Integrative descriptions of two new Macrobiotus species (Tardigrada, Eutardigrada, Macrobiotidae) from Mississippi (USA) and Crete (Greece)

In this paper, we describe two new Macrobiotus species from Mississippi (USA) and Crete (Greece) by means of integrative taxonomy. Detailed morphological data from light and scanning electron microscopy, as well as molecular data (sequences of four genetic markers: 18S rRNA, 28S rRNA, ITS-2 and COI), are provided in support of the descriptions of the new species. Macrobiotus annewintersae sp. nov. from Mississippi belongs to the Macrobiotus persimilis complex ( Macrobiotus clade B) and exhibits a unique egg processes morphology, similar only to Macrobiotus anemone Meyer, Domingue & Hinton, 2014, but mainly differs from that species by the presence of eyes, granulation on all legs, dentate lunulae on legs IV, and of bubble-like structures within the tentacular arms that are present on the distal portion of the egg processes. Macrobiotus rybaki sp. nov. from Crete belongs to the Macrobiotus clade A and is most similar to Macrobiotus dariae Pilato & Bertolani, 2004, Macrobiotus noemiae Roszkowska & Kaczmarek, 2019, Macrobiotus santoroi Pilato & D’Urso, 1976, and Macrobiotus serratus Bertolani, Guidi & Rebecchi, 1996, but differs from them mainly in the morphological details of its egg processes and chorion reticulation, but also by a number of morphometric characters. In light of the specific morphology of the egg processes of Macrobiotus annewintersae sp. nov. and Macrobiotus anemone , that are equipped with tentacular arms instead of proper terminal disc, we also provide an updated definition of the Macrobiotus persimilis complex.


Introduction
Tardigrades are a phylum of micrometazoans distributed worldwide, that inhabit marine and limno-terrestrial environments (Schill 2019). Currently, there are more than 1300 formally recognised tardigrade species (Guidetti and Bertolani 2005;Degma and Guidetti 2007;Degma et al. 2009Degma et al. -2020. In recent years, the number of tardigrade species described with integrative taxonomy has steadily increased (e.g., Surmacz et al. 2019;Bochnak et al. 2020;Kayastha et al. 2020;Tumanov et al. 2020a, b;Guidetti et al. 2021). The accumulation of data from such integrative studies allows at some point for broader examination of phylogenetic relationships within a larger group of organisms. This was the case for the family Macrobiotidae, one of the most speciose and diverse groups among tardigrades, which was recently extensively revised ) and which is partially in focus in this study.
Faunistic and taxonomic studies on the tardigrades of North America are numerous and both local and continental species lists have been compiled (Meyer 2013;Kaczmarek et al. 2016). It is, however, clear from new species in the USA being described (see for example Nelson et al. 2020a), that we are still far from a complete knowledge of the taxonomic diversity of tardigrades in this country. In particular, the tardigrade fauna in the state of Mississippi (USA) has been investigated only once by Hinton and Meyer (2009) who reported only 9 species (from 20 samples). In contrast, the tardigrade fauna in the neighbouring states have been more thoroughly investigated and consequently more than 20 species have been recorded for Alabama, Louisiana and Arkansas, and about 100 species in Tennessee (Bartels and Nelson 2007;Meyer 2013;Kaczmarek et al. 2016;Nelson et al. 2020b).
The first information on Greek tardigrades was provided 85 years ago (Marcus 1936), and since then only a couple of studies have been explicitly devoted to assessing the diversity in this country (Durante Pasa and Maucci 1979;Maucci and Durante Pasa 1982). On the island of Crete, 28 species (from more than 150 samples) have been listed based on two sampling campaigns alone (Maucci and Durante Pasa 1982). Taking into consideration recent progress in tardigrade taxonomy and faunistic studies brought about by the integrative approach, it is more than likely that the region exhibits higher species diversity and additional sampling effort may reveal more species (Vuori et al. 2020).
In this paper, we provide descriptions of two new Macrobiotus species: Macrobiotus annewintersae sp. nov. from Mississippi (USA) and Macrobiotus rybaki sp. nov. from Crete (Greece) and show their phylogenetic position within the genus Macrobiotus. Detailed morphological and morphometric data were obtained using phase contrast and scanning electron microscopy (PCM and SEM, respectively) supported by DNA sequences for four molecular markers (three nuclear -18S rRNA, 28S rRNA, and ITS-2 -and one mitochondrial -COI).
In order to perform the taxonomic analysis, animals and eggs were either extracted from culture (M. annewintersae ssp. nov.), or directly from the sample (M. rybaki sp. nov.) and split into several groups for specific analyses i.e., morphological analysis in PCM and SEM, as well as DNA sequencing (for details see sections "Material examined" provided below in the results section for each species description).

Microscopy and imaging
Specimens for light microscopy were mounted on microscope slides in a small drop of Hoyer's medium and secured with a cover slip, following protocol by Morek et al. (2016). Slides were examined under an Olympus BX53 light microscope with PCM, associated with an Olympus DP74 digital camera or under a Zeiss Axioscope A2 light microscope associated with a MiniVID digital camera. Immediately after mounting, the specimens were checked under PCM for the presence of males and females in each of the studied populations, as the spermatozoa in testes and vasa deferentia are visible for several hours after mounting . To obtain clean and extended specimens for SEM analysis, tardigrades were processed according to the protocol by Stec et al. (2015). Specimens were examined under high vacuum in a Versa 3D Dual-Beam SEM at the ATOMIN facility of the Jagiellonian University, Kraków, Poland or in a Raith e-LINE E-beam SEM at Nanoscience Center of University of Jyväskylä, Jyväskylä, Finland. All figures were assembled in Corel Photo-Paint X6, ver. 16.4.1.1281. For structures that could not be satisfactorily focused in a single light microscope photograph, a stack of 2-6 images were taken with an equidistance of ca. 0.2 μm and assembled manually into a single deep-focus image in Corel Photo-Paint X6.

Morphometrics and morphological nomenclature
All measurements are given in micrometres (μm). Sample size was adjusted following the recommendations by Stec et al. (2016). Structures were measured only if their orientation was suitable. Body length was measured from the anterior extremity to the posterior end of the body, excluding the hind legs. The terminology used to describe oral cavity armature and eggshell morphology follows Michalczyk and Kaczmarek (2003) and Kaczmarek and Michalczyk (2017). 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 ratio (Pilato 1981). Measurements of buccal tube widths, heights of claws and eggs follow Kaczmarek and Michalczyk (2017). Morphometric data were handled using the "Parachela" ver. 1.7 template available from the Tardigrada Register (Michalczyk and Kaczmarek 2013). The raw morphometric data are provided as Suppl. materials 1, 2. Tardigrade taxonomy follows Bertolani et al. (2014) and Stec et al. (2021). Thorpe´s normalisation was performed with the R software (R Core Team 2020) on the morphometric traits following Bartels et al. (2011) (SM.03 Genotyping DNA was extracted from individual animals following a Chelex 100 resin (BioRad) extraction method by Casquet et al. (2012) with modifications described in detail in Stec et al. (2020a). Each specimen was mounted in water and examined under a light microscope prior to DNA extraction. We sequenced four DNA fragments, three nuclear (18S rRNA, 28S rRNA, ITS2) and one mitochondrial (COI). All fragments were amplified and sequenced according to the protocols described in Stec et al. (2020a); primers with original references are listed in Table 1. 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 MEGA7 (Kumar et al. 2016) and submitted to NCBI GenBank (Table 2).

Phylogenetic analysis
The phylogenetic analyses were conducted using concatenated 18S rRNA+28S rRNA+ITS-2+COI sequences from Macrobiotidae, with Richtersius coronifer (Richters, 1903) and Dactylobiotus parthenogeneticus Bertolani, 1982 as outgroups. GenBank accession numbers of all sequences used in the analysis are listed in Table 2. Only species/populations with at least 3 markers were included in the analysis.
The 18S rRNA, 28S rRNA and ITS-2 sequences were aligned using MAFFT ver. 7 (Katoh et al. 2002;Katoh and Toh 2008) with the G-INS-i method (thread=4, threadtb=5, threadit=0, reorder, adjust direction, any symbol, max iterate=1000, retree 1, global pair input). The COI sequences were aligned according to their amino acid sequences (translated using the invertebrate mitochondrial code) with the MUSCLE algorithm (Edgar 2004) in MEGA7 with default settings (i.e., all gap pen-alties=0, max iterations=8, clustering method=UPGMB, lambda=24). Alignments were visually inspected and trimmed in MEGA7. Model selection and phylogenetic reconstructions were undertaken using the CIPRES Science Gateway (Miller et al. 2010). Model selection was performed for each alignment partition (6 in total: 18S rRNA, 28S rRNA, ITS-2 and three COI codons) using PartitionFinder2 (Lanfear et al. 2016), partitions and model selection process together with results are contained in Suppl. material 4. Bayesian inference (BI) phylogenetic reconstruction was performed using MrBayes v3.2.6 (Ronquist et al. 2012) without BEAGLE. Two runs (one cold chain and three heated chains each) of 20 million generations were used with a burn-in of 2 million generations, sampling a tree every 1000 generations. Posterior distribution sanity was checked using Tracer v1.7 (Rambaut et al. 2018). The MrBayes input file with the input alignment is available as Suppl. material 5, and the MrBayes output consensus tree is available as Suppl. material 6. The phylogenetic tree was visualised with FigTree v1.4.4 (Rambaut 2007) and the image was edited with Inkscape 0.92.3 (Bah 2011).
Description of the new species. Animals (measurements and statistics in Table 3): In live animals, body translucent in smaller specimens and opaque whitish in larger animals; transparent after fixation in Hoyer's medium (     ure 2A, C) and SEM ( Figure 2D). A pulvinus is present on the internal surface of legs I-III ( Figure 2B, E). Claws Y-shaped, of the hufelandi type. Primary branches with distinct accessory points, a common tract, and an evident stalk connecting the claw to the lunula (Figure 3). The lunulae I-III are smooth ( Figure 3A, C), whereas lunulae IV are dentate ( Figure 3B, D). A divided cuticular bar with double muscle attachments are poorly visible under PCM ( Figure 3A).
Mouth antero-ventral. Bucco-pharyngeal apparatus of the Macrobiotus type ( Figure 4) with ventral lamina and ten peribuccal lamellae. The stylet furcae typically-shaped, the basal portion is enlarged and has two caudal branches with thickened, swollen, rounded apices. Under PCM, the oral cavity armature is of the patagonicus type, i.e., with only the second and third bands of teeth visible ( Figure 4B, C). However, under SEM the first band of teeth is visible and composed of one row of very small cones situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae ( Figure 5). The second band of teeth is situated between the ring fold and the third band of teeth and composed of 3-4 rows of teeth visible in PCM as granules ( Figure 4B, C). The third band of teeth is divided into a dorsal ( Figure 4B) and a ventral portion ( Figure 4C). Under PCM, the dorsal teeth are seen as three distinct transverse ridges whereas the ventral teeth appear as two separate lateral transverse ridges between which one big tooth (sometimes circular in PCM) is visible ( Figure 4B, C).
Pharyngeal bulb spherical, with triangular apophyses, two rod-shaped macroplacoids and a drop-shaped microplacoid ( Figure 4A, D, E). The macroplacoid length sequence is 2<1. The first and the second macroplacoid have a central and a subterminal constriction, respectively ( Figure 4D, E).
Eggs (measurements and statistics in Table 4): The surface between processes is of the persimilis type, i.e., with a continuous smooth chorion, never with pores or reticulum (Figures 6, 7). Under PCM the surface between the processes is covered with wrinkles that appear as dark thickenings/striae, whereas under SEM the surface appears clearly wrinkled (Figures 6, 7). Processes are of a modified hufelandi type (Figures 6, 7). The proper terminal disc is absent and instead 2-8 thick tentacular arms (typically 5-6) are present in the distal part of the process (Figures 6, 7). The tentacular arms present bubble-like structures (visible in PCM). Under SEM, each tentacular arm is distally divided into many irregular digitations that are sometime covered with micro-granulation ( Figure 7C-F). Also, under SEM micro-pores can be seen on the egg surface between the processes and around the process bases ( Figure 7C, E).
Reproduction / Sexual dimorphism. The species is dioecious. Spermathecae in females as well as testis in males, clearly visible under PCM up to 24 hours after mounting in Hoyer's medium, have been found to be filled with spermatozoa ( Figure 8A, B). The species exhibits secondary sexual dimorphism in the form of clearly visible lateral gibbosities on the hind legs in males (Figure 8B, C).
DNA sequences. 18S rRNA: GenBank: MW588024-MW588025; 659 and 664 bp long.       Figure 17) and by lacking a cavity between the process trunk and tentacular arms that appears in PCM as a clearly refracting dot (the cavity present in M. anemone, Figure 17). Etymology. We dedicate this species to the singer, composer, musician, actor and the 2009 Eurovision Song Contest winner, Alexander Rybak. Material examined. 173 animals and 37 eggs. Specimens mounted on microscope slides in Hoyer's medium (156 animals + 32 eggs), fixed on SEM stubs (15+5), and processed for DNA sequencing (2+0).
Description of the new species. Animals (measurements and statistics in Table 5): In live animals, body translucent in smaller specimens and opaque whitish in larger animals; transparent after fixation in Hoyer's medium ( Figure 9A). Eyes present in live animals and after fixation in Hoyer's medium. Elliptical cuticular pores (0.6-1.5 μm in length) present all over the body and clearly visible under both PCM and SEM ( Figures 9B-D, 10). Patches of fine granulation on the external surface of legs I-III as well as on the dorsal and dorso-lateral sides of legs IV clearly visible under both PCM and SEM ( Figure 10A, B, E, F). A pulvinus is present on the internal surface of legs I-III ( Figure 10C, D).
Claws Y-shaped, of the hufelandi type. Primary branches with distinct accessory points, a common tract, and an evident stalk connecting the claw to the lunula ( Figure 11). The lunulae I-III are smooth ( Figure 11A, D, E), whereas lunulae IV are dentate ( Figure 11B, C, F). A divided cuticular bar and doubled muscle attachments are visible under PCM ( Figures 10C, D, 11A, D, E).
Mouth antero-ventral. Bucco-pharyngeal apparatus of the Macrobiotus type ( Figure 12), with ventral lamina and ten peribuccal lamellae ( Figure 13A). The stylet furcae   typically-shaped, the basal portion is enlarged and has two caudal branches with thickened, swollen, rounded apices. Under PCM, the oral cavity armature is of the patagonicus type, i.e., with only the second and third bands of teeth visible ( Figure 12B, C). However, under SEM the first band of teeth is visible as a row of irregularly distributed small teeth situated anteriorly in the oral cavity, just behind the bases of the peribuccal lamellae (Figure 13A, B). The second band of teeth is situated between the ring fold and the third band of teeth and comprised of 3-4 rows of teeth faintly visible in PCM ( Figure 12B, C) and visible as cones in SEM ( Figure 13A). Teeth of the second band are larger than those in the first band.
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 ( Figures 12B, C,  13A, B). The third band of teeth is divided into a dorsal and the ventral portion. Under both PCM and SEM, the dorsal teeth are seen as three distinct transverse ridges ( Figures 12B, 13A). The ventral teeth appear as two separate lateral transverse ridges between which one conical medial tooth (roundish in PCM) is visible (Figures 12C,  13B). Lateral cribrose area present in the buccal tube behind the third band of teeth ( Figure 13B). Pharyngeal bulb spherical, with triangular apophyses, three anterior cuticular spikes (typically only two are visible in any given plane), two rod-shaped macroplacoids and a dropshaped microplacoid ( Figures 12A, D, E). The macroplacoid length sequence is 2<1. The first macroplacoid has a weak central constriction, whereas the second is weakly constricted only subterminally ( Figures 12D, E). Eggs (measurements and statistics in Table 6): The surface between processes is of the hufelandi type, i.e., covered with a reticulum (Figures 14A, B, 15A-E). Peribasal meshes of slightly larger diameter compared to interbasal meshes ( Figures 14A, B, 15A-D). Typically, the reticulation between neighbouring processes is composed of two rows of peribasal meshes and with a third row of smaller mashes interposed (the third row sometimes missing) ( Figures 14A, B, 15A-D). Mesh diameter is usually larger than the mesh walls and nodes ( Figures 14A, B, 15A-D). The meshes are 0.4-1.4 μm in diameter, with roundish irregular shape. The pillars connecting the reticulum with the chorion surface are visible only under SEM ( Figure 15C). The bases of the processes are surrounded by cuticular thickenings that merge into the bars and nodes of the reticulum ( Figure 15C, D). These basal thickenings appear under PCM as short dark projections around the process bases ( Figure 14A, B).  Processes are of the hufelandi type with very elongated concave trunk and extremely reduced (narrow), round and convex terminal discs with irregularly jagged edges ( Figures 14C-F, 15). Under SEM the surface of the convex terminal discs is covered by small irregular granules and tubercles (Figures 15C-F).
Reproduction / Sexual dimorphism. The species is dioecious. Testis in males, which were clearly visible under PCM up to 24 hours after mounting in Hoyer's medium, have been found to be filled with spermatozoa, (Figure 16). In females spermathecae filled with spermatozoa were not observed. The species exhibits secondary sexual dimorphism in the form of small lateral gibbosities on the hind legs of males ( Figure 16). COI: GenBank: MW593931-MW593932; 658 bp long.
Phenotypic differential diagnosis. By having the OCA of the patagonicus type (only the 2 nd and 3 rd bands of teeth visible under light microscopy), egg chorion of the hufelandi type (covered with a reticulum), and egg processes with reduced (narrow) terminal disc, Macrobiotus rybaki sp. nov. is most similar to four species: Macrobiotus dariae Pilato & Bertolani, 2004, Macrobiotus noemiae Roszkowska & Kaczmarek, 2019, Macrobiotus santoroi Pilato & D'Urso, 1976 and Macrobiotus serratus Bertolani, Guidi & Rebecchi, 1996. The new species differs specifically from:     walls and nodes of the reticulum (very delicate and faint reticulation with mesh of similar sizes distributed uniformly on the egg surface between processes in M. serratus; Figure 18D, E).

Phylogenetic analysis.
The phylogenetic reconstruction ( Figure 19) recovered the genus Macrobiotus as well as the three clades found by Stec et al. (2021) and by Kiosya et al. (2021) to be monophyletic. All three clades have high support values (pp=1). The new species Macrobiotus annewintersae sp. nov. belongs to subclade B, within the Macrobiotus persimilis complex, even though the monophyly of this complex was not strongly supported (pp=0.73). Macrobiotus engbergi Stec, Tumanov & Kristensen, 2020 was recovered as the closest relative of M. annewintersae sp. nov. (Figure 19). The second species analysed in this study, Macrobiotus rybaki sp. nov., belongs to subclade A with its closest relatives being Macrobiotus wandae Kayastha, Berdi, Miaduchowska, Gawlak, Łukasiewicz, Gołdyn & Kaczmarek, 2020 and Macrobiotus vladimiri Bertolani, Biserov, Rebecchi & Cesari, 2011 (Figure 19). The newly found Swedish population identified in this study as Macrobiotus aff. polonicus, as could have been predicted from its morphological similarity with that species, clusters together with two populations of Macrobiotus polonicus Pilato, Kaczmarek, Michalczyk & Lisi, 2003 from Austria and Slovakia ( Figure 19).

Discussion
We identified two new tardigrade species in the genus Macrobiotus using an integrative taxonomy approach combining the analyses of detailed morphological and genetic data. Thanks to the phylogenetic analysis performed in this study we confirmed Macrobiotus annewintersae sp. nov. to belong to the Macrobiotus persimilis complex (as defined by Stec et al. 2021). Nevertheless, the morphological definition provided by Stec et al. (2021) does not encompass the extraordinary egg phenotype exhibited by Macrobiotus annewintersae sp. nov., indicating the need for further amendment of the characters describing this monophyletic group of species. The definition of that complex, regarding the egg processes, states "[…] single-walled egg processes […] in the shape of truncated cones terminated with a well-developed disc and with solid chorion surface […]", It is therefore clear that as M. annewintersae sp. nov. possesses 2-8 tentacular arms on the distal part of its egg processes, as opposed to 'well-developed discs', it falls outside the current definition of the group. Very similar egg processes are also present in M. anemone, which was previously included in the M. persimilis complex by Stec et al. (2021) without any elaboration on that issue (please see Table 5 in Stec et al. (2021) for the list of species included there in the complex). Therefore, to avoid inconsistency in accommodating these two species within the M. persimilis complex, we propose an upgraded definition that reads: species with white body, hufelandi type claws and with single-walled egg processes (without the labyrinthine layer = not reticulated) in the shape of truncated cones terminated with a well-developed disc or tentacular arms and with a solid chorion surface (the surface can be wrinkled and sometimes with faintly visible micropores but never properly porous or reticulated). Furthermore, we propose to tentatively include Macrobiotus andinus Maucci, 1988 within the M. persimilis complex. The species meet now all the criteria except the porous cuticle, (hence it was not considered as a member of the hufelandi group sensu Kaczmarek and Michalczyk (2017), but it is likely that these pores could be visible only under SEM similarly as in same species of the Macrobiotus pseudohufelandi complex .
In their faunistic study devoted to Greek tardigrades Maucci and Durante Pasa (1982) reported Macrobiotus anderssoni Richters, 1907, specifically from the island of Crete. According to the description provided by Maucci and Durante Pasa (1982), their Macrobiotus anderssoni population from Crete is very similar to M. rybaki sp.   nov. described in our study, with the only considerable difference being dentation on lunulae IV, that is present only in M. rybaki sp. nov.. Therefore, it is highly likely that these two populations represent closely related taxa, however, more populations from this region should be examined using an integrative approach to reliably test such a hypothesis. Based on newly found M. anderssoni material, Maucci and Durante Pasa (1982) proposed a redescription of that species. However, the proposed redescription cannot be considered as valid as they failed to designate a neotype. Even if they had done so, several regulations of the International Code of Zoological Nomenclature (ICZN 1999) and the conditions listed in Article 75.3 of the code would not have been fulfilled. Specifically, (i) the authors did not provide reasons for believing the name-bearing type specimen(s) (i.e., holotype, or lectotype, or all syntypes, or prior neotype) to be lost or destroyed, and the steps that had been taken to trace it or them; (ii) the population that they studied did not come, as nearly as practicable, from the original type locality (terra typica of M. anderssoni is Tierra del Fuego in Argentina). Moreover, Roszkowska et al. (2016) have already questioned the identification of the population from Crete, stating that it belongs to an unrecognised species of the Macrobiotus hufelandi group. In light of the discussion in Roszkowska et al. (2016) on the taxonomic uncertainty concerning M. anderssoni, further supported by the newly found egg that fits perfectly with Richters' description and which was found near terra typica, we agree with the authors' claims that it is highly likely that M. anderssoni represents the genus Mesobiotus Vecchi, Cesari, Bertolani, Jönsson, Rebecchi & Guidetti, 2016. Nevertheless, a more robust conclusion can only be made following an integrative redescription of the species, based on a population from Tierra del Fuego or nearby locality, becoming available.
Our study describes yet another two new species of the genus Macrobiotus utilising the integrative taxonomy approach. The detailed morphological examination linked with genetic data in the form of DNA sequences has allowed us also to elucidate the phylogenetic position of the studied taxa and amend the definition of the Macrobiotus persimilis complex. This further underlines the pre-eminence of the integrative approach, compared with classical taxonomy, in more reliably testing species hypotheses.