In the Asia-Pacific region, the COI DNA test revealed the divergence of the bivalve mollusc Mactra chinensis into three species; can these species be distinguished using shell coloration and sperm structure?

According to COI DNA barcoding testing, the marine bivalve mollusc Mactra chinensis, which is native to the Asia-Pacific region, diverged into three species. These species were preliminary characterized as M. chinensis COI clade I, M. chinensis COI clade II and M. chinensis COI clade III. To find out whether it is possible to morphologically distinguish samples representing genetic clades, we examined the color of the shells and the structure of the spermatozoa. It was found that the number of detected coloration types exceeds the number of detected species. In addition, it was shown that individuals belonging to the same genetic clade can have shells of different colors. Consequently, it is impossible to choose one type of shell coloration as a species-specific trait. For sperm, the sperm morphological patterns found in each of the three species are consistent with the M. chinensis sperm model described in previous reports. However, the single sperm variant is also not applicable to discriminate between species derived from M. chinensis, since heterogeneous variants of spermatozoa differing in the length of the acrosomal rod were found. We hypothesized that genetic divergence of species could cause a shift towards predominance of one of the sperm variants, and that species-specific sperm morphs could be quantitatively dominant in molluscs belonging to different clades. However, the dominant sperm morphs were the same in COI clade I and COI clade III. Thus, dominant sperm morphs are useless as species-specific traits. However, shell color and sperm parameters are specific to different geographic regions, and it seems that unique environmental factors can determine shell color and sperm morphology. As a result, both shells and spermatozoa can be used to distinguish the geographical forms of M. chinensis, regardless of the belonging of the forms to a particular genetic clade. Here we propose the introduction of geographic identifiers, in which the shell color and parameters of sperm sets are used as morphological criteria to determine the geographical origin of mollusc specimens belonging to the M. chinensis species complex.

Introduction Reunov et al. Helgol Mar Res (2021) 75:7 chinensis Philippi, 1846, plays an important role in the production of seafood [11,12]. In addition, M. chinensis is a rich source of keratin sulfate, glycosaminoglycan, useful for stimulating neuronal development, nerve growth and repairing damaged areas of the human nervous system, as well as for preventing Alzheimer's disease [13,14].
For many years, the identification of M. chinensis was carried out only on the basis of morphological characters. The first test of the nucleotide sequence of the mitochondrial COI gene, carried out at the Canadian DNA Barcoding Center (Canada), showed that there are three clades of M. chinensis, such as clade COI I, clade COI II and clade COI III, which constitute separate species inhabiting the Russian, South Korean and Chinese parts of the Asia-Pacific region [15]. A second test, carried out at the Institute of Oceanology of the Chinese Academy of Sciences (China), confirmed this conclusion (present report). Thus, M. chinensis is thought to be a complex of three species. Since there are no new names for these new species yet, we preliminarily characterized them as M. chinensis COI clade I, M. chinensis COI clade II, and M. chinensis COI clade III.
The existence of three species implies the possibility of interspecies differences in pharmacological value. Therefore, an accurate definition of the species is important. For greater convenience, it would be important to find out whether the genetic varieties of M. chinensis differ morphologically.
It has been reported that shells of different colors can be found in northern and southern populations of M. chinensis in the Sea of Japan [15]. With this in mind, it would be logical to find out if there are three variants of shell coloration, each of which is typical for a particular species descended from M. chinensis. It would also be interesting to know if more than three shell coloration can be found among the M. chinensis species complex. Depending on how many color options are found, two possible reasons can be discussed that are likely to affect the shell color.
After the discovery of three genetic species of M. chinensis, it would be tempting to suggest that the plasticity of shell coloration may be associated with intraspecific divergence. A similar phenomenon was reported for M. coralline, in which the presence of two species which respectively have a white shell and a shell with brown rays was clearly confirmed by the analysis of genes 12S, 16S, 18S and COI [16]. However, it is also known that environmental factors can influence the coloration of mollusc shells [17]. Thus, more research is needed to understand which factor-intraspecific genetic divergence or environmental influences-is critical for determining the shell color of M. chinensis.
In addition to shells, an analysis of the structure of gametes can be used to detect divergence of the species. It has been established that, compared with the organism as a whole, reproductive cells undergo a faster rate of evolution and, therefore, different sperm morphology found in remote geographical populations of the species is a sign of probable divergence and speciation [18]. In the study of male gametes in M. chinensis, living in the southern and northern regions of the Sea of Japan and belonging respectively to M. chinensis COI clade I and M. chinensis COI clade II, it was found that each male has not one, but three morphological sperm patterns. Moreover, different sperm morphs were quantitatively dominant in representatives of COI clade I and clade II. It has been hypothesized that genetic divergence of species may be responsible for a shift towards a predominance of one of the sperm variants [15]. Given that M. chinensis is genetically divided into three clades, a more detailed analysis of spermatozoa in samples representing M. chinensis COI clade III is needed to test whether male gametes may be specific in this clades.
The aim of this article was to analyze the shell coloration and structure of spermatozoa of three species derived from the marine bivalve mollusc M. chinensis. For shells, to test whether it is possible to distinguish between species by the shell color specific to each species, our experiments consisted of the following: (1) to find out if there are three types of shell coloration, each of which corresponds to one of the three genetic clades; (2) examine the coloration of shells in museums and examine the shells of all available specimens to see if more than three coloration can be found; (3) determine if more than one shell color can correspond to the same genetic clade; (4) use scuba diving to study the molluscs in their natural habitat and check if the coloration of the shells is consistent with the local environment. As for spermatozoa, in order to test whether spermatozoa can be used to distinguish between species derived from M. chinensis, we examined the structure of male gametes in specimens belonging to the species M. chinensis COI clade III, and combined this with data obtained earlier for species M chinensis COI clade I and M. chinensis COI clade II. We hope that our observations will help to draw a conclusion about the applicability of shells and spermatozoa as morphological characters suitable for distinguishing the forms of M. chinensis inhabiting the Asia-Pacific region.

Sample collection
Samples of the bivalve mollusc Mactra chinensis Philippi, 1846, genetically corresponding to the species M. chinensis COI clade I and M. chinensis COI clade II, were collected in the "Russian" and "South Korean" regions of the Sea of Japan according to the protocol that was previously described [15]. Samples genetically consistent with M. chinensis COI clade III were collected by a supplier in the Yellow Sea of China near Yantai City (exact location unknown) and delivered by the supplier to a fish market in Yantai City, where they were identified as M. chinensis and purchased by Dr. K. Lutaenko on July 7, 2018.

Study of shell coloration in museum collections
We studied samples of M. chinensis from the collection of the Zoological Museum of the Far Eastern Federal University (Vladivostok, Russia), collected by Dr. K. Lutaenko, as well as samples from the collection of the Marine Biological Museum of the Chinese Academy of Sciences (Qingdao, China), collected by Dr. J. Zhang.

Investigation of M. chinensis shell coloration by SCUBA
M. chinensis specimens were collected from Peter the Great Bay (Sea of Japan) in September 2020 by Dr. Yana Alexandrova. SCUBA work was done by professional marine biology divers of NSCMB FEB RAS (Vladivostok, Russia). The samples were obtained by manually loosening the bottom substrate at a depth of four meters at the sites of holes, presumably the outlets of siphons of inhabitants, which sometimes turned out to be M. chinensis. Coordinates were identified using manual profiler Cast Away ctd. (SonTek, USA). The shells were delivered to the laboratory, dried and photographed by Dr. Yana Alexandrova under the same lighting conditions.

Transmission electron microscopy
M. chinensis is a dioecious species. Cases of hermaphroditism are unknown. There is no sexual dimorphism in the color and morphology of the shells. We identified males by detecting sperm in cell suspension using light microscopy. Male gonads were dissected, cut into small pieces and fixed overnight in primary fixative containing 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) with osmolarity 1100 mOsm adjusted by sodium chloride. Fixed tissues were washed in buffer, postfixed in 2% OsO 4 in sea water, rinsed in 0.1 M cacodylate buffer and distilled water, dehydrated in an ethanol series, infiltrated and embedded in Spurr's resin. Ultra-thin sections were mounted on slot grids that were coated with formvar film. Sections were stained with 2% alcoholic uranyl acetate and Reynolds lead citrate and investigated using a Zeiss Libra 120 transmission electron microscope (Carl Zeiss, Oberkochen, Germany) and a Philips 410 transmission electron microscope (Philips 123 Electronics, Eindhoven, Netherlands).

Scanning electron microscopy
As for transmission electron microscopy, males were identified by examining the cell suspension using light microscopy. Sperm suspension was collected, pipetted onto a Thermanox coverslip (Cat. #72280) and allowed to settle for 5 min. Coverslips with attached sperm cells were fixed overnight in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) with osmolarity 1100 mOsm adjusted with sodium chloride. Primary fixed materials were washed in 0.1 M cacodylate buffer (pH 7.4) and postfixed in 1% OsO 4 in sea water. After the following washing in buffer, the samples were rinsed in distilled water, dehydrated in a graded series of ethanol solutions, transferred to acetone and critical-point dried in CO 2 . Dried materials were mounted onto aluminum stubs, coated with gold, and examined with a scanning electron microscope LEO-430 (Horus Tech Inc., USA).

Quantitative analysis of sperm morphs
Two samples were taken from three geographic areas such as (i) the Sea of Japan "Vostok" Biological Station (Russia), (ii) the Sea of Japan, Gyeongsanbuk Province, Uljin County (South Korea), and (iii) the Yellow Sea near city Yantai (China). 500 sperm cells of each sample were evaluated, thus 1000 cells were studied from each geographic area, 3000 sperm cells altogether. Sperm phenotypes were identified by scanning electron microscopy, and the frequency of each phenotype was calculated. The results were analysed by the Microsoft Excel program. All values are expressed as means with standard error of the mean (SEM). Percentages were calculated using the Student's t test and *P < 0.05, **P < 0.001 were considered statistically significant.

Genetic analysis
The COI gene analysis was originally carried out at the Canadian DNA Barcoding Center (Canada), and its methodological description can be found in our previous publication [15]. The second test was carried out at the Institute of Oceanology, Chinese Academy of Sciences (China). The protocol is shown below.
Standard DNA barcoding protocols were followed at all stages of the analysis, including DNA extraction, PCR and DNA sequencing. A piece of the mantle or muscle (3-5 mm 3 ) was taken from three molluscs living in the Russian part of the Sea of Japan, from two molluscs living in the South Korean part of the Sea of Japan and from eight molluscs living in the Yellow Sea (China). Tissue samples were then subjected to overnight lysis in CTAB buffer with proteinase K (Invitrogen) followed by DNA extraction on a glass-fiber membrane (PALL) using automated protocol. A 658-bp barcoge region of the mtDNA gene COI was amplified using primer cocktails C_Lep-FolF + C_LepFolR and C_LepFolF + SipR1 (LepF1, ATT CAA CCA ATC ATA AAG ATA TTG G and LepR1, TAA ACT TCT GGA TGT CCA AAA AAT CA; LCO1490, GTC AAC AAA TCA TAA AGA TAT TGG and HCO2198, TAA ACT TCA GGG TGA CCA AAA AAT CA, SipR1, TAA ACT TCT GGR TGR CCA AAR AAY CA).
ABI trace files for each specimen were assembled into contigs using Codon Code Aligner software. In addition to the 13 consensus sequences generated, DNA sequence data for M. chinensis and Mactra sp. collected from the Yellow Sea and the East China Sea [19,20] and for some additional congeneric species in published papers [21,22] were downloaded from Genbank and the BOLD Public Data Portal. The final dataset of 121 partial COI sequences was aligned by eye in BioEdit. All original specimen metadata, sequences, and trace files were uploaded to the Barcode of Life Data Systems (BOLD) at http:// www. bolds ystems. org or GenBank at https:// www. ncbi. nlm. nih. gov/ genba nk/.
To discriminate species, the Automatic Barcode Gap Discovery (ABGD) analysis [23] were used to scan a range of prior intraspecific divergence from P min = 0.001 to P max = 0.1, with 10 steps, 1.5 for the minimum relative gap width, and using Jukes-Cantor (JC69) and Kimura (K80) measure of distance. Nucleotide sequence divergences were also calculated in MEGA X [24] using the Kimura two-parameter (K2P) nucleotide substitution model [25]. Sequences that were split into distinct groups with clear barcoding gaps are hypothesized to represent provisional species. The best-fitted evolutionary models were selected by AICc as implemented in jModel-test2 [26]. The general time-reversible substitution with discretized gamma-distributed rate and invariant sites model (GTR + G + I) was selected as the best-fitted model. Bayesian inference (BI) was conducted using the software MrBayes 3.2.6 [27]. Posterior probability (PP) was estimated using four chains running 10,000,000 generations sampling every 100 generations. Maximum likelihood (ML) analysis was carried out using RAxML 8.1.2 [28] (with bootstrapping) using GTR + G + I as the model. Node support came from the bootstrap scores (BS) in the best-scoring tree of 1,000 bootstrap replicates. The neighbor-joining phylogenetic tree [29] was inferred in MEGA X [24] using theK2P method [25]
Multiple polymorphic sites were detected in the alignment of sequence data for analyzed samples of the M. chinensis collected in the northern and southern areas of the Sea of Japan and in the Yellow Sea. No insertions or deletions have been detected. All nucleotide substitutions were silent (i.e., not resulting in any amino acid changes) and were observed largely in third codon positions. For the analysis of nucleotide  sequences, we used DNA barcoding of the materials collected by us and the data retrieved from Genebank. Total of 121 partial COI sequences were used for genetic analysis ( Table 1). The ABGD analysis based on the COI alignment using the Jukes-Cantor (JC69) and Kimura (K80) models yields 10 distinct groups, which represent 10 candidate species (Fig. 2). Pairwise K2P distances of COI within the 10 groups are much lower (0-0.0206) than those between groups (0.0820-0.2894) ( Table 2). There is no overlap but clear barcoding gap between intraand inter-specific distances of all candidate species. That support the validity of these 10 candidate species.
The phylogenetic trees established using BI, ML and NJ analyses are generally consistent. Thus, a single topology is presented with support values indicated on branches (Fig. 2). In the trees, the 10 species form separate clades, with M. chinensis splits into three clades. The first clade is represented by haplotypes shared by M. chinensis specimens from the northern part (Russia) of the Sea of Japan (YARRA074-12/RRYA-74; YARRA075-12/RRYA-75; YARRA073-12/RRYA-73) and Mactra sp. samples from the northern part of the Yellow Sea (individuals CX2, CX3, and CX4 in [19]). The second clade includes haplotypes shared by the specimens from the southern part (South Korea) of the Sea of Japan (YARRA072-12/  RRYA-72; YARRA071-12/RRYA-71) and Mactra sp. samples from the Yellow Sea (CX1 in [19].  1-48 in [21]) and specimens from the East China Sea (including specimens Z1 and Z6 in [29]; Haplotype 49-57 in [20]. Thus, the analysis confirmed the existence of three COI lines of the species M. chinensis.

Shell coloration of museum samples
Work in the museum collections showed that other shell colors are present in M. chinensis. For example, the malacological collection of the Zoological Museum of the Far Eastern Federal University (Vladivostok, Russia) contains shells from the Ussuriisky Bay (Sea of Japan, Russia), which have brightly visible violet-purple rays against a light background, and the edge of the shell also has a purple color (Fig. 3-4). Shell samples collected in the Gyeonsang area of South Korean part of the Sea of Japan are devoid of any rays and have pale yellow tones (Fig. 3-5). Dr. J. Zhang's collection contains shell specimens collected in the Yellow Sea of China, which are represented by two variants from Qingdao, one of which is devoid of any rays and has a pale pink tint (Fig. 3-6), and another has very light ochroid tone with slightly visible stripes (Fig. 3-7). In the same collection, one can see two variants of mactra from Weihai, which are distinguished by a dark and pale ocher shade ( Fig. 3-8, 9). The same collection has four variants of mactra from the Bohai Sea in China, which are represented by two variants from Qinghuangdao, which have a pearlescent hue and differing intensity of ocher stripes (Fig. 3-10, 11), and two very different variants from Beidaihe, one of which is specimen showing irregular ocher tone and another white with ocher stripes on it (Fig. 3-12, 13). Also, the collection of Dr. J. Zhang allows to see two samples of M. chinensis collected in the East China Sea near Pingtan (China). One of these is stripe-less with very light ochroid tone and another has pale ochroid stripes on the pearly background ( Fig. 3-14,  15).

Investigation of the shell color in samples belonging to the COI clade III
In this part of the study, we used data of Dr. J. Zhang to test if the shell colors are similar in M. chinensis specimens belonging to the same COI clade (sperms were not examined). Samples collected in the Yellow Sea near Yantai (ABCBF460-19, ABCBF461-19, ABCBF462-19,   Fig. 2). Shell coloration of all samples was morphologically compared. Morphologic analyses of the Yantai's M. chinensis samples showed that their shells have characteristic rays radially extending from the beak, and colored in mustard-like overall tone having slight intensity variations between the samples (Fig. 4-1-4). Samples collected near Rushan show variations in brownish tone, which can be bright (Fig. 4-5) or relatively pale (Fig. 4-6-8), as well as vary from a very weak tone of the rays (Fig. 4-5, 6) to relatively bright rays (Fig. 4-7, 8). In any case, each group of specimens is specific in terms of overall hue, and shells from Yantai and Rushan can be classified into shells having a mustard tone and shells having a brownish tone.
In the collection point #1 (Fig. 5-1), the water temperature and salinity of sea water were 19.99 °C and 31.75‰ respectively. Some shells have damaged periostracum (Fig. 6A-1) but another shells appear normal ( Fig. 6A-2, 3). Shells usually have stripes (Fig. 6A-1-3). The central shell areas are stained in either dark brownish-terracotta pigment ( Fig. 6A′-1), or in slightly red terracotta pigments ( Fig. 6A′-2, 3). Periferal areas of shells are stained in the same brownish-terracotta colour but appear more light comparatively to the central areas and are peculiar in having contrasted stripes (Fig. 6A′-1′-3′). Most typical brownish-terracotta pigment is represented by a periostracum piece taken from the central shell area of one of the mollucs (Fig. 6A′′). The sand of the sea bottom substrate was found to have a grayish-brown tone (Fig. 6A′′′).
In the collection point #2 (Fig. 5-2), the water temperature and salinity were 21.28 °C and 31.91‰ respectively. The mactra samples collected have shells which periostracum either have no stripes, or have faint stripes. In most samples a shell surface appears strongly damaged. However, some periostracum areas turned out intact and appear as stained with very dark colour (Fig. 6B-1-3). It was found that this colour is represented by dark brown pigment (Fig. 6B′-1-3). Periferal areas of the shells have similar tone but appear more ligh (Fig. 6B′-1′-3′). A typical dark brown pigment is represented by a piece of periostracum taken from  the central region of the shell of one of the molluscs (Fig. 6B′′). The sand of the sea bottom substrate has a dark gray tone (Fig. 6B′′′).
In the collection point #3 some samples have shells which central parts are stained by very bright foxy pigment ( Fig. 6C-1, C′-1). The central areas of the other samples appear not so red and show variants of brown-foxy pigment (Fig. 6C-2, C′-2 and C-3; C′-3). Periferal areas of all shells found in the collection point #3 appear creamy (Fig. 6C′-1′-3′). We reckon that foxy pigment is most specific for molluscs found in collection point #3 (Fig. 6C′′).
In the collection point #4 some samples have shells which central parts are stained by creamy colour having slight foxy hue (Fig. 6D-1, 2 and D′-1, 2). We reckon that this colour is most specific for molluscs found in collection point #4 (Fig. 6D′′). Periferal areas of all shells found in the collection point #4 appear creamy (Fig. 6D′-1′-3′) and are very close to the colour that is found in peripheral areas of shells found in the collection point #3 (Fig. 6C′-1′-3′). Some samples from the collection points #4 has dark cream tone in both peripheral and central areas (Fig. 6D-3 and D′-3, 3′) but we are not sure that this tone is natural.

Sperm structure analysis
Spermatozoa of M. chinensis specimens, representing the three COI clades, look the same at low magnification of scanning electron microscope. All have compact heads and thin flagella (Fig. 7A-C). Analysis of ultrathin sections by transmission electron microscopy showed that mactra sperm can have different acrosomes, the structure of which is presented in two variants. In the first variant, the acrosome consists of two layers, one of which (anterior) is electron-light, and the other (basal) is electrondense ( Fig. 7D-F). In another embodiment, acrosomes have a central electron-dense rod that has an intracellular part inside of the acrosome and external part protruding above the acrosome (Fig. 7G-I). The nucleus always has an oval shape and is filled with electron-dense chromatin (Fig. 7D-I). In the middle part, centrioles are located perpendicular to each other. One of the centrioles is the basal body of the flagellum. Centrioles are surrounded by a mitochondrial ring consisting of four mitochondria (not shown).
Using SEM, it was found that the sperm sets of all studied M. chinensis specimens were similar in terms of the presence of common morphological variants. There are spermatozoa with concave acrosomes in which there are no protrusions, as was established when examining the acrosomal surface (Fig. 8A). Such spermatozoa with acrosomes without protrusion have been found in molluscs belonging to each of the three COI clades (Fig. 8E, I, L). In some spermatozoa, the acrosomes have a protrusion extending from the center of the acrosomal fossa ( Fig. 8B-D, F-H, J-K, M-N).
Calculations performed by analyzing the data obtained using scanning electron microscopy showed that there are three categories of axial rod length. M. chinensis representing each of the three clades have spermatozoa with acrosomes, the axial rod of which has a length 0.4 μm (Fig. 8F, J, M), and 0.6 μm (Fig. 8G, K,  N). A specific feature of the mactra collected in the Yellow Sea and belonging to COI clade III is the presence of the same three patterns and a fourth sperm pattern with an acrosome, the axial rod length of which reaches 0.8 μm (Fig. 8D, H). Here we characterize sperm patterns as SPERM 1, SPERM 2, SPERM 3, and SPERM 4, respectively.
Quantitative analysis showed that the proportions of sperm samples vary. In COI clade I, the most typical (72%) is SPERM 1. A smaller amount (18%) falls on SPERM 2, and a smaller amount (10%) belongs to SPERM 3 (Fig. 9A). In COI clade II, only 15% falls on SPERM 1. The rest of the spermatozoa is represented by morphs with the numbers of 21% and 64% for SPERM 2 and SPERM 3, respectively (Fig. 9B). In COI clade III, the proportions are presented as 45%, 19%, 25% and 11% for SPERM 1, SPERM 2, SPERM 3 and SPERM 4, respectively (Fig. 9C).

Genetic divergence of mactrids is a common phenomenon
As shown above, there are three M. chinensis genetic lines living in the Asia-Pacific region, specimens of which have been found in the Sea of Japan and the Yellow Sea [15]. Also Ni et al. [20] reported that genetic subdivision occurs in populations of M. chinensis in the East China Sea. It is also known about the intraspecific divergence of Mactra coralline inhabiting the northern Adriatic coasts of Cesenatico (Italy) [16]. The subdivision of the population into two distinct clades was shown for M. coralline habituating in the Gulf of Tunis (northern Tunisia) [21]. Thus, genetic divergence is a common occurrence in mactrids. Mactridae genetic divergence can be triggered by factors such as rivers' outflow, environmental gradient factors and life-history traits [20,21]. In addition, local hybridization can cause divergence, as shown for other bivalve molluscs [30,31]. It is obvious that the hybridization process is capable of inducing species collapse [32].

The intraspecific variety of shell coloration exceeds the number of detected species
According to the analysis of shell coloration in genetically tested samples, the coloration of the shell was specific in the representatives of the "Russian", "South Korean" and "Chinese" M. chinensis that belong correspondingly to the COI clade I, COI clade II, and COI clade III. Based on this, it would be tempting to speculate that shell coloration follows genetic divergency. However, a study of museum specimens showed that M. chinensis is characterized by a wide variety of shell coloration. Museum specimens have never been genetically tested, and the belonging of their shells to one of the discovered species derived from M. chinensis could not be genetically confirmed. However, assuming that morphological criteria have always been considered a reliable method for M. chinensis identification, we believe that all shells belong to this species complex. Taking into account that the number of types of shell coloration clearly exceeds the number of species diverged from M. chinensis, we are confident that none of these species can be distinguished from a single shell pattern.

Individuals belonging to the same genetic clade may have shells of different colors
Further evidence of the impossibility of distinguishing species derived from M. chinensis by the only speciesspecific shell pattern was obtained by genetic analysis of samples from Yantai and Weihai, which have "mustard shells" and "brownish shells", respectively. It turned out that both groups of individuals belong to the same clade (COI clade III) or to the same species M. chinensis COI Percentages were calculated using the Student's t test and *P < 0.05, **P < 0.001 were considered statistically significant clade III. Thus, it appears that belonging to the same genetic lineage does not mean uniformity of shell coloration patterns. It seems likely that shell coloration is not optimized by intraspecific genetic divergence, but is regulated by other factors.

The optimization of shell color appears to be the result of environmental influences and gene work
There is evidence that environmental factors influence shell coloration [33,34]. For example, the geographic distribution of different types of limpet species and the specificity of their shell coloration are determined by the geographic distribution of algae, which are the main dietary component of these molluscs [35]. In Littorina saxatilis, salinity exerts an influence on shell colours [36]. Adaptive plasticity of the color of the shell is known in gastropod Concholepas concholepas, which adapt it to the color scheme of its victims [37]. Also, the ratio of calcium to magnesium in sea water may affect shell colour hue [17]. The bivalve mollusc Enigmonia aenigmatica adapts the shell design to the color of the substrate for camouflage from predators, and it is likely that the substrate affects the color immediately after the larvae have settled [34,38,39]. It has been experimentally shown that molluscs are able to change shell color when switching to a new substrate in order to obtain a shell color that better matches the new substrate, and these data support the existence of ecological control of shell coloration [35,40].
According to some reports, environmental factors can regulate the color of the shell through the work of genes [41][42][43][44][45][46]. The influence of environmental factors on the genome is mediated by epigenetic mechanisms such as DNA methylation, histone modification, and microRNA expression, which are capable of altering gene expression patterns and causing changes in phenotype [47,48]. Various types of allelic increase or decrease in gene expression are also associated with epigenetic mechanisms, the activity of which can be determined by external influences [49]. In the bivalve mollusc Macoma baltica, four alleles at one locus provide the expression of shell color from white to yellow, orange and red [50]. In Pacific oysters, shell pigmentation is controlled by two genetic loci with two alleles that are responsible for the secretion and distribution of pigments. In addition, one independent locus with two alleles controls striped pigmentation [51]. Thus, it seems possible that the optimization of shell color is the result of the combined work of environmental factors and epigenetic mechanisms.

Shell color is specific and uniform at each collection point
To check if effect of environments on shell coloration could be found in M. chinensis we conducted SCUBAstudy in four areas of The Peter the Great Bay (Japan Sea) that is known by various environmental conditions [52]. The shell periostarcum of some samples looked damaged but this seems to be normal. Indeed, over time, periostracum decreases in thickness and undergoes erosion being affected by environmental abiotic and biotic factors [53]. However, we were able to find undamaged samples, or we used undamaged pigment areas of partly damaged samples.
We investigated both the central and marginal shell areas. It was often found that marginal shell areas look lighter than central ones. Although in some samples both the central and peripheral areas were almost similar in terms of color density. Anyway, assuming that marginal periostracum is represented by growing area which color is unstable [54], we reckon that exactly central part has mature color.
Remarkable, in the collection points #1 and #2 the samples have a hue that is specific in each collection point. We suggest that shell uniformity found in these niches occurs because of stability of ecologic parameters characteristic for each of these niches. In the points #3 and #4 shell hue was also specific for each collection zone. However, some tone overlap was found in the samples living in these areas. Probably, some shells' colour mixture may occur along the borders of the neighboring niches. Indeed, both collection points (#3 and #4) are situated very close to each other and their borders overlap. Anyway, although more extended SCUBA study is needed to identify the sizes and borders of the niches, we find it possible to speculate that shell color could be considered as a marker specific for each ecological niche that also could be considered as geographical area. It seems likely that the coloration of the shell is the same in a geographic area, the size of which is determined by the zone of action of the ecological features characteristic of the area.
According to our study using SCUBA, the Peter the Great Bay cannot be considered as a single ecological niche inhabited by M. chinensis with a single shell coloration. We believe that various ecological niches coexist in the bay and the color of the shells is specific for the niche and is uniform in each niche. Based on the available data, we can distinguish at least six color morphs of M. chinensis in the Peter the Great Bay (Fig. 10). However, it is entirely possible that other shell coloration variants could be found.
We found that the color of the substrate was also specific in each area, and noticed that the light/dark color of the substrate corresponded to the light/dark shade of the shells. It was also observed that the presence of brown substrate particles coincided with the brown color of the shell, and the presence of black substrate particles coincided with a darker brown color. This observation suggests the existence of a masking ability of the M. chinensis shell color in relation to the substrate.

The quantitative parameters of heterogeneous sets of spermatozoa are specific for each collection site
We also examined whether the structure of the spermatozoa can be used to recognize species derived from M. chinensis. By the method of transmission electron microscopy, it was found that the ultrastructure of the spermatozoa of the studied samples corresponds to the ultrastructure of the mactra spermatozoa described in previous studies [55][56][57]. However, our study by transmission electron microscopy provided more detailed data on the ultrastructure of acrosomes. In the sperm of M. chinensis derived species, we found two types of acrosomes, namely, acrosomes without an acrosome protrusion and an acrosome with an acrosome protrusion.
Using scanning electron microscopy, allowing evaluation of external cell shape, we found that in the samples belonging to COI clades I and II, the sperm morphotypes are represented by three variants. These are SPERM 1sperm with acrosomes without axial rod, SPERM 2sperm with acrosomes having axial rod with a length corresponding to 0.4 μm, and SPERM 3-sperm with acrosomes having axial rod with a length corresponding to 0.6 μm. Samples belonging to COI Clade III have the same three sperm morphs. In addition, the unique SPERM 4 with an acrosome with a very long axial rod (0.8 μm) was also found in samples of the third clade.
Based on the fact that species derived from M. chinensis have more than one sperm morph, these species can be classified as a species with heteromorphic sperm morphology. Interestingly, sperm heteromorphism was recently discovered in another bivalve mollusc, the Pacific oyster Crassostrea gigas, whose heteromorphic sperm set consists of six morphologically stable morphs [52]. Targeted research is needed to test whether sperm heteromorphism can be found in other bivalve molluscs.
In animals, the phenomenon of intraspecific heteromorphism of spermatozoa is known [58]. The functional parameters of heterogeneous spermatozoa can vary, for example, higher speed-shorter life span and lower speed-higher life span [59]. There is also a hypothesis that spermatozoa of different phenotypes differ in their allelic content, which ensures differences in sperm competition [60]. In C. gigas oysters, in which reproduction occurs by external fertilization in seawater, the causes of sperm plasticity are associated with reproductive adaptation to the aquatic environment, which can be influenced by anthropogenic pollution, intense water current or turbulence, unstable temperature and salinity. The dominant expression of one or another variant of spermatozoa may be associated with a greater degree of adaptability of this variant to a certain type of environment. Parallel expression of additional abundant sperm variants may increase the chances of reproductive success in oysters in unstable conditions [52]. Since M. chinensis also has external fertilization a link between environmental conditions and sperm shape seems possible in this species.
Since each of the three M. chinensis-derived species have heteromorphic spermatozoa, the validity of these species cannot be confirmed using only one variant of the sperm. However, given that different sperm samples dominated in M. chinensis COI clade I and M. chinensis COI clade II [15], we hypothesized that the predominant sperm samples could be used as species-specific traits.
To check this hypothesis we tested if M. chinensis belonging to the COI clade III may have specific dominant sperm morph. This version seemed very likely, given that intraspecific genetic divergence is thought to be accompanied by divergence of sperm structure, as has been shown for some marine invertebrates [61][62][63]. However, the dominant sperm variant of M. chinensis COI clade III was not specific, but coincided with the dominant variant of spermatozoa in the samples belonging to M. chinensis COI clade I.
Thus, it can be concluded that the dominant morphs of spermatozoa also cannot be used to identify species descended from M. chinensis. However, it should be emphasized that the quantitative parameters of sets of spermatozoa are specific in the Russian, South Korean and Chinese regions. We assume that the parameters of sperm sets are individual in each of the geographic zones, each of which presumably represents an ecological niche that determines the specificity of sperm sets. In other words, we believe that the morphological features of M. chinensis male gamete sets are geographically specific.

Geographic identifiers of M. chinensis samples can become a means of geo-authentification (= quality control) in the seafood market
In medicinal plants, the type, content and proportion of active substances can vary depending on environmental factors in the areas of plant growth, and, therefore, the geographical identification of plant materials is a necessary measure to control its pharmacological quality [64,65]. The geographical dependence of the nutritional and medicinal value of species derived from M. chinensis, as well as other bivalve molluscs, has not yet been investigated. If future research reveals the dependence of the clam properties on the collection site, the introduction of geographic identifiers will be an effective means of controlling the quality of samples provided by suppliers. Taking into account that the coloration of shells and the structure of spermatozoa in species derived from M. chinensis are geographically specific, we propose to use these characters to determine the geographic forms of these species.
We propose geographic identifiers that include information such as: (1) the scientific name of the species (Fig. 11A-1, B-1, C-1), (2) the conventional name of the geographical form of the species (Fig. 11A-2, B-2, C-2), (3) the geographical name of the habitat (cultivation) (Fig. 11A-3, B-3, C-3), (4) geographical coordinates ( Fig. 11A-4, B-4, C-4), (5) an image showing the color of the shell (Fig. 11A-5, B-5, C-5), (6) images showing sperm morphs (Fig. 11A-6, B-6, C-6), (7) proportional ratio of sperm morphs with an indication of the dominant type of sperm ( Fig. 11A-7, B-7, C-7), (8) a scientific source on the basis of which an idea of the geographical form of the species was created (Fig. 11A-8, B-8, C-8). Using these identifiers, it will be possible to (i) determine the geographical origin of the molluscs by simply comparing the colors of the shells, and (ii) control the geographical origin through sperm examination, which can be performed in an electron microscopy laboratory. An expanded study of the M. chinensis species complex  A-7, B-7, C-7), (8) scientific source on the basis of which an idea of the geographical form of the species was created (A-8, B-8, C-8). Note the red square in Section 6, which indicates the sperm morph that dominates in the geographic form of the mollusc in the Asia-Pacific region is needed to provide detailed identifiers that will become a reliable tool for tracing the origin of these valuable molluscs in seafood markets.

Conclusions
When testing the mitochondrial COI gene, it was revealed that the bivalve mollusc M. chinensis, which lives in the Asia-Pacific region, diverged into three species. It was found that these species cannot be reliably determined based on analysis of the shell coloration and sperm structure due to the plasticity of these morphological characters. However, since the coloration of the shells and the quantitative parameters of sperm sets are geographically specific, the geographic identification of M. chinensis forms seems possible using these characters. It is proposed to introduce geographical identifiers to control the geographical origin of the clams in seafood markets.