Integrative description of Paramacrobiotus bengalensis sp. nov. (Tardigrada: Eutardigrada: Macrobiotidae), a new limno-terrestrial tardigrade species from the state of West Bengal, India

. Paramacrobiotus bengalensis sp. nov. was discovered in a moss sample collected from a tree in West Bengal, India. We describe this new species using detailed morphological and morphometric data obtained from phase contrast microscopy and scanning electron microscopy, along with molecular and phylogenetic data analyses. Due to the presence of a cap-like structure at the distal portion of egg processes, the new species showed the highest similarity with Paramacrobiotus garynahi (Kaczmarek, Michalczyk & Diduszko, 2005), Paramacrobiotus alekseevi (Tumanov, 2005), Paramacrobiotus ﬁ lipi Dudziak, Stec & Michalczyk, 2020, Paramacrobiotus sagani Daza, Caicedo, Lisi & Quiroga, 2017, Paramacrobiotus vanescens (Pilato, Binda & Catanzaro, 1991) and Paramacrobiotus gadabouti Kayastha, Stec, Mioduchowska & Kaczmarek, 2023. However, it can be differentiated from them by some morphological and morphometric characteristics. The genetic data corroborated our phenotypic outcome further supporting the new species hypothesis.

Introduction environments and in all climatic zones (Nelson et al. 2015). They comprise more than 1400 species belonging to 159 genera and 33 families (Degma & Guidetti 2009-2023. The family Macrobiotidae Thulin, 1928 consists of eutardigrades categorized by high species diversity and characterized by: i) presence of condensed epicuticle layer that lacks pillar-like structures; ii) presence of double Y-shaped claws with a confi guration of 2112 on each leg; iii) presence of ventral lamina on the ventral side of the buccal tube; iv) absence of cephalic papillae; v) free-laid ornamented eggs (Bertolani et al. 1996;Guidetti et al. 2000;Pilato & Binda 2010;Marley et al. 2011). The limno-terrestrial genus Paramacrobiotus was erected by Guidetti et al. (2009) based on molecular and morphological data from the two informal species groups recognized previously within the genus Macrobiotus C.A.S. Schultze, 1834(Guidetti et al. 2009). One of them was Paramacrobiotus richtersi morphogroup which is characterized by: i) animals equipped with three rod-shaped macroplacoids; ii) presence of the microplacoid; iii) areolated eggs with cone-shaped processes . The second group (Paramacrobiotus areolatus morphogroup) differs from the richtersi morphogroup by the complete absence of microplacoid in the pharynx. Currently, the entire cosmopolitan genus Paramacrobiotus comprises 45 species.

Sample collection and processing
The samples were collected from 'Acharya Jagadish Chandra Bose Indian Botanic Garden', Shibpur, Howrah, West Bengal, India (Fig. 1). A moss sample was collected from one tree in August 2021. The sample were packed in a paper bag and then dried at 30°C. Tardigrades were extracted and examined using standard methods (Stec et al. 2015). The map in Fig. 1 was made in Ocean Data Viewer ver. 5.4.0 (Ocean data view http://odv.awi.de).

Microscopy and imaging
Thirty two specimens (one holotype, 28 paratypes, three voucher specimens) and fi ve eggs were mounted on microscope slides in a small drop of Hoyer's medium and secured with coverslips following the protocols described in Morek et al. (2016). The slides were then placed in an incubator and dried for 5 days at 60°C. The dried slides were sealed with transparent nail polish and examined under a Nikon Eclipse Ni-U Phase Contrast Microscope (PCM) associated with a Nikon DS-Fi3 high-resolution Microscope digital camera (Nikon Corporation, Minato, Tokyo, Japan).
Three specimens and two eggs for Scanning Electron Microscopy (SEM) were prepared according to the protocols described by Stec et al. (2015). The specimens were fi rst subjected to ethanol/water series, then acetone ethanol series, then critical point drying with CO 2, and fi nally sputter-coated with a thin layer of gold. Specimens were examined under a Carl-Zeiss EVO-18 Special Edition SEM (Jena, Germany) in the Center for Research in Nanoscience and Nanotechnology, University of Calcutta, Kolkata. All fi gures were assembled in Corel Photo Paint 2017 edition and Paint.net. For deep structures that could not be entirely focused in a single photograph, a series of 2-6 images were taken for ca 0.50 μm and then stacked into a single deep-focus image using Corel Photo Paint 2017.

Morphometrics and morphological nomenclature
Sample sizes for morphometry were adjusted following the recommendations of Stec et al. (2016a). All measurements are presented in micrometres (μm). The structures were measured only when their orientations were suitable. Body length, excluding hind legs, was measured from the anterior extremity to the end of the body. The buccopharyngeal apparatus and claws were classifi ed as described by Pilato & Binda (2010). The terminology used to describe the oral cavity armature (OCA) follows Michalczyk & Kaczmarek (2003). The macroplacoid length sequence is indicated in accordance with Kaczmarek et al. (2014). The measurement of buccal tube length and the level of stylet support insertion point follow Pilato (1981) and Kaczmarek & Michalczyk (2017). The buccal tube width was measured as the external and internal diameters at the level of the stylet support insertion point. The pt index is the ratio of the length of a given structure to the length of the buccal tube and is expressed as a percentage (Pilato 1981). The length of each branch of the claws was measured from the base of the claw (excluding the lunula) to the top of the branch, including accessory points. The height of the processes of the eggs was measured from the base of the process to the apical end. The distance between the egg processes was measured as the shortest distance connecting the bases of two close processes. Morphometric data were handled using the 'Parachela ver. 1.8' template available from the Tardigrada Register, www.tardigrada.net/register (Michalczyk & Kaczmarek 2013). The tardigrade taxonomy followed that of Bertolani et al. (2014) and Stec et al. (2021). Raw measurements are provided in the supplementary material (Supp. File 1).

Genotyping
DNA was extracted from three individual animals following the protocol described by Casquet et al. (2012) with modifi cations described in detail by Stec et al. (2020b). Prior to extraction, each specimen was mounted in water and examined under a microscope at higher magnifi cation for identifi cation. We sequenced four DNA fragments with varying mutation rates: the small ribosomal subunit (18S rRNA, nDNA), the large ribosomal subunit (28S rRNA, nDNA), an internal transcribed spacer (ITS-2, nDNA), and the cytochrome oxidase subunit I (COI, mtDNA). Of the four fragments, only two (18S rRNA and COI) were successfully amplifi ed and sequenced using protocols described in detail by Stec et al. (2020b). Three exuviae were successfully extracted, mounted with Hoyer's medium, and submitted as voucher specimens in Protozoology Section, Zoological Survey of India (ZSI). Primers used in this study and their original references are listed in Table 1. All successfully amplifi ed PCR products were sequenced commercially by Barcode Bioscience Pvt., Ltd (Bengaluru, India). Sequences were manually checked, cleaned, and processed in Bioedit ver. 7.2.5 (Hall 1999) and submitted to GenBank.

Phylogenetic analysis
To verify the phylogenetic position of the new species, a phylogenetic tree was constructed using the concatenated 18S rRNA + COI sequences of the genus Paramacrobiotus with the sequences of fi ve Tenuibiotus Pilato & Lisi, 2011, as an outgroup ( Table 2). The sequences were concatenated using SequenceMatrix (Vaidya et al. 2011). We selected the best partitioning scheme and substitution model for the posterior phylogenetic analysis using Partition-Finder ver. 2.1.1 (Lanfear et al. 2016) under the Corrected Akaike Information Criterion (AICc) and the greedy algorithm implemented in the software (Lanfear et al. 2012). Because the COI gene codes for proteins, we partitioned our alignment of this marker into three data blocks that correspond to three distinct codon locations before partitioning. PartitionFinder recommended the GTR+I model as the best-fi t partitioning scheme for 18S marker gene data block, the GTR+I+G model for the fi rst codon positions of COI, and the GTR+G model for the second and third codon position (Supp. File 3). Bayesian inference (BI) marginal posterior probabilities were calculated using MrBayes ver. 3.2 (Ronquist et al. 2012). Two independent runs, each of four Metropolis coupled Markov chains Monte Carlo (MCMC) method, were launched for 1 × 10 7 generations, and trees were sampled every 1000 generations. An average standard deviation of split frequencies of < 0.01 was used as a guide to ensure that the two independent analyses had converged. The program Tracer ver. 1.7.1 (Rambaut et al. 2018) was then used to ensure that Markov chains had reached stationarity and to determine the correct 'burn-in' for the analysis, which was the fi rst 25% of the generations. The ESS values were greater than 200 and a consensus tree was obtained after summarizing the resulting topologies and discarding the 'burn-in.' Based on the BI consensus tree, clades recovered with a posterior probability (PP) between 0.95 and 1 were considered well supported, those with a PP between 0.90 and 0.94 were considered moderately supported, and those with a lower PP were considered unsupported. Maximun Likelihood (ML) topologies were constructed using IQ-TREE ver. 2.2.0 (Minh et al. 2020). Support for internal nodes was measured using 1000 ultrafast bootstrap replicates (Hoang et al. 2018). Bootstrap (BS) support values ≥ 70% in the fi nal tree were regarded as statistically signifi cant. All fi nal consensus trees were visualised and edited in FigTree ver. 1.4.4 available from http://tree.bio.ed.ac.uk/software/fi gtree.

Species delimitation
We performed a genetic species delimitation analysis using the Multirate Poisson tree process (mPTP) model (Kapli et al. 2017) and assembly species by automatic partitioning (ASAP) (Puillandre et al. 2021). The COI dataset for these analyses included newly generated sequences for the new species, as well as all COI sequences of the genus Paramacrobiotus downloaded from GenBank (Supp. File 4). Outgroups were excluded in both cases to avoid bias produced by a distant relationship between the outgroup and ingroup taxa.
The mPTP web server (https://mptp.h-its.org/#/tree) was used to model the multirate Poisson tree process. It is an improved PTP method that does not require user-defi ned parameters as input and computes support values for each clade using MCMC, which can be used to assess the confi dence of ML delimitation (Supp. File 5). The BI tree generated from the COI dataset was used for the mPTP analysis. We used PartitionFinder, as stated in the preceding section, setting three separate data blocks for each codon position, to generate the phylogenetic tree required for the analysis. The best models selected by the program were: JC+I, HKY+G, K80+G (Supp. File 6). Then we used MrBayes ver. 3.2 and the identical settings as mentioned in the preceding section, we calculated Bayesian inference (BI) marginal posterior probabilities using the COI alignment. The outgroup was removed from the fi nal BI tree that we used for the mPTP analysis.

Material examined
32 specimens (one holotype, 28 paratypes, three voucher specimens) and fi ve eggs were mounted on microscopic slides in Hoyer's medium. Three specimens and two eggs were fi xed for SEM preparation. Three specimens were processed for genotyping.

Description
The body is almost transparent in juveniles, white in adults, and transparent after mounting in Hoyer's medium ( Fig. 2; measurements and statistics provided in Table 3). Eyes are present in live specimens but dissolve after mounting in Hoyer's medium.
Mouth anteroventral, bucco-pharyngeal apparatus of the Macrobiotus type ( Fig. 3) with 10 peribuccal lamellae and ventral lamina. The oral cavity armature is well developed and composed of three bands of teeth ( Fig. 3B-C). The teeth in the fi rst band are granular in shape and smaller than those in the other two bands. The fi rst band of teeth is situated in the anterior portion of the oral cavity behind the bases of the peribuccal lamellae. The second band, situated between the ring fold and the third band ( Fig. 3B-C) is intermediate in size, continuous, and arranged in a row that runs around the oral cavity wall. The second band comprises cone shaped teeth which are parallel to the main axis of the buccal tube. Teeth of the  second band are uniform and regular and are not joined to each other. The third band is located at the rear end of the oral cavity between the second band teeth and the buccal tube opening. The third band is divided into two parts: dorsal and ventral, with three ventral and three dorsal teeth each (two lateral and one median, which is always slightly shorter than the lateral ones). The dorsal and ventral portions are visible under the PCM as one median ridge and two lateral transverse ridges. The medioventral tooth of the third band of teeth is subdivided into two to three smaller teeth (Fig. 3B). Additional granular teeth are absent between the second and third band of teeth on the ventral side. The pharyngeal bulb is spherically shaped with triangular apophyses. Three macroplacoids and rod-shaped microplacoids are present and distinctly visible under PCM (Fig. 3A, D). The macroplacoid sequence is 2 < 1 ≤ 3, and the fi rst macroplacoid is anteriorly thinner and arrow-shaped. The second macroplacoid is bar-shaped without constriction, whereas the third macroplacoid has a distinct sub-terminal constriction (Fig. 3D).
The claws are Y-shaped and of the hufelandi type. The primary claw branches have distinct accessory points, a common tract, and a stalk that connects the claw and lunula (Fig. 4A). The lunulae under all the claws on all the legs are smooth (Fig. 4). Leg cuticle is smooth, without any granulations present in legs I-III. Granulation is present on the hind legs but only faintly visible (Fig. 4D). Cuticular bars under the claws absent. In PCM, muscle attachments under claws I to III are visible ( Fig. 4B-C).
Eggs laid freely, white/colourless with 12-14 cone-shaped processes on the circumference (Fig. 5; measurements and statistics provided in Table 4). The space between processes is areolated with 8 to 10 areolas present around each process (Fig. 6). The surface of the areoles is without pores but sculptured with wrinkles. Processes trunk cone shaped with a cap-like structure on the top (Fig. 7), with fi ne villi-    like protrusions. Under PCM egg processes walls have fi ne reticulation which is caused by the internal labyrinthine layer within the chorion (Fig. 7A-B).

DNA sequences
We obtained sequences for two DNA markers. Out of these two successfully sequenced markers, 18S rRNA was represented by two haplotypes, whereas a single haplotype was found for COI: the 18S rRNA haplotype 1 sequence (GenBank: ON923868), 1017 bp long; the 18S rRNA haplotype 2 sequence (GenBank: ON923866) 1014 bp long; the COI haplotype 1 sequence (GenBank: OP531839), 658 bp long.

Phylogenetic analysis
The phylogenetic reconstruction performed with the BI and ML methods on the concatenated dataset of the two DNA markers showed almost identical topologies, with lower support values for the ML tree (Fig. 8A). Our analysis revealed that the Paramacrobiotus richtersi morphogroup forms a monophyletic clade, whereas the Paramacrobiotus areolatus morphogroup was recovered as a paraphyletic group ( Fig. 8A-B), which was consistent with the results presented by Stec et al. (2020c). Phylogenetic analysis supported the discovery of Paramacrobiotus bengalensis sp. nov. The new species is strongly supported in the monophyletic clade of Paramacrobiotus richtersi in both the BI and ML phylogenetic trees, establishing that it does indeed belong to the Paramacrobiotus richtersi morphogroup.

Ecological information
The moss (species unknown) was collected from the mango tree Mangifera indica L., at a height of approximately 2 m from the ground. The altitude of the type locality is 12 m above sea level. The type locality is situated on the banks of River Ganges.

Phenotypic differential diagnosis
With the presence of microplacoid Paramacrobiotus bengalensis sp. nov. is assigned to Paramacrobiotus richtersi morphogroup. According to  only two species within this group have egg processes terminated with cap-like structures, namely Paramacrobiotus garynahi (Kaczmarek, Michalczyk & Diduszko, 2005) and Paramacrobiotus alekseevi (Tumanov, 2005). This structure was also directly reported for recently described species Paramacrobiotus fi lipi Dudziak, Stec & Michalczyk, 2020 and Paramacrobiotus gadabouti Kayastha, Stec, Mioduchowska & Kaczmarek, 2023. Importantly, our literature studies based on original descriptions of taxa belonging to Paramacrobiotus richtersi morphogroup allowed us to identify two other species of Paramacrobiotus in which the cap-like structure was not directly reported but seems to be present. These are Paramacrobiotus sagani Daza, Caicedo, Lisi & Quiroga, 2017 and Paramacrobiotus vanescens (Pilato, Binda & Catanzaro, 1991). Therefore, we provide below the differential diagnosis that compares our new species with the six taxa mentioned above. The new species can be differentiated from: Paramacrobiotus alekseevi, known only from the type locality in Thailand (Tumanov 2005) and from China (Beasley & Miller 2007), by: the medio-ventral tooth usually subdivided into two (only in rare cases into three) smaller teeth (the medio-ventral tooth always subdivided into three to fi ve smaller teeth in P. alekseevi), the absence of granulation in legs I-III (fi ne granulation present in P. alekseevi).
Paramacrobiotus fi lipi, reported from the type locality in Malaysia (Stec et al. 2020a), by: the absence of granulation on body cuticle (granulation present in P. fi lipi), the absence of granulation in legs I-III (fi ne granulation present in P. fi lipi), a shorter egg processes (process height: 12.5-14.7 μm in new species vs 17.8-25.2 μm, in P. fi lipi), the absence of pores in egg areoles (porous areoles present in P. fi lipi).
Paramacrobiotus gadabouti, reported from the type locality in Madeira (Portugal), but also additional localities in Portugal, France, Tunisia and Australia (Kayastha et al. 2023), by: the presence of eyespots (eyespots absent in P. gadabouti), the absence of granulation in legs I-III (fi ne granulation present in P. gadabouti), the absence of pores in egg areoles (sculptured areoles with pores in P. gadabouti).
Paramacrobiotus garynahi, known only from the type locality in Russia (Kaczmarek et al. 2005), by: the presence of eyespots (absent in P. garynahi), the absence of granulation in legs I-III (granulation present in P. garynahi), smaller eggs ( Paramacrobiotus sagani, reported from the type locality in Colombia (Daza et al. 2017), by: the absence of granulation on body cuticle (granulation present in P. sagani), the absence of cuticular bars under claws (poorly developed bars present in P. sagani), the absence of granulation in legs I-III (fi ne granulation present in P. sagani), the absence of pores in egg areoles (porous areoles present in P. sagani), smaller egg bare diameter (64.6-73.6 μm in new species vs 73.7-87.7 μm in P. sagani).
Paramacrobiotus vanescens, reported from the type locality in Tanzania (Pilato et al. 1991), by: the presence of eyespots (eyespots absent in P. vanescens), an obvious microplacoid in new species (described as, "little, faint, sometimes almost invisible" in P. vanescens), the absence of granulation in legs I-III (fi ne granulation present in P. vanescens), the absence of cuticular bars under the claws (faint bars present in P. vanescens), a shorter egg processes (12.4-14.7 μm in new species vs 16-17 μm in P. vanescens).

Genotypic differential diagnosis
The ranges of uncorrected genetic p-distances between the new species and species of the genus Paramacrobiotus, for which sequences are available in GenBank and are as follows (from the most to the least conservative): 18S rRNA: 0.32%-3.1% (1% on average), with the most similar being Paramacrobiotus metropolitanus Sugiura, Matsumoto & Kuneida, 2022 from Tokyo, Japan (LC637243) and the least similar being Paramacrobiotus areolatus (Murray, 1907)

Molecular phylogeny and species delimitation
The fi fth phylogeny of the genus Paramacrobiotus is presented in this study. The fi rst was presented by Guidetti et al. (2009), who used two distinct phylogenetic analyses based on 18S rRNA and COI sequences to establish this taxon from the genus Macrobiotus Schultze, 1834. Guidetti et al. (2019) conducted a second investigation of the genus Paramacrobiotus, which included 11 species. The authors re-described the nominal species Paramacrobiotus richtersi, as well as several additional new species from this genus. The phylogenetic results obtained by Guidetti et al. (2019) were subsequently corroborated by Stec et al. (2020c) and Kayastha et al. (2023), these studies revealed that the Paramacrobiotus areolatus morphogroup is paraphyletic while Paramacrobiotus richtersi forms a monophyletic clade using two or four concatenated genetic markers, respectively (18S+COI; 18S+28S+ITS2+COI). This outcome caused the suppression of formerly proposed subgeneric division within Paramacrobiotus by . Our study recovered a nearly identical tree topology, which is consistent with the fi ndings of Stec et al. (2020c). Paramacrobiotus bengalensis sp. nov. cluster with other Paramacrobiotus richtersi morphogroup species (Fig. 8A-B). The analysis also indicated that Paramacrobiotus bengalensis is closely related to Paramacrobiotus metropolitanus Sugiura, Matsumoto & Kuneida, 2022 from Tokyo, Japan. This was also evident when examining the genetic distances that showed a signifi cant level of similarity between DNA sequences of nuclear markers (p-distance; 18S rRNA: 0.32%) and the COI dataset, which had the lowest genetic distance among all comparisons with other taxa of Paramacrobiotus (p-distance; COI: 21%). Interestingly, although the eggs of our new species and Paramacrobiotus metropolitanus are different (the latter lacks the cap-like structure), the genetic similarity and hence close phylogenetic relationship is refl ected in the distribution of leg granulation. Both these species have been confi rmed to have granulation present only on the hind legs. This may further support the previous suggestions that the evolution of egg morphology is faster/more dynamic than animal morphology (Guidetti et al. 2013;Stec et al. 2016b;Stec 2022). Finally, both mPTP and ASAP delineation based on COI sequences also clearly supported our study specimens as a distinct new species.

Conclusions
Paramacrobiotus bengalensis sp. nov, is a new species to science. The species is described using an integrative approach of morphometry, microscopic imaging (PCM and SEM), and genetic comparison with two DNA markers (18S rRNA and COI). The diversity of cryptic and pseudocryptic species within this group complicates taxonomic identifi cation (Guidetti et al. 2019). The smaller number of morphological characteristics and variations among this group presents diffi culties in identifying and defi ning a species from this genus exclusively based on traditional taxonomy. Recent studies that utilized integrative methods, namely, DNA sequencing combined with detailed morphological and morphometric data, have proven to be extremely useful in uncovering the evolutionary relationships within this genus and unmasking new species. However, only a few of the 45 currently recognized taxa of Paramacrobiotus have DNA sequences available in GenBank, which generates bias during a phylogenetic analysis and hinders species identifi cation based on DNA. More intensive sampling of taxa of this genus with detailed morphometric data and DNA sequence data is still needed to reveal the hidden species richness within this cosmopolitan group of tardigrades.