RESEARCH ARTICLE
Heterocarinella, a new ammonoid genus: Turrilitidae survived into the Santonian (Late Cretaceous) in the northwestern Pacific realm
2026 年 30 巻 p. 164-173
詳細
2026 年 30 巻 p. 164-173
A new genus and species of the family Turrilitidae, Heterocarinella japonica gen. et sp. nov., is described from the Santonian (Upper Cretaceous) of the Haboro and Kotanbetsu areas in Hokkaido, northern Japan. This taxon is characterized by a persistently turriconic shell, a siphuncle positioned near the upper edge of the flank, and a vertically asymmetrical keel at or below the middle flank of the whorls. Notably, the keel differentiates Heterocarinella japonica gen. et sp. nov. from all other genera belonging to the Turrilitidae. The family Turrilitidae has previously been considered to range from the middle Albian to the late Cenomanian. However, the occurrence of Heterocarinella japonica gen. et sp. nov. demonstrates that this family survived until at least the Santonian. In addition, Tridenticeras from the Turonian to Coniacian of Europe, which is often assigned to the Nostoceratidae, is here re-evaluated as more appropriately belonging to the Turrilitidae. The previous misclassification likely resulted from limited knowledge of the position of the siphuncle, an internal structure, and from the tendency to prioritise stratigraphic consistency over morphological characters in taxonomic assignments. The discovery of Heterocarinella japonica gen. et sp. nov. suggests that the Turrilitidae survived the Cenomanian–Turonian Oceanic Anoxic Event 2 (OAE-2) while progressively narrowing their palaeogeographic range, ultimately contracting only in the northwestern Pacific realm.
ZooBank registration: urn:lsid:zoobank.org:pub:7980761E-A428-4B6D-842B-82CD1F83BABB
The family Turrilitidae Gill, 1871 comprises ammonoids characterized by high-spired, helicospirally coiled shells (= turricone) with whorls in contact with each other, and is known to have occurred worldwide from the middle Albian to the late Cenomanian (Wright et al., 1996). In the northwestern Pacific realm, Turrilitidae ranging from the Albian to the Cenomanian have been recorded from the Yezo Group in Hokkaido, Japan (e.g. Yabe, 1904; Matsumoto et al., 1999; Matsumoto and Takahashi, 2000, 2001a, b). Above the Cenomanian, there are no records of Turrilitidae except for Tridenticeras Wiedmann, 1962 from the Turonian to Coniacian, whose family-level assignment remains controversial. Subsequently, the Nostoceratidae Hyatt, 1894, considered to have been derived from Turrilitoides Spath, 1923 of Turrilitidae, diversified during the Turonian–Maastrichtian and has a rich fossil record (Matsumoto, 1967, 1977; Wright et al., 1996; Schaffert and Larson, 2021). Because the shell morphologies of Turrilitidae and Nostoceratidae are very similar and often difficult to distinguish, stratigraphic occurrence has sometimes been used as the primary criterion for taxonomic assignment rather than morphology (e.g. Kennedy, 1984).
The classification of the superfamily Turrilitoidea Gill, 1871 is based on differences in whorl coiling and shell ornamentation (Wright et al., 1996). The position of the siphuncle is also considered a useful taxonomic character (e.g. Wiedmann, 1962; Matsumoto, 1977; Klinger and Kennedy, 1978), although it has not been sufficiently documented in all taxa. Shell ornamentation in Turrilitidae includes ribs, multiple rows of tubercles, and spiral grooves. In contrast, a keel formed by a linear ridge on the whorl is known in several groups of planispirally coiled ammonoids within the order Ammonitida Haeckel, 1866, whereas it is rarely developed in members of the suborder Ancyloceratina Wiedmann, 1966. Here, I describe a new genus and species of turrilitid heteromorph ammonoid from the Santonian of the Haboro and Kotanbetsu areas, northwestern Hokkaido, which is characterized by a siphuncle positioned near the upper edge of the flank and the presence of a keel on the whorl.
The Cretaceous forearc basin deposits of the Yezo Group are distributed in a north–south-trending belt extending for more than 1,000 km from Sakhalin in the Russian Far East to Hokkaido, northern Japan (e.g. Matsumoto, 1954; Vereshchagin, 1977; Takashima et al., 2004; Shigeta and Maeda, 2005). The Yezo Group is well exposed in the Haboro and Kotanbetsu areas of northwestern Hokkaido, and its stratigraphy there has been studied in detail (e.g. Igi et al., 1958; Tsushima et al., 1958; Yamaguchi and Matsuno, 1963; Hashimoto et al., 1965; Toshimitsu, 1985, 1988; Wani and Hirano, 2000; Okamoto et al., 2003; Kawabe and Okamoto, 2012).
The Yezo Group in the Haboro area is subdivided into the Tenkaritoge, Haborodake, Shirochi, Lower Haborogawa, Upper Haborogawa, and Nagareya formations, in ascending stratigraphic order (Okamoto et al., 2003). The Lower Haborogawa Formation is approximately 950 m thick and mainly consists of mudstone with subordinate sandy siltstone and sandstone. The Upper Haborogawa Formation is approximately 700 m thick and mainly consists of sandy siltstone and silty sandstone. The Lower Haborogawa and Upper Haborogawa formations correspond to lithostratigraphic units Ua–Uf and Ug–Uh of Igi et al. (1958) and Tsushima et al. (1958), respectively.
The occurrence of macrofossils varies among the units (Okamoto et al., 2003). Unit Ua yields abundant fossils both from the host rock and from concretions. In unit Ub-c, concretions are scarce, and bivalve fossils are preserved directly in the host rock. The lower part of unit Ud-e yields well-preserved fossils from concretions, whereas the upper part is characterized by reddish-brown, flattened sideritic concretions that rarely contain fragments of inoceramid shells, but fossils are otherwise scarce. Units Uf–Uh yield abundant fossils from calcareous concretions. Based on inoceramid bivalve and ammonoid biostratigraphy (Toshimitsu et al., 2007), units Ua–Uh are correlated as follows: Ua, upper Turonian–Coniacian; Ub-c–Ud-e, Coniacian; and Uf–Uh, Santonian (Okamoto et al., 2003).
Although the stratigraphic scheme in the Kotanbetsu area differs slightly from that of the Haboro area proposed by Okamoto et al. (2003), unit Uh, from which a specimen described herein was most likely collected, is correlated with the Santonian, generally consistent with the stratigraphic interpretation of the Haboro area (Tsujino et al., 2003; Aiba, 2019, 2025).
Three specimens from Haboro and Kotanbetsu were examined in this study (Figure 1A, B). Specimen MCM-A1322 was collected as a float concretion from a tributary of Machiyoizawa Creek in the Haboro area (Figure 1C). At the sampling site, unit Ud-e is exposed, and unit Uf crops out approximately 100 m upstream (Okamoto et al., 2003). The matrix of the concretion containing MCM-A1322 is sandy siltstone and includes multiple individuals of Baculites cf. bailyi Woods, 1906 and Polyptychoceras sp. The ovate whorl cross-section characteristic of B. cf. bailyi is a feature recognised in Baculites species appearing from the Santonian onwards (Matsumoto and Obata, 1963; Tsujino, pers. comm., 2026). Based on the stratigraphic distribution upstream, the lithology of the matrix, and co-occurring fossils, the concretion containing MCM-A1322 is most likely derived from unit Uf. Specimen MCM-A2638 was recovered from a float concretion of sandy mudstone collected at Kaminosawa Creek in the Kotanbetsu area (Figure 1D). Unit Uh is widely distributed near and upstream from the sampling site. It is composed of sandy mudstone and fine- to coarse-grained sandstone and is correlated with the upper Santonian. Accordingly, MCM-A2638 is inferred to have been derived from unit Uh. Specimen MCM-A1590 is from the Santonian of the Haboro area (Hayakawa, 1998), although it lacks precise locality data. For comparison, the holotype of the nostoceratid Horotateceras tatsuyai Hayakawa, 1998 (MCM-A0612) was examined. This specimen was collected from the Turonian of the Horotatesawa Creek in the Kotanbetsu area.
Specimens were examined and measured from photographs of specimens whitened with ammonium chloride. Internal structures were examined using tomographic images obtained by X-ray computed tomography and three-dimensional reconstructions. Tomographic data were acquired using a Nikon XT H225 ST system at the Tomakomai Techno Centre, Hokkaido, Japan. The resulting tomographic images were analysed and 3D models were reconstructed using Fiji (ImageJ) and Molcer Plus (White Rabbit Co., Ltd.).
The systematic description follows the taxonomic frameworks of Wright et al. (1996) and Hoffmann et al. (2022). The measurement points follow Aiba (2025, fig. 2A).
Institutional abbreviations.—MCM: Mikasa City Museum, Hokkaido, Japan.
Order Ammonitida Haeckel, 1866
Suborder Ancyloceratina Wiedmann, 1966
Superfamily Turrilitoidea Gill, 1871
Family Turrilitidae Gill, 1871
Discussion.—Within the Late Cretaceous Turrilitoidea, the turriconic whorl is a feature shared by the families Turrilitidae Gill, 1871 and Nostoceratidae Hyatt, 1894 (Wright et al., 1996). As a general trend, Turrilitidae exhibit minimal ontogenetic change in coiling. Although some species show slight uncoiling in the later stages of growth, the shell essentially retains a turriconic form with whorls in contact with one another throughout ontogeny (Klinger and Kennedy, 1978). In contrast, some Nostoceratidae develop a hook in the later growth stages or form a three-dimensionally coiled vermicone during the main growth phase. However, some species that do not deviate from a turricone are also known (e.g. Eubostrychoceras muramotoi (Matsumoto, 1967), Hyphantoceras orientale (Yabe, 1904)).
With the exception of the basal Albian genera Proturrilitoides Breistroffer, 1940 and Pseudhelicoceras Spath, 1922, the siphuncle is positioned near the upper edge of the flank in all turrilitid genera for which its position has been documented (Klinger and Kennedy, 1978; Wright et al., 1996). In Nostoceratidae, the siphuncle is located near the middle flank of the whorls in many species, whereas it lies in the lower part of the whorls in Yezoceras Matsumoto, 1977 (Matsumoto, 1977; Aiba et al., 2021), and in some species of Hyphantoceras and related taxa it is situated slightly above the middle flank of the whorls (Aiba, 2025, figs. 2B, 4N; Aiba and Karasawa, 2025, fig. 4C). By contrast, there are no known nostoceratids in which the siphuncle is situated near the upper edge of the flank. Based on these observations, Klinger and Kennedy (1978) regarded the position of the siphuncle as one of the diagnostic characters distinguishing Turrilitidae from Nostoceratidae.
The genus Tridenticeras Wiedmann, 1962 from the Coniacian and Turonian of Europe has been variously assigned either to the Turrilitidae (Wiedmann, 1962; Klinger and Kennedy, 1978) or to the Nostoceratidae (Kennedy, 1984; Santamaría-Zabala, 1992; Vašíček, 1992; Kaplan and Kennedy, 1994; Kennedy et al., 1995; Summesberger and Kennedy, 1996; Wright et al., 1996; Schaffert and Larson, 2021). The type species of Tridenticeras was originally described as Turrilites by Schlüter (1876). Based purely on its turriconic whorls and some other features, Wiedmann (1962) established the genus Tridenticeras within the Turrilitidae. Subsequently, Kennedy (1984) reassigned the genus to the Nostoceratidae based on the stratigraphic gap between Tridenticeras and other turrilitid genera. However, in several species of Tridenticeras, the siphuncle is positioned near the upper edge of the flank, a feature characteristic of the Turrilitidae (e.g. T. undosum (Schlüter, 1876); T. soukoupi Vašíček, 1992; T. binodosum Summesberger and Kennedy, 1996). Nevertheless, its continued assignment to the Nostoceratidae suggests that the significance of this character has hitherto been underestimated. Given that many of the criteria proposed by Klinger and Kennedy (1978) for distinguishing the Turrilitidae from the Nostoceratidae have not been explicitly refuted, it is reasonable to regard the presence of a siphuncle near the upper edge of the flank as a fundamental characteristic of the Turrilitidae (Table 1). Accordingly, the taxonomic assignment of Tridenticeras to the Turrilitidae by Wiedmann (1962) is here considered valid.
| characters | Turrilitidae | Nostoceratidae |
|---|---|---|
| Presence or absence of constrictions | Constrictions absent except at the aperture | Constrictions may be present on the whorls |
| Position of the siphuncle | Siphuncle near the upper edge of the flank in most forms, but the middle flank in early forms such as Proturrilitoides and Pseudhelicoceras | Siphuncle near the middle flank in most forms, but near the lower flank in Yezoceras |
| Rib characteristics | Ribs generally single, except in some species of Mariella, Pseudhelicoceras and Hypoturrilites | Rib bifurcation, looping, and intercalation common |
| Coiling directions | Early forms coiled in both directions, later forms exclusively sinistral | Both dextral and sinistral coiling present throughout |
Genus Heterocarinella gen. nov.
ZooBank lsid: urn:lsid:zoobank.org:act:95C44F23-7812-473E-8928-793382C087EE
Type species.—Heterocarinella japonica Aiba gen. et sp. nov.
Diagnosis.—This genus is characterised by turriconic whorls in contact with each other, each bearing a keel at or just below the middle flank, and a siphuncle located near the upper edge of the flank. The whorl cross-section is vertically asymmetrical with respect to the keel, with a gently curved upper part above the keel apex and a somewhat steeper, sharper lower part. The keel is continuous and is not interrupted by ribs, except for a double collar rib associated with a constriction that appears in the later growth stage.
Discussion.—Heterocarinella gen. nov. exhibits several features consistent with those of other turrilitid genera, including the presence of a siphuncle near the upper edge of the flank, turriconic whorls throughout ontogeny, and a double collar rib associated with a constriction appearing in the later growth stage.
Comparison.—Heterocarinella gen. nov. can be distinguished from all other genera of the Turrilitidae by the presence of a vertically asymmetrical keel on each whorl. The lower-edge ridge of the whorl present in Carthaginites Pervinquière, 1907 and Neostlingoceras Klinger and Kennedy, 1978 (e.g. Matsumoto and Takahashi, 2000, fig. 3; Kennedy, 2020, text-fig. 25), and the spiral groove present in Ostlingoceras Hyatt, 1900, Mesoturrilites Breistroffer, 1953, and Hypostlingoceras Matsumoto and Takahashi, 2000 (e.g. Matsumoto and Takahashi, 2000, figs. 7, 8; Kennedy, 2020, text-fig. 27), are somewhat similar in form to the keel of the new genus. However, the ridge on the lower edge of the whorl is not as sharp as the keel, and the spiral groove differs in having two ridges along its sides. Additionally, the presence of rows of tubercles in these genera clearly distinguishes them from the new genus.
Remarks.—Horotateceras Hayakawa, 1998, belonging to the family Nostoceratidae, shares the presence of a keel with Heterocarinella gen. nov., and such unusual similarity within the superfamily Turrilitoidea may appear to suggest that they are closely related. However, although a keel may sometimes be retained within a lineage, it has also arisen independently in multiple different lineages (e.g. Oppeliidae, Desmoceratidae, Collignoniceratidae: see descriptions in Wright et al., 1996); therefore, its presence or absence does not necessarily indicate phylogenetic affinity. Moreover, the characteristics of the keel differ between Heterocarinella gen. nov. and Horotateceras. The keel of Horotateceras is bilaterally symmetrical and resembles the obtuse keel of Damesites damesi Jimbo, 1894 (Desmoceratoidea). In addition, as in Prionocyclus Meek, 1871 (Acanthoceratoidea), the keel intersects with the ribs, forming a serrated structure (Hayakawa, 1998). Hayakawa (1998) illustrated one specimen, assigned to Heterocarinella gen. nov. herein (MCM-A1590) as uncertain species and noted that its keel differs from that of Horotateceras. Furthermore, CT scan analysis of the single known specimen of Horotateceras (the holotype of H. tatsuyai Hayakawa, 1998: MCM-A0612) has confirmed that the siphuncle is located near the middle flank of the whorl, as in other genera of the Nostoceratidae (Figure 2). Based on these differences, Heterocarinella gen. nov. and Horotateceras can be clearly distinguished.
Occurrence.—Santonian of Hokkaido, Japan.
Heterocarinella japonica sp. nov.
ZooBank lsid: urn:lsid:zoobank.org:act:82BE785B-7792-4CD8-A4EC-4ABC8CBC5385
“Keeled ammonite (gen. et sp. indet.).” Hayakawa, 1998, p. 43, fig. 3.
Turrilitidae gen. et sp. indet. b-1. Fukuoka, 2000, p. 209, lower row, the second from right.
Turrilitidae gen. et sp. indet. b-2. Fukuoka, 2000, p. 209, lower row, the rightmost one.
Type specimens.—MCM-A1322 (Figure 3A–J) is herein designated as the holotype, and MCM-A1590 (Figure 3K–P) and MCM-A2638 (Figure 3Q–V) are herein designated as paratypes. MCM-A1590 is the specimen illustrated by Hayakawa (1998, fig. 3) as “keeled ammonite (gen. et sp. indet.)”. Information on each specimen is summarised in Table 2.
| specimen number | type | locality | collector | number of whorls | a (mm) | b (mm) | c (mm) | d (°) |
|---|---|---|---|---|---|---|---|---|
| MCM-A1322 | holotype | Machiyoi-zawa C., Haboro area (f) | T. Kitayama | 7 ½ | 42.1 | 1.1 | 10.7 | 52 |
| MCM-A1590 | paratype | Haboro area | K. Muramoto | 2 | 15.9 | 3.3 | 8.9 | 85 |
| MCM-A2638 | paratype | Kamino-sawa C., Kotanbetsu area (f) | K. Arai | 3 | 16.8 | 2 | 6.6 | 42 |
Type locality.—Float in a tributary of Machiyoizawa Creek in the Haboro area of northwestern Hokkaido, northern Japan.
Diagnosis.—As for the genus.
Description.—All three specimens described herein are sinistral turricones. The holotype MCM-A1322 comprises approximately seven and a half whorls, with a relatively narrow umbilicus, and is moderately tightly coiled with an apical angle of 52° (Figure 3A–G; Table 2). The final preserved whorl shows a slight increase in coiling tightness compared with the preceding whorls. The whorl section has a slightly compressed sub-elliptical shape, with a concave indentation on the upper part where it overlaps the preceding whorl (Figure 3H). The siphuncle is situated near the upper edge of the flank, and an asymmetrical keel is present just below the middle of the whorl flank (Figure 3H). The shell surface is ornamented with closely spaced, prorsiradiate normal ribs. In the later growth stage, there is a well-developed double collar rib associated with a constriction (Figure 3J). Normal ribs do not interrupt the keel, but a double collar rib associated with a constriction does.
The paratype MCM-A1590 comprises two whorls and has apical angle of 85°, which is larger than that of the holotype (Figure 3K–P; Table 2). However, the entire whorl may have undergone vertical compaction, resulting in secondary deformation. The paratype MCM-A2638 comprises three whorls and has an apical angle of 42°, which is smaller than that of the holotype (Figure 3Q–V; Table 2). Thus, although the whorls are in contact during coiling in all specimens, a certain degree of variation in apical angle is apparent (Table 2). The position of the siphuncle near the upper edge of the flank is confirmed in all specimens.
The suture line is visible near the end of the fifth whorl of the holotype MCM-A1322 (Figure 3I). It consists of a relatively shallowly incised external lobe (“E” in Figure 3I), a deeply incised bifid lateral lobe (“L” in Figure 3I), and a moderately shallow, bifid saddle between them.
Measurements.—Measurements of the specimens are summarised in Table 2.
Remarks.—Two specimens illustrated by Fukuoka (2000) as “Turrilitidae gen. et sp. indet. b-1” and “Turrilitidae gen. et sp. indet. b-2” both possess a turriconic shell in which the whorls are in contact, and a keel-like elevation is present just below the middle flank of the whorl. These features indicate that these specimens are assigned to Heterocarinella japonica gen. et sp. nov. The double collar rib associated with a constriction present in the later growth stage, together with a slight tightening of coiling (= a slight narrowing of the umbilicus), resembles the mature modification seen at the final growth stage of Turrilitoides (e.g. Klinger and Kennedy, 1978, pl. 9M; Jattiot et al., 2023, fig. 3B), and may likewise indicate maturity. The variation in apical angle may reflect dimorphism rather than secondary deformation. Indeed, dimorphism has been documented in some species of Turrilitidae (e.g. Jattiot et al., 2023). However, this possibility cannot be adequately evaluated in the present study because of the limited number of available specimens.
Occurrence.—Of the specimens described herein, two were collected as float, and the other lacks precise locality data. The two float specimens most likely originated from the Santonian, corresponding to Inoceramus amakusensis zone (Toshimitsu et al., 1995, 2007), although their respective sources are inferred to be the lower and upper parts of the Santonian. Accordingly, this species may have occurred throughout the Santonian; however, additional in situ material is required to further constrain its stratigraphic range.
Examination of the siphuncle, an internal shell structure, is highly dependent on the state of preservation of a fossil and, moreover, generally requires either physical sectioning, peeling of the shell or non-destructive techniques such as CT scanning. Consequently, this character is not known from all specimens (e.g. Klinger and Kennedy, 1978; Wright et al., 1996). However, the series of discussions concerning Tridenticeras and the results of the present study indicate that the position of the siphuncle should not be disregarded in the classification of Turrilitoidea. Given the incompleteness of the fossil record, taxonomic judgements based on stratigraphic range are inherently limited. Therefore, as noted by Wiedmann (1962), classification should be based on morphological characters rather than stratigraphic consistency.
The record of Heterocarinella japonica gen. et sp. nov. extends the stratigraphic range of Turrilitidae, which was previously regarded as being restricted to the Albian–Cenomanian (e.g. Kennedy, 1984; Wright et al., 1996), into the Santonian. The occurrence of not only Heterocarinella japonica gen. et sp. nov. but also Tridenticeras suggests that this range extension does not merely represent an exceptional occurrence. Rather, it implies that previous interpretations of the stratigraphic distribution of Turrilitidae may have been constrained by limited knowledge of character states and biases in identification criteria and thus require re-evaluation.
The evolutionary history and diversification of Turrilitidae may therefore have differed from previously inferred. The new record and the revised stratigraphic range suggest the following palaeobiogeographic and evolutionary history. Turrilitid ammonoids were distributed globally until the Cenomanian but suffered a severe decline during the Cenomanian–Turonian Oceanic Anoxic Event 2 (OAE-2), after which their distribution became drastically restricted in the Turonian–Coniacian. Subsequently, they may have undergone a further contraction of their range and ultimately survived only in the northwestern Pacific realm during the Santonian. Such a pattern, in which formerly widespread taxa subsequently underwent range contraction and ultimately persisted and/or became regionally specialised in the northwestern Pacific, is well documented in several Late Cretaceous ammonoid groups (e.g. Tetragonitinae: Shigeta, 1989, 1992, 2024; Gabbioceratinae: Shigeta et al., 2012, 2024; Desmoceratinae: Nishimura and Maeda, 2025; Kossmaticeratidae: Maeda, 1993; Nostoceratidae: Shigeta and Izukura, 2022). Thus, the northwestern Pacific may have functioned as a refugium for ammonoids.
I would like to thank the collectors of the specimens described herein, Toshio Kitayama, Kikuwo Muramoto and Kazato Arai. I am also grateful to Tomoki Karasawa (Mikasa City Museum) and Ryoji Wani (Yokohama National University) for kindly facilitating the examination of these specimens; to Yasuyuki Tsujino (Tokushima Prefectural Museum) for his assistance in identifying the associated fossils; and to Christian Klug (Universität Zürich) and David Peterman (Miami University) for their valuable comments, which improved the manuscript.
