III as well as on the dorsal and dorsolateral sides of leg IV visible in LCM (Fig. 4c, e) and SEM (Fig. 5b, d). A pulvinus is present on the internal surface of legs I-III (Figs. 4d, 5c). In addition to the typical patches of leg granulation, a band of granulation is present on the dorso-and latero-caudal surface of the last body segment (Figs. 2, 4a, 5a). It consists of two lateral patches of dense granulation joined by a band of sparse dorsal granulation (Figs. 2, 4a, 5a). The sparse granulation also extends posteriorly from these dense granulation patches towards the granulation on leg IV, but they never connect (Figs. 2, 4a, 5a). This granulation can be poorly visible under LCM when the cuticle is wrinkled (Fig. 4b).
Claws slender, of the hufelandi type. Primary branches with distinct accessory points, a long common tract, and an evident stalk connecting the claw to the lunula (Fig. 6a-f). Lunulae on all legs smooth and only sometimes faintly crenulated on leg IV ( Fig.  6a-f). Dark areas under each claw on legs I-III were often visible in LCM (Fig. 6a). Paired muscle attachments and faintly visible cuticular bars above them on legs I-III were often visible both with LCM and SEM (Fig. 6a, d), whereas the horseshoe-shaped structure connecting anterior and posterior claw IV was visible only in LCM (Fig. 6b-c).
Mouth antero-ventral. Buccal apparatus of the Macrobiotus type (Fig. 7a), with the ventral lamina and ten peribuccal lamellae ( Fig. 8a-b). The oral cavity armature was well developed and composed of three bands of teeth, all always clearly visible under LCM (Fig. 7b-c).
The first band of teeth is composed of numerous small teeth visible under LCM as granules (Fig. 7b-c) and under SEM as cones ( Fig. 8a-b), arranged in several rows, situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae. The second band of teeth is situated between the ring fold and the third band of teeth and comprises 3-4 rows of teeth visible with LCM as granules (Fig. 7b-c), and as cones in SEM ( Fig. 8a-b) but larger than those in the first band. The most anterior row of teeth within the second band comprises larger teeth than the subsequent posterior rows ( Fig. 7b-c). The teeth of the third band are located within the posterior portion of the oral cavity, between the second band of teeth and the buccal tube opening (Figs. 7b-c, 8a-b). The third band of teeth is divided into the dorsal and ventral portions. Under both LCM and SEM, the dorsal teeth are seen as three distinct transverse ridges, whereas the ventral teeth appear as two separate lateral transverse ridges, between which one large tooth (sometimes circular in LCM) is visible (Figs. 7b-c, 8a-b). In SEM, teeth of the third band have faintly indented margins ( Fig. 8a-b). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids (2<1) and a microplacoid positioned close to them (i.e., the distance between the second macroplacoid and the microplacoid is shorter than the microplacoid length; Fig. 7a, d). The first macroplacoid is anteriorly narrowed and constricted in the middle, whereas the second has a subterminal constriction ( Fig.  7d-e).
Eggs (measurements and statistics in Table 7): Laid freely, white, spherical with conical processes surrounded by one row of areolae (Figs. 9, 10a-f). In SEM, multiple rings of tight annulation were visible on the entire process (Fig. 10a-c), although in some processes, annulation was present only in the upper portion of the process ( Fig. 10d-f) (annulation not visible in LCM

Reproduction
The species is dioecious. Spermathecae in females as well as testes in males have been found to be filled with spermatozoa, clearly visible under PCM up to 24 hours after mounting in Hoyer's medium ( Fig. 11a-b). The species exhibits secondary sexual dimorphism in the form of clearly visible lateral gibbosities on hind legs in males (Figs. 4a, 5a, 11b).

Phenotypic differential diagnosis
By having the processes surrounded by 5-6 areolae, it resembles four other species of the Macrobiotus pallarii complex out of which three are newly described in this study. By the morphology of the animals and eggs, this species can be differentiated specifically from the following: Macrobiotus pseudopallarii sp. nov.: by faintly crenulated lunulae IV (lunulae are gently dentate in M. pseudopallarii sp. nov.) and the sparse granulation on the dorso-caudal end of the body connecting the dense granulation patches between legs III and IV not extending posteriorly to the granulation on leg IV (with sparse granulation extending posteriorly to the granulation on leg IV in M. pseudopallarii sp. nov.; see Fig. 1). Macrobiotus ripperi sp. nov.: by faintly crenulated lunulae IV (lunulae are dentate in M. ripperi sp. nov.), the presence of two lateral patches of dense granulation between legs III and IV (patches of dense granulation are absent in M. ripperi sp. nov.; see Fig. 1), the sparse granulation on the dorsocaudal end of the body connecting the dense granulation patches between legs III and IV not extending posteriorly to the granulation on legs IV (with the dense granulation patches between legs III and IV absent and the sparse granulation extending posteriorly to the granulation on legs IV in M. ripperi sp. nov.; see Fig. 1) and by the presence of granulation on the tips of egg processes (granulation is absent in , and by the labyrinthine meshes within the entire process wall (only small circular bubbles scattered randomly within the process wall are found in M. caymanensis).

Genotypic differential diagnosis
Interspecific genetic p-distances between M. pallarii and other species of the M. pallarii complex are as follows:   Etymology: The name refers to the morphology of the new species, which highly resembles that of Macrobiotus pallarii (Latin "pseudo" = "false").

Description of the new species
Animals (measurements and statistics in Table 8): In live animals, body almost transparent in smaller specimens and whitish in larger animals; transparent after fixation in Hoyer's medium (Fig. 12). Eyes present in live animals and after fixation in Hoyer's medium. Small round and oval cuticular pores (0.5-1.2 μm in diameter), visible under both LCM and SEM, scattered randomly throughout the entire body (Figs. 13a-e, 14a-e). Patches of fine granulation on the external surface of legs I-III as well as on the dorsal and dorsolateral sides of leg IV visible with LCM (Fig. 13c, e) and SEM (Fig. 14c, e). A pulvinus is present on the internal surface of legs I-III (Figs. 13d, 14d). In addition to the typical patches of leg granulation, a band of granulation is present on the dorso-and latero-caudal surface of the last body segment (Figs. 2, 13a, 14a-b). It consists of two lateral patches of dense granulation, joined with each other by a band of sparse dorsal granulation (Figs. 2, 13a, 14a-b). The sparse granulation also extends anteriorly and posteriorly from those dense granulation patches with posterior extension, which connects with the granulation on leg IV (Figs. 2, 13a, 14a-b). This granulation is slightly visible under LCM when the cuticle is wrinkled (Fig. 2, 13b).
Claws slender, of the hufelandi type. Primary branches with distinct accessory points, a long common tract, and an evident stalk connecting the claw to the lunula (Fig. 15a-f). Lunulae on legs I-III smooth, whereas on leg IV gently dentate ( Fig. 15a-f). Dark areas under each claw on legs I-III were faintly visible with LCM (Fig. 15a). Paired muscle attachments and faintly visible cuticular bars above them on legs I-III were often visible both with LCM (Fig. 15a) and SEM, whereas the horseshoe-shaped structure connecting anterior and posterior claws IV was visible only under LCM (Fig. 15b-c).
Mouth antero-ventral. Buccal apparatus of the Macrobiotus type (Fig. 16a), with the ventral lamina and ten peribuccal lamellae ( Fig. 17a-b). The oral cavity armature was well developed and composed of three bands of teeth, all always clearly visible under LCM (Fig. 16b-c). The first band of teeth is composed of numerous small teeth visible with LCM as granules (Fig. 16b-c) and with SEM as cones ( Fig. 17a-b), arranged in several rows, situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae. The second band of teeth is situated between the ring fold and the third band of teeth and comprises 3-4 rows of teeth visible with LCM as granules (Fig. 16b-c) and with SEM as cones ( Fig.  17a-b) but larger than those in the first band. The most anterior row of teeth within the second band comprises larger teeth than the subsequent posterior rows (Fig.  16b-c). The teeth of the third band are located within the posterior portion of the oral cavity, between the second band of teeth and the buccal tube opening (Figs. 16b-c, 17a-b). The third band of teeth is divided into the dorsal and ventral portions. Under both LCM and SEM, the dorsal teeth are seen as three distinct transverse ridges, whereas the ventral teeth appear as two separate lateral transverse ridges, between which one large tooth (sometimes circular in LCM) is visible (Figs. 16b-c, 17a-b). In SEM, teeth of the third band have indented margins (Fig. 17a-b). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids (2<1) and a microplacoid positioned close to them (i.e., the distance between the second macroplacoid and the microplacoid is shorter than the microplacoid length; Fig. 16d-e). The first macroplacoid is anteriorly narrowed and constricted in the middle, whereas the second has a subterminal constriction (Fig. 16d-e).
Eggs (measurements and statistics in Table 9): Laid freely, white, spherical with conical processes surrounded by one row of areolae (Figs. 18a-h, 19a-f). In SEM, multiple rings of tight annulation on the entire process surface were visible (Fig. 19a-b), although in some processes, annulation was present only in the upper portion of the process (Fig. 19c-f) (annulation not visible with LCM because it was obscured by the eminent labyrinthine layer). The upper parts of the processes are covered by granulation, which to a varying extent is distributed on annuli, visible only under SEM (Fig. 19c-f). The labyrinthine layer within the process walls is present and visible as reticulation with circular/ ellipsoidal meshes throughout the entire process (Fig.  18a-c). Only sometimes small areas without reticulation are present in some processes (Fig. 18a, c); however, very rarely, the reticulation can also be considerably reduced (Fig. 18d). The upper part of the process is often elongated into short flexible apices (Figs. 18e-h, 19c-f), which can be occasionally broken. The bottom part of the processes is flattened and extends into the six (only sometimes five) arms that form areolae rims (Figs. 18ad, 19a-d). Each process is surrounded by six (only sometimes five) hexagonal areolae (Figs. 18a-d, 19a-d), which are occasionally falsely subdivided in the middle into two areolae by a thin thickening perpendicular to the process base (Figs. 18c, 19b-d). Areolae rims (walls) thick and usually flat (Fig. 19a-d), with the labyrinthine layer inside the rims visible as bubbles under LCM (Fig. 18a-d). Areolae rims also delimit the areolae at the bases of processes, which forms an irregular collar around process bases (Figs. 18a-d, 19a-d) and makes the process bases penta-or hexagonal in the top view (Figs. 18a-d, 19a-b). The areola surface has wrinkles that are faintly visible under LCM (Fig. 18a-d) but clearly visible under SEM (Fig. 19a-d). Micropores are present within the areolae, but they are distributed only around the areola rims and are usually absent in the central part of the areola (Fig. 19b-d).

Reproduction
The species is dioecious. Spermathecae in females as well as testes in males were found to be filled with spermatozoa, clearly visible under PCM up to 24 hours after mounting in Hoyer's medium ( Fig. 20a-b). The species exhibits secondary sexual dimorphism in the form of clearly visible lateral gibbosities on hind legs in males (Fig. 20b).  ITS-2 sequences (GenBank: MT809090-2), 362 bp long; 2 haplotypes were found, separated by a pdistance of 0.6%. COI sequences (GenBank: MT807920-2), 630 bp long; 2 haplotypes were found, separated by a pdistance of 0.3%.     19 Macrobiotus pseudopallarii sp. nov. from Montenegro (ME.007)eggs seen with SEM: aentire view of the egg; b-fdetails of the egg surface between processes, areolation and egg processes. Filled flat arrowheads indicate thickening perpendicular to the process base that divides the areola in the middle, empty flat arrowheads indicate irregular collars around process bases. Scale bars in μm caymanensis) and by the presence of meshes within the entire process (only small circular bubbles scattered randomly within the process walls are found in M. caymanensis).

Genotypic differential diagnosis
Interspecific genetic p-distances between M. pseudopallarii sp. nov. and other species of the M. pallarii complex are as follows: Etymology: Named after "Ripper", the giant tardigrade-like creature from the TV series "Star Trek:

Description of the new species
Animals (measurements and statistics in Table 10): In live animals, body almost transparent in smaller specimens and whitish in larger animals; transparent after fixation in Hoyer's medium (Fig. 21). Eyes present in live animals and after fixation in Hoyer's medium. Small round and oval cuticular pores (0.5-1.4 μm in diameter), visible under both LCM and SEM, scattered randomly throughout the entire body (Figs. 22a-e, 23a-f). Patches of fine granulation on the external surface of legs I-III as well as on the dorsal and dorsolateral sides of leg IV visible with LCM (Fig. 22c, e) and SEM (Fig. 23d, f). A pulvinus is present on the internal surface of legs I-III (Figs. 22d, 23e). In addition to the typical patches of leg granulation, sparse and uniformly distributed granulation is also present on the dorso-and latero-caudal surface of the last body segment (Figs. 22a, 23a-c). The sparse granulation connects with denser granulation patches on leg IV (Figs. 2, 22a, 23a-c). This granulation is slightly visible with LCM when the cuticle is wrinkled (Fig. 22b). Claws slender, of the hufelandi type. Primary branches with distinct accessory points, a long common tract, and an evident stalk connecting the claw to the lunula (Fig. 24a-f). Lunulae on legs I-III smooth, whereas on legs IV usually clearly dentate (Fig. 24a, c-f). Dentation was rarely absent or most likely not visible under LCM (Fig. 24b). Dark areas   (Fig. 24a). Paired muscle attachments and faintly visible cuticular bars above them on legs I-III were often visible with both LCM (Fig. 24a) and SEM (Fig. 24d), whereas the horseshoe-shaped structure connecting anterior and posterior claw IV was visible only with LCM (Fig. 24b-c).
Mouth antero-ventral. Buccal apparatus of the Macrobiotus type (Fig. 25a), with the ventral lamina and ten peribuccal lamellae (Fig. 26a-b). The oral cavity armature was well developed and composed of three bands of teeth, all always clearly visible under LCM (Fig. 25b-c). The first band of teeth is composed of numerous small teeth visible under LCM as granules (Fig. 25b-c) and under SEM as cones (Fig. 26a-b), arranged in several rows, situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae. The second band of teeth is situated between the ring fold and the third band of teeth and comprises 3-4 rows of teeth visible under LCM as granules (Fig. 25b-c) and under SEM as cones (Fig. 26a-b) but larger than those in the first band. The most anterior row of teeth within the second band comprises larger teeth than the subsequent posterior rows (Fig. 25b-c). The teeth of the third band are located within the posterior portion of the oral cavity, between the second band of teeth and the buccal tube opening (Figs. 25b-c, 26a-b). The third band of teeth is divided into the dorsal and ventral portions. Under both LCM and SEM, the dorsal teeth are seen as three distinct transverse ridges, whereas the ventral teeth appear as two separate lateral transverse ridges, between which one large tooth (sometimes circular in LCM) is visible (Figs. 25b-c, 26a-b). In SEM, teeth of the third band have indented margins (Fig. 26a-b). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped   to them (i.e., the distance between the second macroplacoid and the microplacoid is shorter than the microplacoid length; Fig. 25d-e). The first macroplacoid is anteriorly narrowed and constricted in the middle, whereas the second has a subterminal constriction (Fig.  25d-e).
Eggs (measurements and statistics in Table 11): Laid freely, white, spherical with conical processes surrounded by one row of areolae (Figs. 27a-h, 28a-f). In SEM, multiple rings of tight annulation on the entire process surface were visible (Fig. 28a-b), although in some processes, annulation was present only in the upper portion of the process (Fig. 28e-f) (annulation not visible in LCM because it was obscured by the eminent labyrinthine layer). The upper parts of the processes are smooth and not covered with granulation ( Fig. 28e-f). The labyrinthine layer between the process walls was present and usually visible only in the bottom part of egg processes as one to three rows of meshes and some bubbles scattered randomly on the remaining process (Fig. 27a-c). These rings of reticulation can also sometimes be visible even with SEM (Fig.  28c, e). Sometimes well-developed reticulation with circular/ellipsoidal meshes is observed in the entire process but still with clearly visible places deprived of the reticulation (Fig. 27d). The upper part of the process surface is often elongated into short flexible apices that can be occasionally broken and that can sometimes be bifurcated and have bubble-like structures (Figs. 27e-h, 28a, e-f). The base of the processes extends into the six (only sometimes five) arms that form areolae rims (Figs. 27a-d, 28a-f). Each process is surrounded by six (only sometimes five) hexagonal areolae (Figs. 27a-d, 28a-d), which are occasionally falsely subdivided in the middle into two areolae by a thin thickening perpendicular to the process base (Figs. 27a-c, 28a, c). Areolae rims (walls) are usually thick and flat (Fig. 28a, c-d), but sometimes they can also be very thin (Fig. 28b), with the labyrinthine layer inside the rims visible as bubbles under LCM (Fig. 27a-d). The areola surface has wrinkles that are faintly visible under LCM (Fig. 27a-d) but clearly visible under SEM (Fig.  28a-f). Micropores are present within the areolae, but they are distributed only around the areolae rims and usually absent in the central part of the areola (Fig. 28b-f).

Reproduction
The species is dioecious. Spermathecae in females as well as testes in males were found to be filled with spermatozoa, clearly visible under PCM up to 24 hours after mounting in Hoyer's medium (Fig. 29a-b). The  species exhibits secondary sexual dimorphism in the form of clearly visible lateral gibbosities on hind legs in males (Fig. 29b).

Phenotypic differential diagnosis
By having the processes surrounded by 5-6 areolae, it resembles four other species of the Macrobiotus pallarii complex out of which two are newly described in this study. By the morphology of the animals and eggs, this species can be differentiated from the following: Macrobiotus pallarii: by lunulae IV being dentate (the lunulae are only faintly crenulated in M. pallarii), by the absence of two lateral patches of dense granulation between legs III and IV (dense granulation patches between legs III and IV are present in M. pallarii; see Fig. 1), by the presence of a sparse granulation connecting the dense granulation patches between legs III and IV extending posteriorly to the granulation on legs IV (sparse granulation does not extend posteriorly to the granulation on legs IV in M. pallarii; see Fig. 1) and by and the absence of granulation on the egg processes tips (granulation is present in M. pallarii.; character visible only under SEM). Macrobiotus pseudopallarii sp. nov.: by lunulae IV being dentate (the lunulae are only gently dentate in M. pseudopallarii sp. nov.), by the absence of two lateral patches of dense granulation between legs III and IV (dense granulation patches between

Genotypic differential diagnosis
Interspecific genetic p-distances between M. ripperi sp. nov. and other species of the M. pallarii complex are as follows:  Macrobiotus margoae Stec, Vecchi & Bartels, sp. nov.

Description of the new species
Animals (measurements and statistics in Table 12): In live animals, body almost transparent in smaller specimens and whitish in larger animals; transparent after fixation in Hoyer's medium (Fig. 30). Eyes present in live animals and after fixation in Hoyer's medium. Small round and oval cuticular pores (0.7-1.7 μm in diameter), visible under both LCM and SEM, scattered randomly throughout the entire body (Figs. 31a-d, 32a-d). Patches of fine granulation on the external surface of legs I-III  (Fig. 2, 31b, d) and SEM (Fig. 32b, d). A pulvinus is present on the internal surface of legs I-III (Figs. 31c, 32c). In addition to the typical patches of leg granulation, other types of cuticular granulation are absent.
Claws slender, of the hufelandi type. Primary branches with distinct accessory points, a long common tract, and an evident stalk connecting the claw to the lunula (Fig. 33a-d).
Lunulae on legs I-III smooth, whereas those on legs IV clearly dentate (Fig. 33a-d). Dark areas under each claw on legs I-III faintly visible under LCM (Fig. 33a). Paired muscle attachments and internal strengthening above them on legs I-III were often visible under both LCM (Fig. 24a) and SEM, whereas the horseshoe-shaped structure connecting anterior and posterior claws IV was visible only with LCM (Fig. 33b).
Mouth antero-ventral. Buccal apparatus of the Macrobiotus type (Fig. 35a), with the ventral lamina and ten peribuccal lamellae (Fig. 35a-b). The oral cavity armature was composed of three bands of teeth, from which only the second and third bands were always clearly visible under LCM (Fig. 34b-c), whereas the first band was only visible under SEM ( Fig. 35a-b). The first band of teeth is composed of numerous small teeth visible as globular cones with SEM ( Fig. 35a-b), arranged in several rows, and situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae. The second band of teeth is situated between the ring fold and the third band of teeth and comprises 3-4 rows of teeth visible with LCM as granules (Fig. 34b-c) and with SEM as cones ( Fig. 35a-b) but larger than those in the first band. The posterior row of teeth within the second band seems to comprise larger teeth than the previous anterior rows (Fig. 34b-c). The teeth of the third band are located within the posterior portion of the oral cavity, between the second band of teeth and the buccal tube opening (Figs. 34b-c, 35a-b). The third band of teeth is divided into the dorsal and ventral portions. Under both LCM and SEM, the dorsal teeth are seen as three distinct transverse ridges, whereas the ventral teeth appear as two separate lateral transverse ridges, between which one large tooth (sometimes circular in LCM) is visible (Figs. 34b-c, 35a-b). In SEM, only teeth of the dorsal portion in the third band have clearly indented margins (Fig. 35a). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids (2<1) and a microplacoid positioned close to them (i.e., the distance between the second macroplacoid and the microplacoid is shorter than the microplacoid length; Fig. 34d-e). The first macroplacoid is anteriorly narrowed and constricted   in the middle, whereas the second has a subterminal constriction (Fig. 34d-e).
Eggs (measurements and statistics in Table 13): Laid freely, white, spherical with conical processes surrounded by one row of areolae (Figs. 36a-h, 37a-f). In SEM, multiple rings of tight annulation on the entire process surface were visible (Fig. 37a-f) (annulation not visible in LCM because it was obscured by the eminent labyrinthine layer). The upper parts of the processes are smooth and not covered by granulation (Fig. 37c-f). The labyrinthine layer between the process walls was present and usually visible only as small circular bubbles scattered randomly on the process (Fig. 36a-b). Rarely, these bubbles are almost not visible (Fig. 36c-d). The upper part of the process is often elongated into short flexible apices that can be occasionally broken and that can sometimes be bifurcated and have bubble-like structures (Figs. 36e-h, 37c-f). The base of the processes extends

Reproduction
The species is dioecious. Spermathecae in females as well as testes in males have been found to be filled with spermatozoa, clearly visible under PCM up to 24 hours after mounting in Hoyer's medium ( Fig. 38a-b). The species exhibits secondary sexual dimorphism in the form of clearly visible lateral gibbosities on hind legs in males (Fig. 38b).

Phenotypic differential diagnosis
By having the processes surrounded by 5-6 areolae, it resembles four other species of the Macrobiotus pallarii complex out of which two are newly described in this study. By the morphology of the animals and eggs, this species can be differentiated from the following:

Dichotomous diagnostic key to species of the Macrobiotus pallarii complex
To facilitate species identification, we provide a dichotomous key to the valid species of the M. pallarii complex. When differentiating these species, we stress the need for analyzing the largest possible number of individuals and eggs, as some characters, such as egg process reticulation, show intraspecific variability (see the Taxonomic Account above), and some more conservative characters, such as dorso-caudal granulation, may not always be easily detected due to suboptimal orientation of specimens on slides or/and microscope quality.

Discussion
Macrobiotus pallarii was described from Italy in 1954, and it has been reported from Europe and North America [25,89,90]. We recovered a population of this species from the type locality (Silvana Mansio, Cosenza, Italy) in an environment matching the original species description (moss on rock on the ground in a sparse forest [22]). The assignment of population IT.337 to M. pallarii was confirmed by a comparison with the type series from the Maucci collection (Natural History Museum of Verona). The analysis of the topotypes of M. pallarii, with four distinct populations (that would have been attributed to the same species prior to this study), showed that four pseudocryptic species can be distinguished from morphological, morphometric, and genetic viewpoints. Due to the presence of these morphological and genetic differences, we described them as three new species (see the Taxonomic Account above).

Species delineation in the Macrobiotus pallarii complex
Congruent with the results of species delimitation of different populations of the Paramacrobiotus areolatus group by [7], the COI showed higher divergence than ITS-2, and a strikingly similar pattern was found for a pair of closely related P. areolatus group populations (IT.048 and PT.006 in [7]) with contrasting species delimitation results based on the two analyzed markers (distinct species according to COI, same species according to ITS-2). Similar differences in species delimitation based on COI and ITS-2 markers were obtained for some species in the genus Milnesium, although in contrast to the present study, the precedence was given to ITS-2, as no phenotypic differences between units delineated with COI were detected [12,91]. Thus, it seems that there is no simple rule to decide which genetic marker is better suited for species delineation in tardigrades. Rather, both phenotypic and genetic traits should be used in tandem in taxonomic analyses.

Species diversity in the Macrobiotus pallarii complex
Macrobiotus pallarii, M. ragonesei, M. caymanensis, and the three new species form the M. pallarii complex (defined by [19] as having cone-shaped egg processes separated by one row of areolation). The identification of pseudocryptic species in the M. pallarii complex reopens the question about the identity of M. aviglianae, which has been synonymised with M. pallarii by [24]. The analysis of the M. aviglianae-type material (slides CT1954, CT1955, CT1966, CT4051 and CT4485 of the Maucci collection) showed that it was not in good condition. Thus, it is not possible to verify whether M. aviglianae does not differ from M. pallarii in any subtle traits, such as in the pattern of dorso-caudal granulation or granulation on the tips of egg processes. Moreover, it is also not possible to test whether M. aviglianae represents an independent phyletic lineage or if it falls within the intraspecific genetic variability of M. pallarii. Thus, only the discovery of a new population of M. aviglianae in its type locality will allow for a proper test of its synonymy with M. pallarii. The morphometric characters of the M. caymanensis type overlap with those of the newly described M. margoae sp. nov., with which it also shares a similar appearance of the egg process wall. Thus, it is probable that these two species are phylogenetically close; however, it will be possible to test this hypothesis only after genetic data for M. caymanensis are obtained. For M. ragonesei, detailed morphometric data are not available, precluding its inclusion in the PCA, but based on morphology alone, it is clear that the species deviates from all other known members of the complex by a much higher number of areolae around each egg process (11-12 vs. 5-6 in the remaining species). Macrobiotus ragonesei was collected only once in the type locality in the Democratic Republic of Congo (Central Africa); however, it is possible that the population collected and identified as Macrobiotus cf. pallarii by [92] from Swaziland (South Africa) represents M. ragonesei or a related species because eggs from the Congolese population resembled those of M. ragonesei ("processes are rounded rather than tapering, and the number of areolae surrounding them is ten to twelve rather than eight or nine" [92]). Another species, Macrobiotus insularis Pilato, 2006 [93], known only from the Andaman Islands in the Indian Ocean, is currently assigned to the Macrobiotus polyopus group [93]. However, because of its similarity with the species of the M. pallarii complex, it would not be surprising if after a reanalysis of the type material, the species is moved to the latter complex. Additionally, M. caymanensis was originally described as a member of the M. polyopus group, which suggests that both groups have not been sufficiently differentiated and doubts about the assignment to one or the other group are possible. In our opinion, the M. polyopus group can be differentiated from the M. pallarii complex by the morphology of egg processes and areola walls (rims) (Fig. 39a). Specifically, process base width and interprocess distances were similar in the M. polyopus group (Fig. 39b), whereas process bases were much wider than the distance between neighboring processes in the M. pallarii complex (Fig. 37b). Moreover, areolae are always delimited by high and thin walls in the M. polyopus group (usually nearly half the process height, Fig. 39c-d), whereas areolae walls are low and often thick, with the labyrinthine layer visible in LCM, in the M. pallarii complex (Fig.  37c-d). Thus, currently, there are six confirmed species in the M. pallarii complex, with further potential members: M. insularis (pending the verification of the type material) and M. aviglianae (pending an examination of a topotypic population). However, the number of species constituting the complex is likely to be much higher, as has been shown in other eutardigrade groups when tools of integrative taxonomy and large-scale sampling have been employed, for example, in the genus Milnesium [94], the genera Astatumen and Platicrista [95], the Ramazzottius oberhaeuseri complex [38], the genus Macrobiotus (e.g., [19]), the Paramacrobiotus richtersi and areolatus complexes ( [5,7], respectively), or in the genus Richtersius [6,13].

Biogeography of the Macrobiotus pallarii complex
Species of the Macrobiotus pallarii complex are not commonly found, but the complex has a wide, possibly global distribution. Although the majority of records are concentrated in Europe (Finland, Greece, Hungary, Italy, Montenegro, Norway, Poland, Republic of Bulgaria, Russian Federation, Turkey, Yugoslavia [25,82,[96][97][98]), there are also reports from Asia (North Korea, Russian Federation [25]), North America (United States, Cayman Islands [28,[86][87][88]), and Africa (Democratic Republic of the Congo, Swaziland [27,92]). In light of our findings, all these records except the type localities should be considered dubious and be designated as species of the Macrobiotus pallarii complex, i.e., as Macrobiotus aff. pallarii spp., keeping open the possibility they could represent yet undescribed new species.

The importance of disentangling pseudocryptic species complexes
Thanks to the use of integrative taxonomy, pseudocryptic and cryptic tardigrade species complexes are being detected with increasing frequency (chronologically [6,7,11,13,14,19,38,[99][100][101][102][103]). Other than for purely taxonomic purposes, acknowledging and discerning such entities may significantly impact studies concerning biodiversity, biogeography, ecology or physiology. Specifically, biodiversity studies and taxonomic checklists need precise taxonomic identifications to be comparable across different geographical areas. However, as noted, e.g., by [82], numerous identifications cannot be confirmed due to the presence of (pseudo) cryptic species complexes. Furthermore, biogeographical studies supported by molecular data on tardigrades are still scarce. In one of them, Gąsiorek et al. [104] showed evidence suggesting allopatric speciation in a complex of tardigrades (Echiniscus virginicus complex) that are almost indistinguishable morphologically but genetically divergent, underlining the need for genetic verification of faunistic records to draw sound biogeographic conclusions. Regarding ecology, for example, Faurby et al. [105] found that Echiniscoides species have different geographical distributions and thus also various ecological requirements despite being very closely related. The M. pallarii complex also seems to be ecologically and physiologically diversified, as M. ripperi sp. nov. performs very well in laboratory culture (D.S. and M.V. personal observation), whereas the other examined species (M. pallarii, M. pseudopallarii sp. nov., M. margoae sp. nov.), while able to reproduce in culture, they did not thrive. Whether these differences in reproductive success under uniform laboratory conditions are a result of natural variability in the physiology or different levels of alignment of the ecology of these species and laboratory conditions (i.e., the similarity of laboratory conditions, such as temperature, fodder type or photoperiod, to the conditions to which the species is adapted in nature) has yet to be tested. Nevertheless, in both cases, species identity, despite the striking morphological similarity, is an important variable in the equation and cannot be ignored. Differences in ecological and physiological traits in species that exhibit very little morphological differentiation have also been demonstrated for cryptic species complexes in other taxa (e.g., bryozoans, diatoms and butterflies [106][107][108]). The presence of cryptic species that show physiological and life history divergence can also influence laboratory studies when the adopted model species belong to one of those complexes. For example, studies on the model chordate sea squirt Ciona intestinalis (Linnaeus, 1767 ) [109] assumed that all worldwide natural Fig. 39 Macrobiotus aff. polyopus species from Papua New Guineaeggs seen with SEM: aentire egg; b-ddetails of the egg surface between the processes, areolation and egg processes. Scale bars in μm populations belong to a single species [110]. However, it was later demonstrated that cryptic species that produce sterile hybrid offspring are present [111]. Acknowledgment of the presence of cryptic species in C. intestinalis turned out to be essential for experimental design and data management [110], as, for example, new laboratory strains could not be obtained by crossing the cryptic species. There are also similar examples concerning tardigrades.
Specifically, individuals attributed to Paramacrobiotus richtersi were used by [112] to test the effects of a lower earth orbit environment on tardigrade physiology and reproduction. However, it was later revealed that individuals of that population belong to a different cryptic species of the P. richtersi complex, Paramacrobiotus spatialis Guidetti et al., 2019 [5]. As species in the P. richtersi complex show different geographic distributions [5] and are likely to have different environmental requirements and tolerances, their misidentification (due to the recognition of the species complex years later) could affect the replicability of the 2011 study if the experiment was replicated with another species of the complex identified as "P. richtersi". Similarly, more tardigrade species used in experimental studies were found to belong to cryptic (or pseudocryptic) species complexes and misidentified for the nominal species of the complex, for example Hypsibius exemplaris was misidentified as H. dujardini [113], Milnesium inceptum was misidentified as M. tardigradum [114] and Richtersius sp. 4 was misidentified as R. coronifer [6]. Importantly, while discrepancies stemming from results obtained with different (pseudo) cryptic species could be a problem when not recognized, such species may also provide a useful tool to investigate tardigrade speciation and how isolation between such morphologically similar species is maintained by using the tools of experimental taxonomy, such as interpopulation crosses [7] and mate choice assays, as well as omics techniques.

Conclusions
Integrative taxonomy revealed that multiple pseudocryptic species, each probably with a limited geographic distribution, are present under what was once considered a single cosmopolitan species, "Macrobiotus pallarii". Species of the Macrobiotus pallarii complex can be differentiated principally by dorsal granulation and egg ornamentation but also statistically by morphometric traits. Other than for purely taxonomic reasons, disentangling tardigrade cryptic and pseudocryptic species complexes is fundamental for the proper interpretation of biogeographical and ecological studies and the replicability of experimental works. Due to the rising popularity of tardigrade species as laboratory models (e.g., [17,[115][116][117][118][119][120][121][122][123][124]), the presence of cryptic or pseudocryptic complexes must be taken into account to avoid incongruent results stemming from taxonomic misidentifications.