RESEARCH ARTICLE
Taxonomic revision of the unionid bivalves from the Lower Cretaceous Tetori Group in Japan with reference to their habitats in meandering river systems
2025 年 29 巻 p. 204-215
詳細
2025 年 29 巻 p. 204-215
Unio ogamigoensis, the first unionid bivalve reported from the Tetori Group, is redescribed as Margaritifera ogamigoensis on the basis of the hinge teeth and muscle attachment scars of the holotype and newly added specimens from the Kuwajima Formation of the Tetori Group, central Japan. This species occurs within crevasse splay sandstones that are intercalated with the fluvial floodplain deposits of the meandering river system. Nakamuranaia kagaensis, which is commonly found in muddy fine-grained sandstone and mudstone in the overbank deposits, is also redescribed from Archaeounio kagaensis, i.e., Archaeounio becomes a synonym of Nakamuranaia. The mode of occurrence of the two species indicates that M. ogamigoensis was a rapidly flowing stream dweller, whereas N. kagaensis inhabited stagnant shallow oxbow lakes.
The Tetori Group, distributed in the Hokuriku and Hida regions of central Japan, is one of the representative Early Cretaceous clastic deposits in Japan. Nonmarine molluscan fossils, especially unionid bivalves, are useful as indicators of freshwater depositional conditions and are considered to be good keys to classify and correlate the formations within the Tetori Group (e.g. Maeda, 1961; Tamura, 1990; Matsukawa et al., 2003; Sano, 2015) and other Cretaceous deposits in East Asia (e.g. Yang, 1979; Matsukawa et al., 1998; Sha et al., 2012; Sano, 2015).
Unio ogamigoensis Kobayashi and Suzuki, 1937 was the first unionid bivalve from the Tetori Group to be described. The holotype of this species was excavated at Ogamigo, Shokawa, Takayama, Gifu Prefecture, but no detailed information about the fossil locality, stratigraphy or co-occurring fossils is available. Subsequently, specimens assigned to U. ogamigoensis have been reported from various localities in the Tetori Group and other formations and were used for geological correlation of freshwater strata (Maeda, 1962; Tamura, 1990; Matsukawa and Ido, 1993; Kozai and Ishida, 2003; Isaji et al., 2005; Kozai et al., 2005; Yamashita et al., 2011). However, most records of these unionid fossils in the literature lack taxonomic descriptions and detailed comparison with the holotype. Thus, only the holotype, so far, actually represents U. ogamigoensis.
This paper reports the results of a taxonomic re-examination of U. ogamigoensis on the basis of the holotype and two recently collected specimens considered conspecific with this species. In addition, other unionid species, Archaeounio kagaensis Koarai and Matsukawa, 2016, that had been confused with U. ogamigoensis are examined, and the differences in habitat inferred from the depositional environments of the two species are discussed.
The Tetori Group was established by Maeda (1961) for the Mesozoic clastic deposits distributed in the Hokuriku and Hida regions of central Japan (Figure 1A). Later, the Tetori Group in Maeda’s (1961) sense was subdivided into the Middle to Upper Jurassic Kuzuryu Group and the unconformably overlying Lower Cretaceous Tetori Group (sensu stricto) (Sano, 2015; Yamada and Sano, 2018).
The lower part of the Tetori Group (s. s.), yielding numerous plant and animal fossils, is assigned to the Kuwajima Formation in the Shiramine area of Ishikawa Prefecture, the Okurodani Formation in the Shokawa area of Gifu Prefecture and the Itsuki Formation in the Itoshiro area of Fukui Prefecture (e.g. Maeda, 1961; Sano, 2015; Isaji, 2023). These formations are assumed to be equivalent mainly based on the co-occurrence of nonmarine molluscan fossils and show the transition of depositional facies from brackish to freshwater environments (Figure 1C) (Maeda, 1961; Okazaki and Isaji, 2008; Sano, 2015; Isaji, 2023). Neomiodontid and cyrenid bivalves indicate brackish conditions for the intertidal and shallow subtidal environments at the delta-front (Isaji, 2023). Viviparid gastropods and sphaeriid bivalves are major indicators of freshwater conditions in the floodplain of the meandering river system (Isaji, 2023).
The Kuwajima-kasekikabe site is stratigraphically located in the upper part of the Kuwajima Formation (Maeda, 1961) (Figure 1B, C). At the Kuwajima-kasekikabe site, the clastic deposits feature a combination of underlying very coarse-grained sandstones and overlying alternating beds of fine-grained deposits. These depositional facies concur with the generalized model for sedimentation in meandering rivers (e.g. Allen, 1965; Reineck and Singh, 1980; Walker and Cant, 1984); specifically, the former is regarded as a channel deposit, and the latter is regarded as an overbank deposit formed in a fluvial environment (Figure 1D).
The channel deposit consists of massive and very coarse-grained sandstones that usually contain numerous pebbles to cobbles of orthoquartzite, mud crusts and driftwoods at the base. The sandstone bed also has a sharp erosional base and is fining upward. The alternating beds of mudstones and fine-grained sandstones in the overbank deposits yielded numerous remains of aquatic animals, including molluscs, ostracods and vertebrates. Molluscs are characterized by a shallow-lacustrine viviparid-sphaeriid assemblage represented by Campeloma onogoense (Kobayashi and Suzuki, 1937) and Sphaerium coreanicum (Kobayashi and Suzuki, 1936). These alternating layers of fine-grained deposits are often intercalated with medium- to coarse-grained sandstones that exhibit lenticular bedding. These sandstones formed when a stream broke its natural levees and represent crevasse splays. The sandstones lack the viviparid-sphaeriid assemblage but very rarely yield the large-sized unionid bivalves described in this paper. In situ tree trunks, numerous rootlets and occasional coal occurring in the overbank deposits indicate that subaerial deposition of sediment continued long enough to allow the development of a well-vegetated floodplain (e.g. Ogura et al., 1951; Maeda, 1955, 1958; Taira and Matsuo, 1983; Masuda et al., 1991; Isaji et al., 2005).
The precise geological age of the Kuwajima Formation at the Kuwajima-kasekikabe site is unknown because this nonmarine strata are barren of reliable index fossils and are not interbedded with marine strata containing ammonites or other biostratigraphic markers. The underlying beds of the lower part of the Kuwajima Formation are dated to 130.7 ± 0.8 Ma via laser ablation inductively coupled plasma‒mass spectrometry (LA-ICP-MS) dating of zircons from the tuff beds at Setono in the Shiramine area (Matsumoto et al., 2006). The age of a tuff bed in the overlying Akaiwa Formation along the Osugidani River in the Shiramine area was determined as 121.2 ± 1.1 Ma via LA‒ICP‒MS U–Pb dating of zircons (Sakai et al., 2019). Thus, the Kuwajima Formation at the Kuwajima-kasekikabe site is considered to have been deposited somewhere in the period ranging from approximately 130 to 120 Ma and is roughly constrained to the Hauterivian to Barremian stages.
The holotype of Unio ogamigoensis Kobayashi and Suzuki, 1937, reexamined in this paper, was collected in the early 20th century from Ogamigo, Shokawa, Takayama, Gifu Prefecture (Figure 1A) and is now stored in the University Museum, the University of Tokyo (UMUT MM 07001). According to the specimen label, the type specimen must have been collected by Kenjiro Fujii (1866–1952), who was a professor in the Department of Botany at Tokyo Imperial University. Fujii was studying the Ginkgoaceae at the time, and he presumably found the specimen when he visited Ogamigo, where Ginkgo fossils had been reported by Yokoyama (1889).
Although the exact locality of the holotype is unknown, the depositional conditions of this unionid can be inferred from the sedimentary structures observable in the surrounding rock (Figure 2). Figure 2B shows many very small ripples (small arrows) with wave heights of less than 1 mm in the parallel sand laminae in the muddy matrix. These ripples are considered combined-flow ripples because of their asymmetric forms with rounded crests (Yokokawa, 1995). A small combined-flow ripple (white triangular arrowhead) with a wave height of approximately 10 mm can be observed in Figure 2C, D, along with parallel laminae and very small combined-flow ripples (small white arrow) in Figure 2D. The large white arrow in Figure 2A indicates the current direction estimated by flow ripples. The white dots compose the line connecting the crests of the small combined-flow ripples exposed on the fracture surfaces shown in Figure 2C, D.
Taking these sedimentary structures into account, the holotype seems to have been carried with both valves open and got stuck with the anterior part of the left valve convex up on the crest of a ripple. It is therefore likely that the specimen had been transported post-mortem, but because the valves are connected, it probably was not transported very far.
Two large specimens (>100 mm in length) with elongated outlines were collected from the medium- to coarse-grained sandstones excavated from the Kuwajima Formation as a result of the tunneling work at the Kuwajima-kasekikabe site in Hakusan, Ishikawa Prefecture (Figure 1B, D). These specimens were not collected directly from the outcrop but undoubtedly came from the alternating sandstone and mudstone beds in the overbank deposits of the meandering river systems (Isaji et al., 2005). These large specimens are very rare, so only two specimens are available. The two specimens are laterally compressed and are preserved with closed valves. They were found with the commissural plane parallel to the bedding plane of the sandstones. It is thus obvious that these two specimens were not preserved in life position within the rock and were transported from their original habitats.
Another medium-sized bivalve (<50 mm in length) with elliptical outlines is abundant in the very fine-grained sandstones and mudstones composing the overbank deposits in the Kuwajima-kasekikabe site. These bivalves commonly occur with closed valves and are often found in life positions. These specimens have been erroneously reported as Unio? ogamigoensis (Tamura, 1990; Matsukawa and Ido, 1993) and Unio ogamigoensis (Isaji et al., 2005) but have not been systematically described and belong to a different taxon. Other unionid specimens from Yanagidani, Hakusan (Figure 1A) are used in this paper for comparison. These specimens are topotypes of Archaeounio kagaensis Koarai and Matsukawa, 2016 (Loc. HD01 in Koarai and Matsukawa, 2016).
All the molluscan fossils were coated with ammonium chloride and photographed using a digital camera. Other than the holotype, the specimens from Hakusan discussed in this paper are stored in the Shiramine Institute of Paleontology (numbered with the prefix SBEI) and the Natural History Museum and Institute, Chiba (numbered with the prefix CBM PS). The classification follows Bouchet et al. (2010).
Class Bivalvia Linnaeus, 1758
Order Unionida Gray, 1854
Superfamily Unionoidea Rafinesque, 1820
Family Margaritiferidae Henderson, 1929
Genus Margaritifera Schumacher, 1815
Type species.—Margaritifera fluviatilis Schumacher, 1815: Recent.
Margaritifera ogamigoensis (Kobayashi and Suzuki, 1937)
Unio ogamigoensis Kobayashi and Suzuki, 1937, p. 41, pl. 4, fig. 16; Shikama, 1943, p. 181, pl. 19, fig. 18; Shikama, 1952, p. 114, pl. 19, fig. 18; Shikama, 1964, p. 132, pl. 43, fig. 21.
Unio? ogamigoensis Kobayashi and Suzuki, Hayami, 1975, p. 93; Koarai and Matsukawa, 2016, p. 142, fig. 22CC.
Materials.—UMUT MM 07001 (Holotype), SBEI 2472, SBEI 2473.
Description.—The shell is medium-sized for the genus and elongate and elliptical in shape. The shell is thick-walled, slightly inflated, equivalve and inequilateral. The anterodorsal margin is short and slightly arcuate, grading into a rounded anterior margin. The dorsal margin is broadly arcuate or somewhat straight, forming an obtuse angle with a straight posterodorsal margin. The ventral margin is long and rather straight, graduating to a posteroventral margin. The posteroventral margin is faintly arcuate, forming an acute angle (approximately 60–70°) with a posterodorsal margin at the posterior extremity (Figure 3C, K, L). The acute posterior extremity is not observed in any specimen (Figure 3E, H) due to poor preservation. The carina is distinct in a large specimen and extends from the umbonal region to the posterior extremity. The umbo is very low and situated at anterior one-fifth of the entire shell length. The ligament is moderately strong, occupying approximately 35% of the entire shell length. The nympha is clearly delimited by a slender ridge at the two-third of the dorsal margin from the umbo. The shell surface is ornamented with irregularly spaced concentric growth lines. The hinge plate is moderately long and broad and bears one pseudocardinal tooth (pct in Figure 3G) and one posterior lateral tooth (plt in Figure 3F) in the right valve and two pseudocardinal teeth (pct-a, pct-p in Figure 3J) and one posterior lateral tooth (plt in Figure 3I) in the left valve. The pseudocardinal tooth of the right valve is very stout and prominent and bears some grooves on the tip (pct in Figure 3G). The anterior pseudocardinal tooth (pct-a in Figure 3J) of the left valve is very small and weak. The posterior pseudocardinal tooth (pct-p in Figure 3J) of the left valve is similar to that of right valve in stoutness and also bears some grooves on the tip. The posterior lateral teeth of both valves are long and not prominent and originate around the umbonal region and terminate in front of the posterior margin (plt in Figure 3F, I). The anterior adductor muscle scar is large and semicircular in shape (aa in Figure 3). The anterior margin of the adductor muscle scar is arcuate and shallowly depressed and is located along the anterior shell margin. The posterior margin of the adductor muscle scar is sharply depressed and straight. The surface of the adductor muscle scar is very roughly ornamented by sharp grooves and ridges arranged antero-posteriorly and shallow striations along with the dorsal-ventral axis (white solid lines in Figure 3). The anterior pedal retractor muscle scar (apr in Figure 3) is small and deeply depressed and is located very close to the posterodorsal end of the anterior adductor muscle scar. The pedal protractor muscle scar is small and elongate antero-posteriorly in shape (pp in Figure 3). It is slightly depressed and located close to but distinctly separated from the center portion of the posterior termination of anterior adductor muscle scar. The shape of posterior adductor muscle scar and pallial line are ambiguous in the specimens observed due to poor preservation.
Dimensions (in mm).—UMUT MM 07001: L = 57.4, H = 22.5. SBEI 2472: L = 96.4+, H =38.5. SBEI 2473: L =99.0+, H =42.3+.
Occurrence.—The holotype (UMUT MM 07001) was collected from Ogamigo, Takayama, Gifu Prefecture, but its precise locality is unknown (Kobayashi and Suzuki, 1937). Two specimens (SBEI 2472 and 2473) were collected from the Kuwajima-kasekikabe site, Hakusan, Ishikawa Prefecture, Japan.
Remarks.—This species is characterized by an elongate shell without marked sculptural features on the shell surface, except for growth lines. The species was originally assigned to the genus Unio, but this was based solely on the external shell features and not on its internal characters. The holotype is preserved as an inner mould of a specimen with articulated valves with some outer shell impressions. The hinge teeth of the holotype are not observable because the cavity formed by the dissolution of the shell was subsequently filled with sand.
Comparison with the two newly referred specimens reveals that the holotype is a young adult of the present species. The two large specimens also reveal internal features, such as the hinge structure and muscle attachment scar. The hinge is composed of large and stout pseudocardinal teeth and less strongly developed posterior lateral teeth. The anterior adductor muscle scars are large with arborescent striae. These internal features alongside the shallow umbonal cavity and low beak are considered diagnostic characteristics of the family Margaritiferidae (e.g. Lopes-Lima et al., 2018).
The earliest fossil assigned to the Margaritiferidae was found in Upper Triassic strata in southern China (Liu, 1981; Fang et al., 2009; Van Damme et al., 2015). Araujo et al. (2017) discussed the origin and phylogeny of Margaritiferidae and stated that the abundance of margaritiferid fossils from early Mesozoic strata in China may suggest an Asian origin for Margaritiferidae. Sixteen genus names have been proposed from the Northern Hemisphere for extant and fossil Margaritiferidae, including Mesozoic species (Araujo et al., 2017). Araujo et al. (2017) considered that all these genera, including two nomina dubia, are currently attributed to the genus Margaritifera according to Fang et al. (2009), Bolotov et al. (2015) and Van Damme et al. (2015). In addition, molecular phylogenetic analysis proposed that all extant species belong to Margaritifera (e.g. Huff et al., 2004; Graf and Cummings, 2007; Takeuchi et al., 2016; Araujo et al., 2017; Kakino et al., 2023). Meanwhile, Zotin (2018) attributed the Mesozoic species to the extinct genus Palaeomargaritifera Ma, 1984 on the basis of the idea that Margaritifera larvae parasitize fish of the salmonid family, which emerged after the Eocene. However, these fossil species should be included in Margaritifera, as there are no clear differences in shell morphology or size from those of the extant Margaritifera. For these reasons, we follow Araujo et al. (2017) and argue that Unio ogamigoensis Kobayashi and Suzuki, 1937 should be recombined as Margaritifera ogamigoensis (Kobayashi and Suzuki, 1937).
Numerous Mesozoic margaritiferids have been described from China (e.g. Gu et al., 1976; Liu, 1981; Chen, 1984; Ma, 1984, 1996; Wei, 1984; Guo, 1988; Yin, 1989; Fang et al., 2009). Of those, Margaritifera pujiangensis (Ma, 1996) from the Early Cretaceous Shouchang Formation of Zhejiang Province is somewhat similar to M. ogamigoensis in having an acute posterior end of the shell. Other Chinese Mesozoic margaritiferids differ from M. ogamigoensis in having a rounded posterior shell margin or are not available for comparison due to poor preservation by literature.
Superfamily Trigonioidoidea Cox, 1952
Family Nakamuranaiadidae Guo, 1981
Genus Nakamuranaia Suzuki, 1943
Type species.—Leptesthes chingshanense Grabau, 1923; Chingshan Formation, Lower? Cretaceous, Shantung, China.
Nakamuranaia kagaensis (Koarai and Matsukawa, 2016)
Unio? ogamigoensis, Tamura, 1990, p. 31, pl. 10, figs. 6–15.
Archaeounio kagaensis Koarai and Matsukawa, 2016, p. 142, fig. 22B.
Nakamuranaia sp., Isaji, 2023, p. 15, fig. 7A–C.
Materials.—SBEI 2474–2481; CBM PS-2252, 2255, 2307, 2309.
Description.—The shell is elliptical in shape, poorly inflated and moderately thick. The anterodorsal margin is slightly concave, graduating to a rounded anterior margin. The ventral margin is broadly arcuate, forming an obtuse angle with the posterior margin at the posteroventral corner. The posterodorsal margin is long and nearly straight, forming an obtuse angle with the obliquely truncated posterior margin. The carina is distinct and extends from the umbonal region to the posterior extremity. The umbo is moderately prominent and more or less directed forwards. The shell surface is ornamented with irregularly spaced concentric growth lines. The hinge plate is narrow and bears two pseudocardinal teeth (pct in Figure 4I, K) and one posterior lateral tooth (plt in Figure 4I, K) in the right valve and one pseudocardinal tooth (pct in Figure 4J, L) and two posterior lateral teeth (plt in Figure 4J, L) in the left valve. The pseudocardinal teeth are short and originate from under the umbo. The tooth on the left valve is oblique to the anterodorsal margin (Figure 4J, L). The dorsal pseudocardinal tooth on the right valve is shorter and weaker than the ventral tooth, which is slightly arcuate (Figure 4I, K). The posterior lateral teeth on each valve are at least three times longer than the pseudocardinal teeth. They originate around the central portion of the posterodorsal margin and terminate in front of the posterior margin (Figure 4I, J). No crenulations or striations are visible in any teeth. The anterior pedal retractor scar is minute and deeply depressed and distinctly separated from the anterior adductor muscle scar (apr in Figure 4I–L). The shapes of the adductor muscle scars and the pallial line are ambiguous in the specimens observed due to poor preservation.
Dimensions (in mm).—SBEI 2474: L = 36.7, H = 23.3. SBEI 2475: L = 34.5, H = 23.6. SBEI 2476: L = 36.6, H = 27.8. SBEI 2477: L = 32.6, H = 22.6. SBEI 2478: L = 44.7, H = 30.4. SBEI 2479: L = 44.0, H = 26.7. SBEI 2480: L = 37.4+. SBEI 2481: L = 36.1+. CBM PS-2307: L = 33.2, H = 20.7. CBM PS-2309: L = 37.4, H = 24.8. CBM PS-2252: L = 38.2, H = 24.9. CBM PS-2255: L = 41.5, H = 24.5.
Occurrence.—This species commonly occurs as an autochthonous or parautochthonous element in the alternating beds of muddy fine-grained sandstone and mudstone and is usually accompanied by viviparid and hydrobiid gastropods, sphaeriid bivalves, ostracods and charophycean algae, which are considered to be shallow lake inhabitants.
Remarks.—This species commonly occurs in the lower part of the Tetori Group, such as the Kuwajima, Okurodani and Itsuki formations (e.g. Kobayashi and Suzuki, 1937; Maeda, 1961; Tamura, 1990; Gifu-ken Dinosaur Research Committee, 1993; Isaji et al., 2005; Koarai and Matsukawa, 2016; Isaji, 2023). The specimens from the Kuwajima-kasekikabe site usually occur in muddy fine-grained sandstones and show remarkable morphological variations, including elongated elliptical, trigonal and subcircular shapes, caused by the deformation of clastic rocks due to diagenetic pressure.
In previous studies without a detailed systematic description, the specimens from the Kuwajima-kasekikabe site were attributed to Unio? ogamigoensis (Tamura, 1990; Matsukawa and Ido, 1993), Unio ogamigoensis (Isaji et al., 2005) and Unio (?) ogamigoensis (Matsuura, 2001). However, the examined specimens differ from U. ogamigoensis (=M. ogamigoensis in this paper) in having a small and deeply isolated pedal retractor muscle scar, which is a diagnostic feature of the superfamily Trigonioidoidea, and in having blade-like hinge teeth, indicating that the examined specimens belong to the genus Nakamuranaia Suzuki, 1943. Similar blade-like hinge teeth are also present in Nagdongia Yang, 1975, but we refer these specimens to Nakamuranaia following Stiller and Chen (2019), who treated Nagdongia as a subgenus of Nakamuranaia (Isaji, 2023). The taxonomic position of the Trigonioidoidea is debatable, specifically whether it arose from the Trigoniida as discussed by Stiller and Chen (2019), but this is beyond the scope of this paper.
The present specimens seem to be referable to Archaeounio kagaensis Koarai and Matsukawa, 2016 described from Yanagidani, Hakusan (Figure 1). Figure 4K, L shows the internal hinge features of topotype specimens of A. kagaensis. According to Koarai and Matsukawa (2016), the genus Archaeounio has a diagnostic feature that never occurs in the subfamily Unioninae (i.e., two pseudocardinal teeth on the left valves). However, their observation of the hinge teeth of the left valve disagrees with those of the specimens from the Kuwajima-kasekikabe and Yanagidani sites, which have one pseudocardinal tooth on the left valve (Figure 4J, L). In fact, the figure of an internal mould of the left valve (Koarai and Matsukawa, 2016, p. 142, fig. 22BB) shows only one groove that appears to be a mould of the pseudocardinal tooth, suggesting that the description by Koarai and Matsukawa (2016) is questionable. Therefore, we argue that Archaeounio kagaensis Koarai and Matsukawa, 2016 should be transferred to Nakamuranaia kagaensis (Koarai and Matsukawa, 2016). This means that the genus Archaeounio Koarai and Matsukawa, 2016 becomes a junior synonym of Nakamuranaia Suzuki, 1943.
Two specimens assigned herein to Margaritifera ogamigoensis (Kobayashi and Suzuki, 1937) co-occurred with Nakamuranaia kagaensis (Koarai and Matsukawa, 2016) in the overbank deposits of the meandering river system of the Kuwajima Formation at the Kuwajima-kasekikabe site, but the lithofacies containing the two species are quite different. Specifically, N. kagaensis is common and highly abundant in the shallow lacustrine muddy fine-grained sandstone and mudstone. This appears to support the conclusion of previous studies on extant unionid bivalves that stable environments tend to support greater population densities (e.g. Allen and Vaughn, 2010; Niraula et al., 2017). In contrast, M. ogamigoensis does not occur in muddy deposits and is very rare in the medium- to coarse-grained sandstones of crevasse splay deposits. The rock containing the holotype of M. ogamigoensis is considered to represent a similar depositional environment, as it is a medium-grained sandstone with ripples. It is thus reasonable to conclude that N. kagaensis inhabited shallow lakes on floodplains, similar to other freshwater gastropods and bivalves such as Campeloma onogoense and Sphaerium coreanicum, whereas M. ogamigoensis was a stream dweller that was occasionally transported to the floodplains when a stream’s natural levees broke. As summarized in Figure 5, a variety of aquatic and semiaquatic environments formed by the meandering river system appear to provide habitats suitable for both molluscan assemblages, which includes stream dwellers, lake dwellers, and more terrestrially-oriented species such as stylommatophoran land snails (Isaji, 2010).
Most extant margaritiferids prefer flowing streams, especially in the adult stage of their life cycle (e.g. Vannote and Minshall, 1982; Skinner et al., 2003; Bolotov et al., 2015; Araujo et al., 2017). On the other hand, many researchers have noted that juvenile and adult margaritiferids have different habitat preferences from each other (Okada and Ishikawa, 1959; Buddensiek et al., 1993; Hastie and Young, 2003). Kobayashi and Kondo (2008) also reported that the habitat preferences of Margaritifera laevis (Haas, 1910) change during its lifetime, with juveniles and young adults (shell length < 30 mm) inhabiting muddy bottoms in small tributaries and adults preferring sand and gravel bottoms in larger tributaries to anchor their bodies.
In the case of M. ogamigoensis, no juveniles or young adults have yet been found in muddy deposits. Notably, distinguishing juveniles or young adults of M. ogamigoensis from N. kagaensis can be difficult. In addition, even adult shells have not been found in very coarse-grained sandstones deposited at the bottoms of large streams, where adult individuals seemingly preferred to burrow (Figure 5). The preference of M. ogamigoensis for rapidly flowing streams explains its low fossilization potential, resulting in its extremely rare occurrences in the Kuwajima Formation and other equivalent formations of the Tetori Group. It can therefore be argued that there is insufficient information available to discuss the paleoecology of M. ogamigoensis throughout its life history.
I would like to express my sincere thanks to Kento Otsuka (Shiramine Institute of Paleontology) and the Kuwajima Fossil Survey Team for collecting fossils. I am grateful to the Culture Division, Hakusan City for financial assistance and for their interest in this research. I also wish to thank Takenori Sasaki (The University Museum, The University of Tokyo) for permission to observe a holotype specimen (UMUT MM 07001). Thanks are also due to Simon Schneider (CASP, The University of Cambridge) and an anonymous reviewer for their critically reading the manuscript and providing valuable comments.
