Integrative description of Macrobiotus canaricus sp. nov. with notes on M. recens (Eutardigrada: Macrobiotidae)

. In this paper we describe Macrobiotus canaricus sp. nov., a new tardigrade species of the Macrobiotus hufelandi group from the Canary Islands. Moreover, with the use of DNA sequencing, we confirm that Macrobiotus recens Cuénot, 1932 represents the hufelandi group, even though eggs laid by this species do not exhibit the typical hufelandi group morphology. Our study is based on both classical taxonomic methods that include morphological and morphometric analyses conducted with the use of light and scanning electron microscopy, and on the analysis of nucleotide sequences of four molecular markers (three nuclear: 18S rRNA, 28S rRNA, ITS-2, and one mitochondrial: COI). Our analyses revealed that M. canaricus sp. nov. is most similar to Macrobiotus almadai Fontoura et al. , 2008 from the Archipelago of the Azores, from which it differs by the absence of granulation patches on the external and internal surfaces of legs I–III as well as by the absence of a cuticular pore in the centre of the external patch on legs I–III. Molecular sequences allowed us to pinpoint the phylogenetic positions of M. canaricus sp. nov. and M. recens within the M. hufelandi group.

All figures were assembled in Corel Photo-Paint X6, v. 16.4.1.1281. For structures that could not be satisfactorily focused in a single photograph, a stack of 2-8 images was taken with an equidistance of ca 0.2 μm and assembled manually into a single deep-focus image.

Morphometrics and morphological nomenclature
All measurements are given in micrometres (μm). Sample size was adjusted following recommendations by Stec et al. (2016b). 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 egg shell morphology follows Michalczyk & Kaczmarek (2003) and Kaczmarek & Michalczyk (2017b). Macroplacoid length sequence is given according to Kaczmarek et al. (2014). Buccal tube length and the level of the stylet support insertion point were measured according to Pilato (1981). The pt index is the ratio of the length of a given structure to the length of the buccal tube expressed as a percentage (Pilato 1981). All other measurements and nomenclature are given in accordance with Kaczmarek & Michalczyk (2017b). In brief, the buccal tube width was measured as the external and internal diameter at the level of the stylet support insertion point. Lengths of the claw branches were measured from the base of the claw (i.e., excluding the lunula) to the top of the branch, including accessory points. Distance between egg processes was measured as the shortest line connecting base edges of the two closest processes. Morphometric data were handled using the 'Parachela' v. 1.3 template available from the Tardigrada Register (Michalczyk & Kaczmarek 2013). Tardigrade taxonomy follows Bertolani et al. (2014).

Comparative material
The taxonomic key for the hufelandi group provided by Kaczmarek & Michalczyk (2017b) was used to determine whether the isolated species had previously been described. After one species could not be identified with the key, we compared it with the original description of the most similar species of the hufelandi group, Macrobiotus almadai Fontoura et al., 2008. Additionally, we also compared our new species with three paratypes and one egg of M. almadai kindly loaned to us by Professor Paulo Fontoura. Moreover, thanks to the courtesy of Matteo Vecchi, we analysed photomicrographs of animals and eggs identified as 'M. recens' by Maucci (1979)  *Eggs were taken from the culture and incubated individually until hatching, then the juveniles were used for DNA sequencing whereas egg chorions (isogenophores) were mounted on microscope slides in Hoyer's medium.

STEC D. et al., Tardigrades from the Canary Islands
is outdated and, in our opinion, an accurate identification of the species is not possible until a modern redescription is available (see Discussion for more details). Therefore, we identified our population as M. cf. recens. Nevertheless, given that our population fits the description of M. recens, it has to represent at least a closely related species and, as such, it can be used to estimate the phylogenetic position of M. recens s.str.

Genotyping
The DNA was extracted from individual animals following the Chelex ® 100 resin (Bio-Rad) extraction method of Casquet et al. (2012) with modifications described in detail in Stec et al. (2015). We sequenced four DNA fragments, three nuclear (18S rRNA, 28S rRNA, ITS-2) and one mitochondrial (COI). All fragments were amplified and sequenced according to the protocols described in Stec et al. (2015); primers and original references for specific PCR programs are listed in Table 2. Sequencing products were read with the ABI 3130xl sequencer at the Molecular Ecology Lab, Institute of Environmental Sciences of the Jagiellonian University, Kraków, Poland. Sequences were processed in BioEdit v. 7.2.5 (Hall 1999) and submitted to GenBank.

Comparative molecular analysis
For molecular comparisons, all published sequences of the four above-mentioned markers for species of the hufelandi complex were downloaded from GenBank (listed in Table 3). Additionally, we also sequenced the four DNA fragments for a new population of Macrobiotus macrocalix  collected in Łękawica, southern Poland by DS in October 2015 (see Table 1 for sample details). The sequences were aligned using the default settings (in the case of COI) and the Q-INS-I method (in the case of ribosomal markers: 18S rRNA, 28S rRNA and ITS-2) of MAFFT v. 7 (Katoh et al. 2002;Katoh & Toh 2008) and manually checked against non-conservative alignments in BioEdit. Then, the aligned sequences were trimmed to 763 (18S rRNA), 712 (28S rRNA), 317 (ITS-2) and 618 (COI), bp. All COI sequences were translated into protein sequences in MEGA7 v. 7.0 (Kumar et al. 2016) to check against pseudogenes. According to the recommendation by Srivathsan & Meier (2012), uncorrected pairwise distances were calculated using MEGA7 instead of genetic distances corrected by Kimura 2 parameter model (K2P).

Data deposition
Raw morphometric measurements underlying the description of Macrobiotus canaricus sp. nov. are given in Supplementary Materials (SM.1) and are deposited in the Tardigrada Register (Michalczyk & Kaczmarek 2013) under www.tardigrada.net/register, whereas raw morphometric data underlying the description of the population of M. cf. recens are given in Supplementary Materials (SM.2). The DNA

Phylogenetic analysis
To verify the phylogenetic positions of the new species and M. cf. recens, phylogenetic trees were constructed using (1) all macrobiotid 18S rRNA sequences available from GenBank, (2) concatenated 18S rRNA+28S rRNA+ITS-2+COI sequences of macrobiotid species for which at least three of these markers were sequenced, and (3) all published M. hufelandi group COI sequences. In addition to the sequences of the hufelandi group listed in Table 3 The sequences were aligned using the default settings (in the case of COI) and with the Q-INS-I method (in the case of ribosomal markers: 18S rRNA, 28S rRNA and ITS-2) of MAFFT v. 7 (Katoh et al. 2002;Katoh & Toh 2008) and then edited and checked manually in BioEdit. The alignments of 18S rRNA and COI sequences were trimmed to 739 bp and 618 bp, respectively, whereas the aligned sequences that were used to construct the concatenated data matrix were trimmed to: 728 bp (18S rRNA), 754 bp (28S rRNA), 570 bp (ITS-2) and 621 bp (COI). The sequences were concatenated in SequenceMatrix (Vaidya et al. 2011). The concatenated data matrix comprises species for which at least three of the aforementioned molecular markers are available. This resulted in only two gaps within the matrix: for the 18S rRNA sequence of Me. insanis (sequence too short) and the ITS-2 sequence of Me. ethiopicus (sequence not available). Using PartitionFinder v. 2.1.1 (Lanfear et al. 2016) under the Bayesian Information Criterion (BIC), the best scheme of partitioning and substitution models were chosen for posterior phylogenetic analysis. We ran the analysis to test all possible models implemented in the program. As the COI is a protein coding gene, before partitioning, we divided our alignments of this marker into 3 data blocks constituting separated three codon positions. As best-fit partitioning scheme, PartitionFinder always suggested the retention of all predefined partitions separately. See Table 4 for specific substitution models suggested for all tested data sets and partitions.
Bayesian inference (BI) marginal posterior probabilities were calculated for both data sets using MrBayes v. 3.2 (Ronquist & Huelsenbeck 2003). Random starting trees were used and the analysis was run for ten million generations, sampling the Markov chain every 1000 generations. An average standard deviation of split frequencies of < 0.01 was used as a guide to ensure the two independent analyses had converged. The program Tracer v. 1.7 (Rambaut et al. 2018) was then used to ensure Markov chains had reached stationarity and to determine the correct 'burn-in' for the analysis, which was the first 10% of generations.
The ESS values were greater than 200 and a consensus tree was obtained after summarising the resulting topologies and discarding the 'burn-in'.

Etymology
The specific epithet refers to the Canary Islands, the place where the new species was found.

Description
Animals (measurements and statistics in Table 5) Body white in adults, after fixation in Hoyer's medium transparent (Fig. 1A). Eyes present both in live animals and in specimens mounted in Hoyer's medium. Round and oval pores (0.4-0.7 μm in diameter), visible under PCM and SEM, scattered randomly on entire body cuticle ( Fig Table 4. The best scheme of partitioning and substitution models chosen for posterior phylogenetic analysis using PartitionFinder v. 2.1.1 under the Bayesian Information Criterion (BIC). The analyses were run to test all possible models implemented in the program.

Data set
Substitution model for the given partition 18S 28S Mouth antero-ventral. Bucco-pharyngeal apparatus of the Macrobiotus type, with the ventral lamina and ten small peribuccal lamellae followed by six buccal sensory lobes ( Fig. 4A-C). Under PCM, the oral cavity armature is of the maculatus type, i.e., only the third band of teeth is visible (Fig. 4A).  (2018) Under SEM, the oral cavity is always composed of three bands of teeth ( Fig. 4B-C). The first band of teeth is composed of numerous extremely small cones arranged in one or two rows, situated anteriorly in the oral cavity, on the basal part of the peribuccal lamellae ( Fig. 4B-C, filled arrowhead). The second band of teeth is situated between the ring fold and the third band of teeth and consists of cones, clearly larger than those of the first band ( Fig. 4B-C, empty arrowhead). 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 (Fig. 4B-C). The third band of teeth is discontinuous and divided into a dorsal and a ventral portion. Under PCM, the dorsal teeth form a transversal ridge weakly divided into three teeth, whereas the ventral teeth appear as two separate lateral transverse ridges between which a roundish median tooth is visible (Fig. 4A). Under SEM, the dorsal teeth are divided into three separate teeth: one median and two lateral, the median tooth has a slightly serrated edge (Fig. 4B). The ventral teeth are also separated into one median and two lateral teeth (Fig. 4C). The medio-ventral tooth is much smaller than the medio-dorsal tooth ( Fig. 4B-C). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids and a small microplacoid (Fig. 4A). The first and the second macroplacoids have a fine central and a subterminal constriction, respectively. The macroplacoid length sequence is 2 < 1.  Eggs (measurements and statistics in Table 6) Laid freely, white, spherical or slightly oval (Figs 6A-B, 7A). The surface between processes of the hufelandi type, i.e., covered by a reticulum with very thin walls (Figs 6D-E, 7A-F). Peribasal meshes slightly larger and with slightly thicker walls compared to interbasal meshes (Figs 6D-E, 7B-F). The mesh diameter is always larger then mesh walls and nodes/knots (Figs 6D-E, 7B-F). The meshes are 0.3-1.0 μm in diameter, polygonal but with rounded edges. Under SEM, meshes deep and empty inside (Fig. 7C-F). Processes in the shape of inverted goblets with concave conical trunks and well-defined terminal discs (Figs 6C-F, 7A-F). Terminal discs strongly serrated, with a concave central area (Figs 6C-F, 7B-F). Sparse ultragranulation on the edges of terminal discs visible only under SEM (Fig. 7E-F). Three to five microgranules (0.25-0.30 μm in diameter), covered with ultragranulation, present in the centre of the terminal disc (visible only under SEM; Fig. 7B-F, empty arrowheads).

Reproductive mode
The examined population is dioecious (gonochoristic). Males were identified using aceto-orcein staining, which revealed testicles filled with spermatozoa. However, no morphological secondary sexual dimorphism, such as gibbosities on hind legs in males, was identified.

DNA sequences
We obtained sequences for all four of the above-mentioned molecular markers. The two conservative nuclear markers (18S rRNA, 28S rRNA) were represented by single haplotypes, whereas ITS-2 and COI exhibited three and two haplotypes, respectively. The p-genetic distance between the ITS-2 haplotypes ranged from 0.5 to 1.1% and between COI haplotypes it was equal to 1.3%. The 18S rRNA sequence (GenBank: MH063925) was 1033 bp long. The 28S rRNA sequence (GenBank: MH063934) was 721 bp long. The ITS-2 haplotypes 1-3 were 413 bp long (GenBank: MH063928, MH063929 and MH063930, respectively). The COI haplotypes 1-2 were 658 bp long (GenBank: MH057765 and MH057766, respectively).

Phenotypic differential diagnosis
By the oral cavity armature of the maculatus type and hufelandi type of egg shell ornamentation, smooth lunules under claws of all legs and granulation at least on legs IV, the new species is similar to M. almadai Fontoura et al., 2008, M. humilis Binda & Pilato, 2001, and M. rawsoni Horning et al., 1978, but can be differentiated specifically from: Macrobiotus almadai, known only from the Azores (Fontoura et al. 2008), by the presence of the external and the internal patch of granulation on legs I-III (legs I-III smooth in M. almadai) and by the presence of a single large pore in the centre of the external patch on legs I-III (occasionally, regular cuticular pores may be present on some legs, but such pores are small and never present on all legs in the same place in M. almadai).
Macrobiotus humilis, reported only from its type locality in Sri Lanka (Binda & Pilato 2001 Macrobiotus rawsoni, known only from its type locality in New Zealand Kaczmarek & Michalczyk 2017a), by the presence of granulation on all legs (granulation present only on legs IV in M. rawsoni), the presence of a subterminal constriction in the second macroplacoid (second macroplacoid without constrictions in M. rawsoni), the absence of cuticular bars under the claws on  M. rawsoni), a different morphology of reticulation on the egg surface between processes (several lines of mesh between neighbouring egg processes in the new species vs two lines of mesh between neighbouring egg processes in M. rawsoni) and by a smaller mesh size in the chorion reticulum (0.3-1.0 μm in diameter in the new species vs 1.8-2.5 μm in diameter in M. rawsoni).

Description of the population from Gran Canaria
Animals (measurements and statistics in Table 7) Body white in juveniles and slightly yellowish in adults, after fixation in Hoyer's medium transparent (Fig. 8A). Eyes present in live animals and in specimens mounted in Hoyer's medium. Elliptical and sometimes roundish pores (1.0-1.8 μm in diameter), visible under PCM and SEM, scattered randomly on entire body cuticle (Fig. 8B-E), including the external and internal surface of all legs (Fig. 9A-I). Inside pores several granules, visible only under SEM, always present (Fig. 8E). Granulation patch on external and internal surfaces of legs I-III present (Fig. 9A-E). Single pore present at centre of each external granulation patch (Fig. 9A-C). Granulation patch on external surface larger and more distinct than the one on internal surface (Fig. 9A-E). Faint cuticular fold present on external surface of legs I-III just above claws (Fig. 9A-B, empty arrowhead), whereas on internal surface of legs I-III there is a cuticular bulge resembling pulvinus (Fig. 9D-E, filled arrowhead). Both external fold and internal bulge visible only if legs are fully extended and correctly oriented on slide (particularly cuticular fold above claws). Granulation on legs IV always clearly visible and consists of two granulation patches: the distal STEC D. et al., Tardigrades from the Canary Islands patch with densely distributed granules situated just above claws and the proximal patch being wider with more sparsely distributed granules located immediately above distal patch (Fig. 9G-I).
Mouth antero-ventral. Bucco-pharyngeal apparatus of the Macrobiotus type ( Fig. 10A-C), with ventral lamina and ten small peribuccal lamellae followed by six buccal sensory lobes. Under PCM, oral cavity armature of the hufelandi type, i.e., with all three bands of teeth always visible (Fig. 10B-C). First band of teeth composed of numerous very small cones arranged in four to six rows situated anteriorly in oral cavity, just behind bases of peribuccal lamellae (Figs 10B-C, 11A-B, filled arrowhead). Second band of teeth situated between ring fold and third band of teeth and comprises 4-5 rows of small cones, slightly larger than those of first band (Figs 10B-C, 11A-B, empty arrowhead). Teeth of the third band located within posterior portion of oral cavity, between the second band of teeth and buccal tube  (2018) opening (Figs 10B-C, 11A-B). Third band of teeth discontinuous and divided into dorsal and ventral portions. Under PCM, dorsal teeth seen as three distinct transversal ridges, whereas ventral teeth appear as two separate lateral transversal ridges and a roundish median tooth (Fig. 10B-C). Under SEM, both dorsal and ventral teeth clearly distinct (Fig.11A-B). Medio-ventral tooth rarely divided into two or three smaller teeth (Fig. 11B). Under SEM, margins of dorsal and latero-ventral teeth slightly serrated (Fig. 11A-B). Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids and

Reproductive mode
The examined population is dioecious (gonochoristic). Males were identified using aceto-orcein staining, which revealed testicles filled with spermatozoa. However, no morphological secondary sexual dimorphism, such as gibbosities on hind legs in males, was identified.

DNA sequences
We obtained sequences for all four of the above-mentioned molecular markers. The two conservative nuclear markers (18S rRNA, 28S rRNA) were represented by single haplotypes, whereas both ITS-2 and COI exhibited two haplotypes. The p-genetic distance between the ITS-2 as well as between the COI haplotypes was 1.1%. The 18S rRNA sequence (GenBank: MH063927) was 1033 bp long, the 28S rRNA sequence (GenBank: MH063936) was 725 bp long, the ITS-2 haplotype 1 and 2 sequences (GenBank: MH063932 and MH063933, respectively) were 420 bp long; the COI haplotype 1 and 2 sequences (GenBank: MH057768 and MH057769, respectively) were 658 bp long.

Genotypic differential diagnosis
The ranges of uncorrected genetic p-distances between the new species and species of the Macrobiotus hufelandi complex, for which sequences are available from GenBank, are as follows:

Molecular phylogeny
Phylogenetic analyses conducted on macrobiotid 18S rRNA sequences as well as on the concatenated macrobiotid data set unambiguously confirmed that the two studied species represent the M. hufelandi group (Figs 15-16). The phylogeny based on the COI sequences of the hufelandi group also corroborated these results, since none of the two species were recovered external to the species of the hufelandi group (Fig. 17). In all analyses, two clades within the hufelandi group were present, although the species composition varied slightly between phylogenies based on different markers. One clade grouped exclusively species that exhibit modified egg processes (M. paulinae Stec et al., 2015, M. polypiformis, M. papei, M. shonaicus Stec et al., 2018a, M. scoticus Stec et al., 2017band M. kristenseni Guidetti, Peluffo, Rocha, Cesari & Moly de Peluffo, 2013; 'the kristenseni clade' henceforth. The other clade comprised mostly species with typical inverted goblet-shaped egg processes ('the hufelandi clade' hereafter). In contrast to our predictions, M. cf. recens, with its atypical egg processes, was always embedded within the hufelandi clade. The two clades were well supported in phylogenies based on the concatenated data set and on COI sequences, but weakly supported in the 18S rRNA tree (Figs 15-17). Moreover, in the 18S rRNA analysis the kristenseni clade, in addition to the majority of species with modified egg processes, comprised M. sapiens Binda & Pilato, 1984 (DQ839601) and undetermined species of the M. hufelandi group (HQ604971), of which at least the first species exhibits the typical egg morphology. In contrast to other analyses, the 18S rRNA phylogeny recovered a clade, with X. pseudohufelandi (Iharos, 1966) and M. polonicus Pilato, Kaczmarek, Michalczyk & Lisi, 2003, that was in a sister relationship to all other species of the hufelandi group, suggesting that the hufelandi group is polyphyletic or that Xerobiotus belongs to the hufelandi group (Fig. 15).

Discussion
Thanks to the detailed morphological and molecular analyses, we were able to describe a new species M. canaricus sp. nov. and characterise a population of M. cf. recens, both collected on the Canary Islands. Moreover, by the use of phylogenetic inference, we confirmed the affinity of both studied species with the M. hufelandi group.
Macrobiotus recens was described in 1932 by Lucien Cuénot, but soon after, some researchers started questioning the status of this species and classified it as 'M. hufelandi forma recens' (Marcus 1936;Ramazzotti 1945Ramazzotti , 1962Ramazzotti , 1972Rudescu 1946;Grigarick et al. 1973). However, Maucci (1979) analysed several Portuguese populations he identified as M. recens morphologically and confirmed that the taxon is indeed a 'bona species'. However, it should be noted that the locus typicus for M. recens is La Tardière, a commune in the Vendée department, in the Pays de la Loire region in western France. Thus, similarly to our study, the comparison presented by Maucci (1979) has to be treated with caution, as it may be based on a related species rather than M. recens s.str. Since the original description is outdated and incomplete, a confident identification of M. recens is currently not possible. Although Pilato & Bertolani (2004) supported the identification of the Portuguese populations studied by Maucci (1979)  , then this study will be an addition to our poor knowledge of the intraspecific variability in tardigrades. If, however, the current record represents a new species, then, with the data already presented in our study, only a new name will have to be proposed to erect a new species.
Recently, Stec et al. (2018a) showed two well supported COI lineages within the M. hufelandi group: one clade grouping species with hufelandi-type egg processes in the shape of inverted goblets, and the other with modified egg process (conical processes or processes with filaments growing out of terminal discs). Thus, M. cf. recens, with its conical egg processes, should be expected to cluster with species with modified egg processes. However, in contrast to this prediction, M. cf. recens is embedded within the clade with species exhibiting typical egg morphology in our COI analysis (Fig. 17). Moreover, the   (2018) same position of the species was recovered in the 18S rRNA phylogeny (Fig. 15) and in the analysis based on four concatenated markers (Fig. 16). Also, M. sapiens, with typical egg processes, should be present in the hufelandi clade, but in our analyses it clustered with species exhibiting mostly modified processes (Fig. 15). Importantly, however, the 18S rRNA sequence identified as M. sapiens (DQ839601; Schill & Steinbrück 2007) comes from Croatia, whereas the type locality of the species is in Sicily (Binda & Pilato 1984). Thus, given that there are no type/neotype sequences for M. sapiens and Schill & Steinbrück (2007) did not provide SEM photomicrographs of Croatian eggs, it is not possible to verify the identity of the DQ839601 sequence. In other words, the sequence may represent a similar species of the hufelandi group that may exhibit terminal disc filaments. Nevertheless, regardless of the phylogenetic position of M. sapiens, the affinity of M. cf. recens with species exhibiting typical egg processes shows that the morphological criterion proposed by Stec et al. (2018a) to distinguish the two clades is not Fig. 17. The Bayesian Inference (BI) phylogeny constructed from COI sequences of the species of the Macrobiotus hufelandi group. Numbers at nodes indicate the Bayesian posterior probability. Species of the hufelandi group with typical and atypical egg processes are indicated by blue and red fonts, respectively. See Table 2 for details on the species sequences used in the analysis. The outgroup is marked with grey. The scale bar represents substitutions per position. universal. Therefore, more species representing the two clades need to be sequenced to elucidate the taxonomic status of the two clades.
Another species whose phylogenetic position diverges from the predicted is M. polonicus. This species, with typical egg processes, should be embedded within the hufelandi clade. However, in the 18S rRNA phylogeny, not only does it not cluster within the hufelandi clade, but it clusters with Xerobiotus pseudohufelandi in a clade that is in a sister relationship to the entire hufelandi group. The M. polonicus + X. pseudohufelandi clade was also found to be a sister group to all other hufelandi group species by Bertolani et al. (2014). The genus Xerobiotus Bertolani & Biserov, 1996 shares a similar morphology of the buccal apparatus, egg shell and spermatozoa with species of the M. hufelandi group, but it differs from them by having strongly reduced claws. Given that the M. polonicus + X. pseudohufelandi clade contains only a single hufelandi group sequence, the clade could be a statistical artefact or possibly the 18S rRNA marker is too conservative to solve the relationships between the hufelandi group and the genus Xerobiotus. Therefore, it is crucial to analyse more populations of M. polonicus, other species of the persimilis subgroup (sensu Kaczmarek & Michalczyk 2017b) and species of the genus Xerobiotus, both in terms of the 18S rRNA marker and additional, more variable DNA fragments. If, however, more species of the hufelandi group turn out to cluster with the genus Xerobiotus, then the taxonomic status of the genus and of the hufelandi group should be reconsidered.
Although our multilocus phylogeny supports the presence of two distinct evolutionary lineages within the hufelandi group, the taxon sample size is still very low and more precise conclusions concerning the phylogeny of the group cannot currently be made. Thus, to provide more reliable conclusions, much more effort should be made to obtain multilocus molecular data linked to morphology for a larger number of species and populations. Particularly, it would be beneficial to increase the sample size, both in terms of markers as well as species, for the M. polonicus + X. pseudohufelandi clade to test whether the genus Xerobiotus and the hufelandi group are monophyletic. Another clade which would benefit from an increased sample size is the clade that comprises species with modified egg processes. Currently, the clade consists of species with two very different process morphotypes, i.e., species having processes with flexible filaments on the terminal discs and species having conical processes devoid of terminal discs. Thus, it might be possible that increased sampling could reveal the presence of two further clades that differ by the egg process morphology. Nevertheless, our findings support the previous results that egg morphology seems to be evolving faster than animal morphology, which underlines the usefulness of chorion ornamentation in the delineation of closely related species (Guidetti et al. 2013;Stec et al. 2016aStec et al. , 2017aStec et al. , 2018a.