﻿Three new species of Macrostomum (Platyhelminthes, Macrostomorpha) from China and Australia, with notes on taxonomy and phylogenetics

﻿Abstract In this paper, three species of the macrostomid free-living flatworm genus Macrostomum are described. Two species, Macrostomumlittorale Wang & Shi, sp. nov. and M.shekouense Wang & Shi, sp. nov., were collected from coastal water at Shenzhen, Guangdong Province, China. One species, M.brandi Wang & Shi, sp. nov., was collected from Perth, Western Australia and Queenscliff, Victoria, Australia. Macrostomumlittoralesp. nov. differs from congeneric species within the genus in the length of the stylet, diameter of stylet opening, and the bend of the stylet. Macrostomumshekouensesp. nov. and M.brandisp. nov. differ from similar species within the genus in the stylet morphology, position of the female antrum and developing eggs, or presence or absence of the false seminal vesicle. Phylogenetic analysis based on cytochrome c oxidase subunit I (COI) gene shows that M.littoralesp. nov. and M.hystrix are sister clades on two well-separated branch, M.shekouensesp. nov. and M.brandisp. nov. are sister clades on two well-separated branches. Accordingly, both morphological and molecular evidence support M.littoralesp. nov., M.shekouensesp. nov., and M.brandisp. nov. as three new species.


Introduction
Macrostomum Schmidt, 1848 is a genus of the family Macrostomidae (Platyhelminthes; Macrostomorpha), with more than 160 species described to date from around the world (Tyler et al. 2006(Tyler et al. -2021Zhang et al. 2021). These free-living flatworms are transparent, hermaphroditic, mostly 1-2 mm in body length, and have a relatively simple general anatomy. Macrostomum is attracting interest because it contains Macrostomum lignano Ladurner, Schärer, Salvenmoser & Rieger, 2005, a versatile model organism increasingly used in evolutionary, developmental, and molecular biology (Ladurner et al. 2005;Ladurner et al. 2008;Mouton et al. 2009;Vizoso et al. 2010;Wasik et al. 2015;Wudarski et al. 2020). A recent study revealed that the unusual karyotype of M. lignano hinders it from becoming a full-fledged genomic model organism, and M. cliftonense Schärer & Brand, 2020 was suggested as the potential primary Macrostomum model to replace M. lignano . To explore and describe new species of Macrostomum will obviously be of benefit to these research fields by providing more biological and genetic resources since Macrostomum is a species-rich genus (Tyler et al. 2006(Tyler et al. -2021Brand et al. 2022a).
Taxonomy of the genus Macrostomum is particularly challenging, due to the difficulty of their study. In addition, there is considerable convergent evolution of the copulatory organ morphology, particularly in the morphology of penis stylet (Schärer et al. 2011;Brand et al. 2022b). The penis stylet for Macrostomum species is considered as the most typical and significant taxonomic feature for identification (Ferguson 1954;Rieger 1977;Rieger et al. 1994). According to morphology of the genital organs and mating behavior, species can be divided into two groups showing a hypodermic mating syndrome (HMS) or reciprocal mating syndrome (RMS) (Schärer et al. 2011). HMS species have extremely similar stylet morphologies, showing a clear case of convergent evolution, which means that those species are hard to distinguish based on stylet morphology only (Schärer et al. 2011). Some molecular markers, such as nuclear ribosomal RNA genes (18S rDNA and 28S rDNA), were used to determine the molecular phylogenetic placement or help identify Macrostomum species. However, the rDNA regions are too conserved to successfully resolve interrelationships between some Macrostomum species, such as M. lignano and M. janickei . We therefore also used a partial (mitochondrial) cytochrome c oxidase I (COI) gene sequence, a more rapidly evolving marker that was suggested to be used as potential molecular marker resolving interrelationships between species in the Macrostomorpha by several authors (Janssen et al. 2015;Xin et al. 2019;Schärer et al. 2020;Zhang et al. 2021). However, the sequence of COI gene is available for approximately only ten Macrostomum species. To understand the evolutionary relationships within Macrostomum, much greater species representation would be required.
The first species of Macrostomum described from China was Macrostomum intermedium Tu, 1934(Tu 1934. Seventy years later, our laboratory reported a second Macrostomum species, Macrostomum xiamense Wang & Luo, 2004. Since then, 22 species of Macrostomum have been reported from China, of which 21 species were newly described, 13 from freshwater and eight from brackish environments (Lin et al. 2017a;Xin et al. 2019;Zhang et al. 2021). China is likely to have a high richness and diversity of Macrostomum species, since 14 new species were reported in the last five years from south China alone, mainly in Guangdong Province.
In this paper, we describe three new species of Macrostomum, two species from China and one from Australia. Twenty COI gene sequences of seven species are provided, and phylogenetic analyses inferred from partial 18S rDNA, 28S rDNA, and COI gene sequences of Macrostomum taxa are presented.

Sample collection and rearing
Specimens of Macrostomum littorale sp. nov. and M. shekouense sp. nov. were collected in 2018 from Waterlands Resort located at the estuary of Pearl River in the west of Shenzhen (22°43.32'N, 113°45.88'E) and from the seashore at Shekou peninsula (22°28.77'N, 113°55.12'E), Guangdong Province, China, respectively.
Samples were collected by washing off sediment and organisms from floating plants or underwater stones and using 750-, 125-, and 75-μm mesh nets sequentially. The material retained by the 125-μm and 75-μm mesh nets was transported to the laboratory. All living flatworms were maintained in the water of the original location with a 12:12 h light/dark period at room temperature (25 ± 1 °C). The flatworms were fed with Paramecium sp. every two days.
In addition to our own specimens, we also analyzed specimens of M. brandi sp. nov. that were previously deposited by Brand et al. (2022a) Fig. S1D), Australia (see below for more details).

Specimen preparation, observations, and data processing
The procedures of specimen preparation followed the method described by Zhang et al. (2021). In brief, the specimens were fixed in Bouin's solution after being anesthetized successively with 5% (for 1-2 min) and 7% (for 1-2 min) ethanol in habitat water. The dehydrated specimens were embedded in paraffin and serially sectioned along the sagittal plane (thickness 6 μm) using a manual rotary microtome (Leica RM2235, Leica Biosystems, Germany). After being stained with either hematoxylin-eosin (H&E) or modified Cason's Mallory-Heidenhain stain and hematoxylin (Yang et al. 2020), the specimens were mounted in neutral balsam for histological observation. Lactic acid phenol liquid (lactic acid: phenol = 1:1 in volume) was used for penis stylet dissection.
Images were captured and measured with Olympus DP 2-BSW and Image-Pro Plus software v. 6.0. Type material has been deposited in the Institute of Zoology, Chinese Academy of Sciences, Beijing (IZCAS), with the abbreviation PLA-Ma (Platyhelminthes-Macrostomum) followed by catalog numbers.

DNA extraction, amplification, and sequencing
After being deprived of food for three days, three individuals of each species (nontypes) were placed into liquid nitrogen for 15 s for the following DNA extraction. An E.Z.N.A TM Mollusk DNA Isolation Kit (Omega, Norcross, GA, USA) was used to extract DNA. PCR reactions were performed using KOD One TM PCR Master Mix (TOYOBO Co. LTD, Japan) with a Thermal Cycler (Applied Biosystems 2720, Thermo Fisher Scientific, USA). The primers for 18S rDNA, 28S rDNA and COI sequences and PCR amplification procedures are listed in Table 1. The amplified 18S rDNA, 28S rDNA, and COI fragments were approximately 1250, 1200, and 610 bp long, respectively. PCR products were separated on a 1.2% agarose gel and purified using a Gel DNA Extraction Kit (Chinatopbio, Shenzhen, China), and were inserted into the pESI-Blunt vector (Hieff Clone TM Zero TOPO-Blunt Cloning Kit, Yeasen, Shanghai, China), respectively. The amplified DNA fragments were sequenced by Sanger sequencing by Beijing Genomics Institute (BGI, Shenzhen, China) or Beijing TsingKe Biotech Co., Ltd (Beijing, China).

Molecular phylogenetic analyses
Newly obtained sequences have been deposited in the GenBank database at NCBI. Sequences used for phylogenetic analyses in this study were obtained from GenBank under accession numbers shown in Suppl. material 1, 2: Tables S1, S2. In total, 43 18S rDNA sequences, 54 28S rDNA sequences, and 46 COI sequences from 42 Macrostomum species were included. The 18S rDNA and 28S rDNA of five species of Psammomacrostomum and one COI sequence of Psammomacrostomum sp. 5 were selected as outgroups for rDNA and COI trees, respectively.
Alignments were performed with the online version of the software MAFFT v. 7 (Kuraku et al. 2013;Katoh et al. 2017), applying the E-INS-i interactive refinement method. Surprisingly, a single-base deletion was found in the COI  Zhang, 2021, andM. zhujiangense Wang &Fang, 2016 when it was translated by ORFfinder in NCBI with the genetic code 9, and we discuss this observation in more detail below. For alignment and analysis, the missing bases were filled with N. Ambiguous sites (e.g., containing gaps and poorly aligned sites) were removed by Gblocks v. 0.91b (Castresana 2000) with default settings. The final length of the aligned 18S rDNA, 28S rDNA, and COI sequences were 1195, 776, and 406 bp, respectively. Uncorrected pairwise distances between species were calculated in MEGA v. 6.06 (Tamura et al. 2013 (Nylander 2004) was used applying the AIC; GTR+I+G model was chosen for all datasets. The gene partitions in the concatenated dataset were defined as 18S rDNA/28S rDNA. PartitionFinder v. 2.1.1 (Lanfear et al. 2017) was used to select the best-fit models for each partition based on AIC; for ML and BI methods, the models for both 18S rDNA and 28S rDNA were GTR+I+G. After that, phylogenetic trees were constructed by both ML and BI. For ML, analyses were performed in IQ-TREE v. 1.6.2 (Nguyen et al. 2015) with 1000 bootstrap replicates. Model parameters were calculated independently for each gene partition of the concatenated dataset. For BI, analyses were performed in MrBayes v. 3.2.6 (Ronquist et al. 2012) with two simultaneous runs of one cold and three heated chains. Partitioned analysis was performed on the concatenated dataset. The Markov Chain Monte Carlo (MCMC) algorithm was run for 1,500,000 generations for all datasets in four simultaneous chains. Every 1000 th generation was sampled. Burn-in was chosen as 25% of the results. Effective sample size (ESS) values of each parameter in the .p files were checked by TRACER v. 1.7.1 (Rambaut et al. 2018) to ensure good convergence. All trees were visualized using FigTree v. 1.4.3 (Rambaut 2009) and Adobe Photoshop CC 2017.

Molecular phylogeny
Twenty fragments of sequences of COI were amplified and sequenced from two specimens of Macrostomum littorale sp. nov., and 18 specimens of the other six The individuals of Macrostomum littorale sp. nov. and M. hystrix fell into one supported clade with high support values (1.00 PP, 97% BP) in the phylogenetic tree resulting from the analysis of the concatenated 18S and 28S rDNA dataset (Fig. 1). The individuals of M. shekouense sp. nov. cluster together with M. brandi sp. nov. with high support values (1.00 PP, 100% BP) in the phylogenetic tree resulting from the analysis of the concatenated dataset ( Fig. 1). However, in the more rapidly evolving COI gene tree, M. littorale sp. nov. occupied a separate branch by 0.99 PP, 82% BP, supporting a separation between M. littorale sp. nov. and its congener, M. hystrix (Fig. 2). Similarly, the individuals of M. shekouense sp. nov. clustered in a clade with high support values (0.99 PP, 93% BP), as a sister clade of M. brandi sp. nov. (Fig. 2). Type material. Two specimens: holotype IZCAS PLA-Ma0140, collected by Fan Xin from Waterlands Resort located at the estuary of pearl River in the west of Shenzhen, Guangdong,China (22°43.32'N,113°45.88'E) in April 2018, unsectioned wholebody mounted in neutral balsam. The paratype (collection date and locality same as the holotype), is one serially-sectioned specimen mounted in neutral balsam (IZCAS PLA-Ma0141). Digital photomicrographs of the holotype specimen and the sectioned paratype specimen, as well as photomicrographs of five non-type specimens (IZCAS PLA-Ma0140a-e) imaged in vivo, were deposited on the Macrostomorpha Taxonomy and Phylogeny website (at https://macrostomorpha.myspecies.info) and can also be accessed at https://doi.org10.5281/zenodo.4585492.
Habitat. Specimens were collected from Waterlands Resort at the estuary of the Pearl River at Shenzhen, Guangdong Province, China (Suppl. material 4: Fig. S1A). The animals were collected from floating plants. Salinities of the water in habitat were 9‰-11‰.
Diagnosis. A Macrostomum species with dorsoventrally flattened body, two almost round eyes, and rounded rostrum (Fig. 3A). Body length and width of mature worms are 920 ± 109 μm and 192 ± 49 μm, respectively. Testes clearly larger than ovaries. Stylet (62 ± 7.9 μm) is a hook-like and gradually narrowing funnel with an 105° bend in the 66% position (when measured from the proximal to the distal part). The stylet opening is 6 ± 0.9 μm in diameter, willow leaf-shaped, located on the concave side of the subterminal region of the stylet (Fig. 3F, G, H). Muscular walls of vesicula seminalis and vesicula granulorum thickened. No bristles and brush on sperm (35 ± 1.1 μm) (Fig. 3E).
Remarks. A comparison between Macrostomum littorale sp. nov. in this study and eleven similar species (stylet hook-like) within the genus is shown in Table 2. The main difference between M. littorale sp. nov. and M. astericis Sopott-Ehlers, 1976, M. qiaochengense Wang &Fang, 2017, andM. sp. 1 is the location of penis stylet opening (pso), which is at the concave side of the curved tube in M. littorale sp. nov., while it is located at the convex side of the curved tube in the other species. Moreover, the length of the stylet in M. hystricinum and M. astericis is shorter than that in M. littorale sp. nov. The distal opening of M. qiaochengense (diameter 20 ± 1.6 μm) is much larger than that of M. littorale sp. nov. (diameter 6 ± 0.9 μm).
Habitat. Specimens were collected from the seashore at Shekou peninsula (Suppl. material 4: Fig. S1B). The animals were collected from underwater stones. Salinities of the water in habitat were 16‰-21‰.

Diagnosis.
Macrostomum with slightly dorsoventrally flattened body and wider arc-shaped tail (Fig. 6A). Muscular wall of both vesicula seminalis and vesicula granulorum thickened. The stylet, hook-like in shape, is a gradually narrowing funnel, including a 90° bending at the 65% position from proximal end (Fig. 6D). No bristles and brush on sperm (Fig. 6E).
Etymology. The specific epithet refers to the locality where the species was found, which is the Shekou peninsula, Guangdong, China.

Macrostomum brandi Wang & Shi, sp. nov.
http://zoobank.org/53CB59EA-8C1F-4B5D-B9CC-2F2D1FBC1535 Figs 9, 10 Type material. Macrostomum brandi sp. nov. was first collected in 2017 by Brand et al. (2022a) and therein referred to as Macrostomum sp. 81 (as well as M. sp. 81 and Mac081). The present description of the species is based on photomicrographs and videos that Brand et al. (2022a) deposited with multiple specimens of the species. As the holotype we designate their transcriptome-sequenced specimen MTP LS 2864 (transcriptome accession SAMN15061091). The digital type materials of the specimen are available on Zenodo (https://zenodo.org/record/5656981) (for details see Brand et al. 2022a).
Habitat. The type specimen was collected from shallow subtidal sediment in Perth (Suppl. material 4: Fig. S1C). The longitude and latitude of the sampling site was described in part Sample collection and rearing. Salinities of the water in the habitat was 30‰.
Etymology. Species name in honor of Lukas Schärer's former PhD student Jeremias Brand, with whom Lukas Schärer collected this species in both Perth, Western Australia and Queenscliff, Victoria, Australia.
in length at the anterior and posterior body end, respectively. Tufts of sensory hairs sparsely distributed along the body edges. Rhabdite rods mainly distributed on the dorsal side of the body (Fig. 9A).
Pair of ovaries lie directly behind the testes and show the short elliptic shape, 124 ± 31 μm (n = 8) in length and 30 ± 7 μm (n = 8) in width (Figs 9C, 10A). Female gonopore opening ventrally at female antrum, surrounded by numerous cement glands. Female antrum lies at the ventral side at 50% of body length.
Remarks. A comparison between Macrostomum shekouense sp. nov., M. brandi sp. nov., and ten similar species with hook-like stylets within the genus is shown in Table 2. For the 12 listed species, the female antrum of only three species (M. shekouense sp. nov., M. brandi sp. nov., and M. obelicis) lies at the ventral side at 50% of body length, while that of the other nine species lies considerably further towards the posterior.
Based on the 28S rDNA phylogenetic tree in Brand et al. (2022a) it seems clear that there currently are no other known species that are very close to Macrostomum brandi sp. nov. M.shekouense sp. nov., M. brandi sp. nov., and M. obelicis are very similar in stylet morphology, particularly with respect to the position of the stylet opening, as well as the bending angle and position of the curve in the stylet, although the stylet is a little larger in M. brandi sp. nov. than that in M. shekouense sp. nov. and in M. obelicis. In terms of overall morphology of the flatworms, M. shekouense sp. nov. is much more similar to M. brandi sp. nov. This includes a central position of the developing eggs at ~ 50% (rather than an even more anterior position at ~ 40% in M. obelicis), and the linear anterior-posterior arrangement of the false seminal vesicle, the true seminal vesicle, and a posteriorly pointing stylet (while the M. obelicis lacks a false seminal vesicle, and has a seminal vesicle and stylet that are oriented laterally in an opposite direction). In agreement with this, the molecular phylogenetic analyses based on 18S-28S rDNA show that M. shekouense sp. nov. is a very close relative of M. brandi sp. nov.
However, in the COI gene tree, M. shekouense sp. nov. occupied a separate branch by 0.99 PP, 93% BP, supporting a separation between M. shekouense sp. nov. and M. brandi sp. nov. (Fig. 2). GDs based on the COI sequence within the genus Macrostomum were also calculated, showing that GDs between individuals of M. shekouense sp. nov. and M. brandi sp. nov. were between 10.1% and 10.9% (Suppl. material 3: Table S3), while the two specimens from Perth and Queenscliff differed by only 0.2%. Hebert et al. (2003) found that GDs between species are ordinarily greater than 3% for a range of invertebrates. Moreover, we note that M. littorale sp. nov. and M. hystrix show a clearly different stylet morphology, while the GD between these two species (8.9%) is less than that between M. shekouense sp. nov. and M. brandi sp. nov.
In addition, we calculated and compared the ratio of the length of two lines of the stylet as shown in Fig. 8 and Fig. 10, dd and vl. This suggests that the dd-to-vl ratio in M. brandi sp. nov. (5.2 ± 0.62, n = 7) is significantly higher than that in M. shekouense sp. nov. (3.1 ± 0.26, n = 5) (2-tail p < 0.001, Independent Sample Test) (Suppl. material 7: Fig. S4).
Accordingly, both morphological and molecular evidence supports that M.shekouense sp. nov. and M. brandi sp. nov. are two new species.

Discussion
It is a challenging task to identify Macrostomum species for various reasons. Firstly, these small, fragile microturbellarians are difficult to study. Secondly, there is considerable convergent evolution of the copulatory organ morphology in Macrostomum species, particularly the morphology of the penis stylet (e.g., Schärer et al. 2011;Brand et al. 2022b). Brand et al. (2022a) suggested that investigations of species within the hypodermic clade without support from molecular data thus require considerable caution due to the striking cases of convergent evolution found in HMS species. Molecular markers such as the mitochondrial COI, showing a more rapid evolutionary rate than nuclear 18S and 28S rDNA, was suggested to better resolve species-level relationships ). However, mitochondrial COI are currently not available in most species of Macrostomum to date. The present study has now identified a primer pair, which was used to amplify and sequence eight new mitochondrial COI sequences of three Macrostomum species, as well as extracting three COI sequences from previously deposited transcriptomes (Suppl. material 2: Table S2). The COI gene tree supported a separation between M. janickei and M. lignano (see also Schärer et al. 2020), and it supported separation between M. littorale sp. nov. and M. hystrix, as well as between M. shekouense sp. nov. and M. brandi sp. nov., although in all these cases these species pairs could not be separated by 18S and 28S rDNA trees (Figs 1, 2). Accordingly, with an increase in the number of COI sequences provided for Macrostomum species, molecular phylogenetic analyses based on mitochondrial COI gene could be a useful tool for helping to identify Macrostomum species in combination with morphological characters. In addition, it is suggested that mitochondrial COI sequences of Macrostomum species could be amplified using primer pair provided either by Janssen et al. (2015) or by this study showing in Table 1. In this study, a significant difference was found in dd-to-vl ratio of stylet in M. shekouense sp. nov. and M. brandi sp. nov. We propose that the dd-to-vl ratio of stylet could serve as an additional character for the delimitation of the hypodermic species with hook-like stylets, considering that the values of dd and vl vary with the change of stylet length, the bending angle and position of the curve in the stylet, and the diameter of proximal opening of stylet.
It is interesting to note that the COI gene sequences of both M. littorale sp. nov. and M. sp. 34 show single-base deletions at two positions, position 336 and position 434 when counting from the ATT start codon of the 1,548 bp COI gene of M.lignano (Egger et al. 2017, GenBank: MF078637), leading to a frameshift mutation and a TAG stop codon only ten bases (and three more stop codons in the sequenced fragment alone). Furthermore, two species, M. taurinum and M. zhujiangense, show a single-base deletion at position 434, resulting in the frameshift mutation of ten amino acid residues followed by a stop codon (Fig. 11). A similar phenomenon was previously reported in M. hystrix , for which the COI gene sequence shows a single-base deletion at position 336, leading to a frameshift mutation and a TAG stop codon only ten bases on (and eight more stop codons in the sequenced fragment alone). The single-base deletion at position 336 or at position 434 would be expected to lead to a truncation of the resulting COI protein.
Given the importance of the COI protein in the electron transport chain of mitochondrial oxidative phosphorylation, this finding is surprising and requires further examination and validation, as already noted by Schärer et al. (2020). Further studies are needed to elucidate the phenomenon of single-base deletion in the COI gene of Macrostomum species. It will be interesting to further explore the mechanism of gene expression of the COI gene in the Macrostomum species, particularly the posttranscriptional processing of precursor-messenger RNA, since at least in M. hystrix the frames-shifted transcript seems to be abundantly expressed .