Daphnia japonica sp. nov. (Crustacea: Cladocera) an eastern Palearctic montane species with mitochondrial discordance

The Daphnia longispina complex (Crustacea: Cladocera) contains several keystone freshwater species such as D. longispina O.F. Müller (D. rosea Sars is a junior synonym), D. galeata Sars, D. cucullata Sars, and D. dentifera Forbes. The complex is common throughout the Holarctic, but there are several geographic regions where local forms have been assigned to European species names based on a superficial morphological resemblance. Here we examine the species status of a form that was previously assigned to D. rosea from a montane bog pond on Honshu, Japan. We used two nuclear non-coding loci (nDNA), mitochondrial sequences (the ND2 protein-coding region) and morphology for evidence. The mitochondrial gene evidence supported the existence of a divergent lineage that is more closely related to D. galeata than to D. dentifera. However, morphology and the nuclear DNA data indicated a lineage that is most closely related to D. dentifera. As our evidence supported the existence of a cohesive divergent lineage, we described a new species, Daphnia japonica sp. nov. Recognition of local and subalpine diversity in this group is critical as ongoing anthropogenic disturbance has been associated with introductions, local extirpations, and hybridization.


INTRODUCTION
Crustaceans of the genus Daphnia O.F. Müller (Crustacea: Cladocera) are model organisms in aquatic toxicology, ecology, and evolutionary genetics (Benzie, 2005;Smirnov, 2017). Although Daphnia is the best studied cladoceran genus, there is still taxonomic confusion. Presently, Daphnia is composed of three subgenera D. (Australodaphnia) Colbourne, Wilson & Hebert; D. (Ctenodaphnia) Dybowski & Grochowski and Daphnia s. str. DNA sequence evidence (nuclear and mitochondrial) supported these divergent subgenera (Adamowicz et al., 2009;Colbourne, Wilson & Hebert, 2006;Kotov, 2016;Omilian & Taylor, 2001). The subgenus Daphnia is further subdivided into two species groups: D. pulex and D. longispina. The Daphnia longispina group is common in Holarctic standing waters, but rare in tropical and subtropical morphology, and protein-coding mitochondrial gene sequences and link this form with additional montane populations and formally describe a new endemic species, Daphnia japonica sp. nov.

MATERIALS AND METHODS
The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid:zoobank.org: pub:7AAA1961-81B8-46D8-8D0A-9FE1F5CD776D. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central SCIE and CLOCKSS.

Ethics statement
Plankton samples from Japan were collected by Dr. S. Ishida during his Ph.D. dissertation work, such sampling does not require special permission in Japan. The species was not assessed as endangered at the time of collection and is currently not subject to specific regulations, however all efforts were taken to ensure that the collection and preservation of animals was performed with due consideration of their welfare. The number of individuals taken did not represent a significant proportion of the clonal populations present at each site.

DNA sequencing
Ethanol-preserved specimens of the Daphnia longispina group (see Table 1 and Table S1 in supplement, which includes sequences from Genbank and those from this study) were exposed to DNA extraction (Quickextract DNA extraction from Epicentre as modified by Ishida et al. 2007), PCR (followed Ishida & Taylor, 2007a but with 50 C PCR Annealing temperature), and bidirectional Sanger sequencing (TACGEN). Sequences were assembled and compared to representative sequences of named Holarctic species (e.g., Daphnia dentifera, Daphnia galeata, and Daphnia tanakai) and one ND2 sequence from an unnamed lineage (Petrusek, Thielsch & Schwenk, 2012) in the Daphnia longispina complex using Geneious (https://www.geneious.com). Sequences (and their genomic locations) were identified from the reference genome of Daphnia galeata (Nickel et al., 2021) by BLASTn and included for each gene tree. A region from the ND2 gene of the mitochondrial genome was amplified and sequenced using the primers and protocol of Ishida & Taylor (2007a). As part of a larger effort to identify informative, single copy nuclear regions, we pooled genomic DNA of 500 individuals of Daphnia dentifera (4 mg dry mass). This total DNA was extracted using Nucleospin gDNA Cleanup (Macherey-Nagel) and exposed to restriction enzyme digestion with Rsa I (New England Biolabs).
Purified DNA was size selected (1-3 kbp) using a 2% agarose gel, and further purified using Nucleospin gel Cleanup. After dephosphorylation using Shrimp Alkaline Phosphatase, we performed cloning with a Zero Blunt TOPO PCR cloning kit (Invitrogen, Waltham, MA, USA). Colonies were then sequenced and primers designed. Of these, two nuclear regions were successfully amplified and sequenced from the target specimens of Misumi-ike. We used the following PCR primers (custom oligos from Integrated DNA Technologies): Locus 1 (F: TTTACCGATGGGCCGACCAGATTAGAG, R: GCATCCACTTGTCAGCGCCGTTTGTCA), Locus 2 (F: CCTGTTAAAATCAACAATAACAAATAGGAA, R: GCCAATTTTATACGATTTGATGTTATGC). A nucleotide BLAST to the reference Daphnia galeata genome indicated that these loci are present as intergenic single copies and located on contig CAKKLH010000168 from bases 1032297 to 1032629 (locus 1) and contig CAKKLH010000342 from bases 156848 to 157308 (locus 2). Cloning (with  (Gouy, Guindon & Gascuel, 2010) using the invertebrate mitochondrial genetic code. The nuclear loci (nDNA) were aligned using MAFFT (Katoh & Standley, 2014). Codon-specific partitioned models were estimated in IQtree (Nguyen et al., 2014) for the ND2 alignment. Standard model-fitting was applied to the nDNA alignment. The best fit substitution models were TN+F+G4 for nuclear locus one, and HKY+F+G4 for locus 2. Best fit substitutions models for the partitioned codon alignment of the ND2 gene were HKY+F+I, TPM3+F +G4, and TN+F. The best fit model for the amino acid alignment of ND2 was mtVer+G4. Tests of branch-specific relaxed purifying selection were made using the RELAX routine of HyPhy (Wertheim et al., 2015). Maximum Likelihood estimates of phylogenies were made in IQtree with optimal models. Two support values were estimated: approximate likelihood ratio tests and ultrafast bootstraps. Average K2+G distances among clades were estimated in Geneious (https://www.geneious.com) with the species identification plugin.

Plankton sampling and morphological analyses
As specimens from Misumi-ike have been associated with Daphnia dentifera (Daphnia rosea sensu Uéno, Hoshino & Mizuno (1959)), our taxonomic sampling targeted the Daphnia longispina complex of the Palearctic. We also emphasized comparison of the specimens from Misumi-ike with Japanese populations of this complex (this includes two collections, Hourai-numa and Imori-ike, with putative new species from Ishida et al., 2011). Additional sequences were obtained from GenBank, including the only existing reference genome in the group (Daphnia galeata from Nickel et al., 2021). Zooplankton samples were collected and examined using standard methods (as in Garibian et al., 2021) with a net mesh size of 50 µm. Specimens of Daphnia were presorted with ethanol-preserved samples being selected under a binocular stereoscopic microscope Leica MZ7.5. They were then studied in toto under optical microscopes (Olympus CХ 41) in a drop of a glycerol-ethanol mixture. Then, 10 parthenogenetic females, five ephippial females, two juvenile and three adult males were dissected under a stereoscopic microscope using tungsten needles (Dumont & Negrea, 2002) to analyse appendages and postabdomens. Drawings were prepared using a camera lucida attached to the optical microscope. Morphological comparisons of the named species in the group in the far eastern Palearctic, and Daphnia cucullata were made with information in the literature (Table 2 and in the differential diagnoses). Taxonomic descriptions used standard anatomical terms for the genus Daphnia following Kotov et al. (2021).

RESULTS
The alignment lengths were 930 nt for the ND2 gene, 480 nt for nuclear locus 1, and 501 nt for nuclear locus 2. The mitochondrial locus ( Fig. 1) supported the genetic uniqueness and cohesiveness of the divergent lineages of Daphnia: D. dentifera, D. longispina, D. turbinata, D. galeata, D. cucullata and D. japonica sp. nov. Phylogenies of the nuclear loci supported the existence of D. dentifera, D. galeata and D. japonica sp. nov. as lineages (Fig. 2). However, for the nuclear loci, D. dentifera and D. japonica were not reciprocally monophyletic. No mitochondrial or nuclear sequence had a significant base composition difference from other sequences in their respective alignments (based on goodness of fit tests in IQtree). The test of relaxed selection for the mitochondrial branch leading to D. japonica was not significant (K = 0.17, p = 0.36). However, one published ND2 sequence (Daphnia longispina, JX069351) had the longest branch on the amino acid tree (Fig. 1B) and significant relaxed selection (K = 0.31, p = 0.001).   Figure 2 Maximum Likelihood phylograms of the Daphnia longispina complex based on nuclear non-coding loci one and two. Sequenced specimens are from the Daphnia galeata mendotae, The nuclear genes showed monophyly of D. japonica sp. nov. with specimens from populations proposed by Ishida et al. (2011) as a new species. The ND2 sequences preserved the open reading frame. The phylogeny of ND2 revealed a placement of D. japonica sp. nov. within the D. galeata/D.cucullata clade using both nucleotide (Fig. 1A) and amino acid alignments (Fig. 1B). In contrast, nuclear locus 1 ( Fig. 2A) and nuclear locus 2 (Fig. 2B) placed D. japonica sp. nov. within D. dentifera. However, the branches grouping D. japonica within D. dentifera were poorly supported, suggesting that the hypothesis of a sister group association between D. japonica and D. dentifera is also plausible. The average genetic distance of the novel lineage was 0.18 (K2P+G4) from D. galeata for ND2 and 0.12 (K2P+G4) from D. dentifera for nuclear locus one. For nuclear locus one, the divergent lineages from Honshu (i.e., D. japonica sp. nov. and the populations identified by Ishida et al., 2011) had an average within group genetic distance of 0.035.  reduced, small mound present on each first, second and third segments. Gonopore opens subdistally, without genital papilla. Antenna I (antennule) long, almost straight; antennular setae small, its tip projected beyond antennular body tip. Male seta (flagellum) located on low process, longer than aesthetascs, bisegmented, with hooked tip. Length of females 0.9-2.1 mm, adult males 1.3-1.6 mm.

TAXONOMY
Adult parthenogenetic female. Body subovoid in lateral view, maximum height in body middle (Fig. 3A). Dorsal margin regularly convex, depression between head and rest of body absent. Postero-dorsal angle with very short caudal spine, ventral margin regularly convex. Head very large, with low anterior crest and long, straight rostrum, its tip subdividing into two lobes by a "line" of pre-rostral fold (Figs. 3B and 3C); posterior margin of head flat or slightly convex, without medial keel (Fig. 3B, arrow); ventral margin of head concave. Compound eye large, lying ventral to middle body axis and out of anterior most extremity of head; ocellus small. Labrum as rectangular main body with large, setulated distal labral plate (Figs. 3B, 3D). Carapace subovoid, spinules present on caudal spine and occupying more than a third of dorsal and a half of ventral valve margin (Figs. 3A, 3E). Inner face of posterior valve portion with fine setules, separated into groups by longer setules (Fig. 3F). Abdomen with first (basal most) abdominal process almost straight, directed anteriorly, second (middle) process shorter, third process as massive mound on the segment, fourth process absent (Figs. 3G and 3H). Postabdomen elongated, tapering distally, with ventral margin straight. Preanal margin long, concave; preanal angle smooth, postanal angle smooth. Paired spines on postanal angle and anal portion, their size increasing distally (Figs. 3H and 3I). Postabdominal seta as long as preanal margin, its distal segment shorter than basal armed by a continuous row of thin setules (Fig. 3H).
Juvenile female. Body more elongated, with straight posterior margin and longer caudal spine (Fig. 5A). Head with straight ventral margin, rostrum short, dorsal margin convex, with 1-3 neckteeth, dorsal organ in posterior portion of head ( Fig. 5B: arrow), but anterior to small notch or depression.
Ephippial female. A depression between head and valves, dorsal margin of valves straight (Fig. 5C). Ephippium with two resting eggs, with axes being almost perpendicular to ephippial dorsal margin, egg chambers not separated from each other, much of ephippium darkly pigmented and covered with sculpturing of polygonal cells, postero-dorsal portion of valves with caudal spine incorporated into ephippium (Figs. 5C, 5D). Dorsal wall of carapace forming dorsal plate without spinules (Fig. 5E).
Juvenile male of pre-reproductive instar. Body elongated, dorsal margin straight, rudimentary depression between head and valves, postero-dorsal angle with short caudal spine (Fig. 5F). Head with rounded rostrum (Figs. 5G and 5H), strong rounded depression dorsal to antenna I base (Fig. 5H: arrow), ventral margin straight, anterior-most extremity of head not projected; 4-5 neckteeth located on special projection ("pedestal") (Fig. 5I). Eye medium-sized, ocellus small. Valve with antero-ventral angle not prominent, small denticles at ventral margin, but no setae. Antenna I short, with fine antennular sensory and short male seta (shorter than aesthetascs) located on top of low distal process (Fig. 5J: arrow).
Adult male. Body elongated, dorsal margin of valves straight, depression between head and valves shallow, postero-dorsal angle with long caudal spine (Fig. 6A). Head with rounded rostrum, massive projection posterior to antenna I base (Fig. 6B: arrow). Distal-most head extremity projected ventrally, fully occupied by large compound eye, ventral head margin slightly concave. Valve with antero-ventral angle distinctly prominent ventrally, entire ventral margin with numerous setae, located submarginally (on inner face of valve) in anterior and posterior portions of valve (Figs. 6C and 6D). A row of setules on inner face of posterior margin subdivided into groups by longer setules (Fig. 6E). Abdomen with all abdominal processes reduced, only small mound present on each first (basal most), second and third segments (Figs. 6F and 6G). Postabdomen with slightly convex ventral margin; preanal margin concave, preanal angle expressed, postanal angle rounded. Gonopore (Fig. 6F: arrow) opens subdistally, without genital papilla. On outer surface of postabdominal claws, three pectens composed of thin setules with a modest separation of proximal pectens (Fig. 6H). Antenna I long, almost straight; antennular setae small, its tip projected beyond antennular body tip. Male seta (flagellum) located on low process, longer than aesthetascs, bisegmented, with hooked tip (Figs. 7A and 7B). Antenna II as in female (Figs. 6I, 7D and 7E), but sensory seta on inner side of basal segment in its distal portion long (Fig. 7C). Maxilla II (rarely studied in Daphnia) present in male (Fig. 7F: mxII) together with maxilla I and paragnath (pgn).
Differential diagnosis. D. japonica sp. nov. differs from the species of the D. longispina species complex known from Japan in a series of characters (See Table 2). D. japonica can be differentiated from D. galeata by the absence of the helmet and a straight or slightly curved posterior head margin lacking a medial keel in females and a very massive antero-ventral head extremity fully occupied by a large compound eye. Unlike the males of D. japonica, the males of the lacustrine D. ezoensis Ueno, 1972 possess a rostrum and lack pronounced ventral anterior margin setation. The females of D. ezoensis have a lower head shape (see Tanaka, 1997) as compared to D. japonica sp. nov. Daphnia dentifera is most easily confused with D. japonica. The females of D. japonica sp. nov. have a longer rigid seta on limb II as compared to females of D. dentifera, and the medial keel is fully absent in D. japonica sp. nov. but is present in D. dentifera. In contrast to D. cucullata, aesthetascs in D. japonica sp. nov. are protruding on a small mound located far from the rostrum tip. In contrast to D. longispina, the ephippium of D. japonica sp. nov. lacks spinules on the posterior margin. Finally, the male abdominal processes is are small and indistinct in D. japonica sp. nov., which is not the case in other species considered here.
Distribution. D. japonica was likely studied and illustrated from Ôtori-Misumi-Ike as D. rosea by Uéno, Hoshino & Mizuno (1959). However, we are uncertain of the distribution of D. japonica beyond the type location of Misumi-Ike. We found nearly identical nuclear sequences to those of the type location in Iwate (40.6096 N 140.9390 E,685 m.a.s.l.) and Lake Imori-ike (in reality, this is a pond c.a. 20 × 50 m), Niigata (36.6314 N,138.5361 E, 1,335 m.a.s.l.) (populations also studied by Ishida et al., 2011). We tentatively propose that D. japonica is distributed in mountain ponds and boglakes, on Honshu Island, Japan.

Evidence of a new species
The genetic and morphological evidence supports the existence of a novel species of Daphnia present on Honshu, Japan. The strong support for monophyly and pronounced divergence at two nuclear loci and the mtDNA locus is consistent with the species hypothesis. The nDNA supports a grouping of the Misumi-Ike population with specimens from Imori-ike and Hourai-numa populations that contained a putative new species proposed by Ishida et al. (2011). mtDNA sequences of specimens from all three populations are also divergent from known species (Misimu-ike from this study and Imori-ike, Hourai-numa from Ishida et al., 2011;Beninde, 2021). Although we used the same collections as Ishida et al. (2011) for Hourai and Imori, assignment of the earlier mitochondrial lineages (Hourai-numa, Imori-ike) to D. japonica awaits direct comparison with the mitochondrial markers used by Ishida et al. (2011). The open reading frames of ND2 sequences from Misumi-Ike and the lack of significant relaxed selection are inconsistent with the mitochondrial pseudogene hypothesis. Also, the strong divergences (18% for mtDNA) between this lineage and known species make incomplete lineage sorting less likely as an explanation for the novelty of this group. As we included samples from other Holarctic regions in the mtDNA analysis, it is unlikely that the divergences represent cryptic invaders. We failed to detect other lineages coexisting with D. japonica sp. nov., but increased sampling may change this finding. The discordance in the association of mtDNA and noncoding nDNA/morphology is consistent with ancient asymmetric introgression (and perhaps capture of a divergent mtDNA lineage). Beninde (2021) found a similar discordance for this lineage (not including samples from Misumi-Ike) using genome-scale data. It is unknown if the putative mitochondrial donor species is extant.
As the Daphnia longispina complex is a geographically widespread syngameon (Taylor, Sprenger & Ishida, 2005), introgression has been invoked for polymorphic populations with discordant mtDNA haplotypes (Thielsch et al., 2017;Schwenk, Ender & Streit, 1995). Recent stabilized introgressants have been proposed for the formation of D. mendotae (Taylor & Hebert, 1993;Taylor, Sprenger & Ishida, 2005) and some members of the D. pulex group (Marková et al., 2013). Ishida et al. (2011) predicted the existence of stabilized ancient (i.e., pre-glacial) introgressed lineages in weakly glaciated regions such as Japan. We have no calibrated molecular clock for among-population divergences that includes Misumi-ike specimens. If the lineages of Ishida et al. (2011) do belong to D. japonica, then the divergences for mtDNA among populations suggest that the species origin predates the last glaciation. Indeed, Beninde (2021) estimated that this lineage shared a common ancestor with D. dentifera/D. longispina about 15 MYA.
There is increasing support from the genomic studies that stabilized members of syngameons can and should be taxonomically recognized despite a lack of complete reproductive isolation (Cannon, 2021). Presently, it is unknown if D. japonica sp. nov. can coexist with extant members of the complex. However, members of the D. longispina complex are almost never detected in very low pH waters (Iwabuchi et al., 2017). Indeed, one of the most acid-tolerant species, D. dentifera, fails to survive and grow at a pH of 4.5. In contrast, D. japonica sp. nov. has been detected in Misumi-Ike when the pH was measured as low as 4.4 (Uéno, Hoshino & Mizuno, 1959). Given the genomic, morphological and ecological cohesion of the divergent lineages that we observed, we propose a new species Daphnia japonica sp. nov.

Morphological evolution and phenotypic plasticity
In addition to genetic markers, we were able to identify several taxonomically informative morphological traits. Several of these traits (such as head shape) are usually phenotypically plastic, but plasticity need not preclude taxonomic informativeness. For example, the reaction norms of helmet shape for the longispina complex in Japan are nonoverlapping between Daphnia galeata and Daphnia dentifera (Ishida et al., 2011). Likewise, fully developed inducible neckteeth appear to be absent from D. galeata. Juveniles in a few populations bear a single tooth on the head (Glagolev, 1986), but such populations, when examined genetically, have been assigned to early generation hybrids (Taylor & Hebert, 1992). We have also found that the rigid seta length ratios are informative in East Asia for this species complex. Although this seta length ratio is thermally plastic in Daphnia galeata (Kim & Taylor, 2022), D. japonica has a larger warm-water ratio (>10 C) than is known for D. galeata. Non-plastic morphological differences among species of the Daphnia longispina complex (especially among the parthenogenetic females) are few in number. However, the abdominal processes of parthenogenetic females and males appear to have diagnostic value in the group (Brooks, 1957;Ueno, 1972).
As the mitochondrial gene trees group D. japonica sp. nov. as a divergent lineage related to D. galeata/D. cucullata (i.e., discordant with morphology), there may be an undetected species in this group that donated this genome to D. japonica sp. nov. The results support the contention that the analyses of both nuclear and mitochondrial markers are necessary when divergent mitochondrial haplotypes are detected in this species complex (Thielsch et al., 2017). Presently, D. japonica sp. nov. is an endemic taxon known only from Honshu.
Our findings follow a series of discoveries of montane and subarctic lineages of the Daphnia longispina complex (Taylor & Hebert, 1994;Taylor, Hebert & Colbourne, 1996;Nilssen et al., 2007;Zuykova et al., 2018b). Populations from such regions with a longispina-like morphology must be carefully assessed for cryptic species. Some of these are likely undescribed species, such as the Daphnia from Berse in Norway (Petrusek et al., 2008;Zuykova, Bochkarev & Katokhin, 2013;Zuykova et al., 2018a). Understanding hidden diversity in these relict habitats warrants research priority as these waters are very sensitive to anthropogenic changes such as eutrophication, climate warming, and species introductions.

The Eastern Palearctic as a center of cladoceran endemism
Here we describe a new species of Daphnia that is likely present in but a few mountain lakes and ponds in Japan. The daphniid fauna of many Japanese and other eastern Palearctic waters have recently been studied in detail with genetic markers (Ishida, Kotov & Taylor, 2008;Ishida et al., 2011;Ishida & Taylor, 2007a;So et al., 2015). Also noteworthy is that well-studied temperate species have populations in East Asia that are genetically divergent from western Palearctic populations (Ishida & Taylor, 2007a;Ishida & Taylor, 2007b;Ma et al., 2014;Tokishita et al., 2017). Palearctic montane and subarctic specialists such as Daphnia umbra and Daphnia lacustris are presently unknown from the eastern Palearctic (Zuykova et al., 2018b). Korovchinsky (2006) proposed that the subtropics and temperate regions were centers of ancient (Tertiary) cladoceran endemism. Indeed faunistic, taxonomic and molecular studies have revealed numerous endemics in East Asia: the Russian Far East, China, Korea and Japan (Ishida, Kotov & Taylor, 2008;Kotov et al., 2021;Kotov et al., 2011a;Kotov et al., 2011b;Ma et al., 2019;Makino et al., 2020;Maruoka et al., 2018;Smirnov & Sheveleva, 2010;Yamamoto, Makino & Urabe, 2020). Direct phylogeographic evidence of a pre-Pleistocene age of some phylogeographic patterns have also been obtained (Kotov et al., 2021). Other taxa in this region are apparently of southern origin as they are thermophiles with a pronounced northern range limit (Garibian, 2017;Kotov, 2016;Kotov, Jeong & Lee, 2012;Kotov et al., 2011a;Kotov et al., 2011b). Regarding the Holarctic pattern, it is expected that East Asia would have suffered less extinction of temperate endemics compared to other regions (such as the Nearctic) because of the reduced influence of Pleistocene glaciations. D. japonica sp. nov. appears to represent such a pre-Pleistocene relict in agreement with Korovchinsky (2006).

CONCLUSIONS
We find genetic and morphological evidence that supports the existence of a new endemic species of Daphnia from East Asia. We describe Daphnia japonica sp. nov. from a montane bog on Honshu Island. The mitochondrial phylogeny was discordant with the morphological and nDNA similarity to D. dentifera. The finding continues a trend of hidden diversity in ecologically sensitive montane habitats and in East Asia for Daphnia.