RESEARCH ARTICLE
A tarsometatarsus of a plotopterid bird from the lower Oligocene Yamaga Formation, Ainoshima, Japan
2025 年 29 巻 p. 292-299
詳細
2025 年 29 巻 p. 292-299
A tarsometatarsus belonging to the family Plotopteridae (Aves, Suliformes) is described from the lower Oligocene Yamaga Formation of the Ashiya Group on the island Ainoshima in Kitakyushu, Fukuoka, western Japan. It resembles Phocavis in gracility and size but shares with Copepteryx and Hokkaidornis features such as an elongated trochlea metatarsi IV, splayed trochlea metatarsi II, and an intertrochlear fossa between the trochleae metatarsorum III and IV. Comparison with modern waterbirds indicates that the intertrochlear fossa represents an ancestral character. Intertrochlear structures exhibit variability among plotopterids, suggesting that these structures underwent distinct evolutionary modifications within each lineage. The specimen does not appear to be a referable to Copepteryx or Hokkaidornis, implying a greater morphological diversity of the family in Japan.
The family Plotopteridae Howard, 1969 were wing-propelled diving seabirds that flourished from the late Eocene to the middle Miocene along the northern Pacific Rim (Howard, 1969; Hasegawa et al., 1979; Olson, 1980; Olson and Hasegawa, 1985, 1996; Goedert, 1988; Sakurai et al., 2008; Dyke et al., 2011; Kaiser et al., 2015; Mayr and Goedert, 2016; Ohashi and Hasegawa, 2020). While similarities in forelimb (Mayr, 2004) and brain (Kawabe et al., 2014) morphologies suggest a close relationship between the family Plotopteridae and penguins (order Sphenisciformes), Plotopteridae is now generally considered to be more closely related to the order Suliformes (Howard, 1969; Olson and Hasegawa, 1979, 1996; Sakurai et al., 2008; Smith, 2010; Ando and Fordyce, 2013; Ando and Fukata, 2018; Mayr et al., 2020; Mayr et al., 2025).
To date, 13 species from 10 genera of Plotopteridae have been described. A variety of plotopterid specimens encompassing six genera and eight species have been discovered in the uppermost Eocene to Oligocene series along the Pacific coast of North America (Phocavis maritimus Howard, 1969; Tonsala hildegarde, Olson, 1980; Phocavis maritimus Goedert, 1988; Klallamornis buchanani Dyke et al., 2011; Stemec suntokum Kaiser et al., 2015; Olympidytes thieli Mayr and Goedert, 2016; Klallamornis abyssa Mayr and Goedert, 2016; and Klallamornis? clarki Mayr and Goedert, 2016). Mayr and Goedert (2018) considered Plotopterum, Phocavis, and Stemec to be the most basal taxa within the family. Since these genera are all found in North America, this led to the hypothesis that plotopterids originated off the western coast of North America during the late Eocene (Mayr and Goedert, 2022).
The Oligocene strata of northern Kyushu, Japan also preserve a rich record of plotopterids, with four genera and five species recognized to date (Copepteryx hexeris Olson and Hasegawa, 1996; C. titan Olson and Hasegawa, 1996; Sternornis kanmonensis Ohashi and Hasegawa, 2020; Empeirodytes okazakii, Ohashi and Hasegawa, 2020). With the exception of C. hexeris, these taxa were established based on isolated material. Additionally, other indeterminate plotopterid materials from the region are often fragmentary or isolated (e.g. Ando and Fukata, 2018; Mori and Miyata, 2021). Consequently, the morphological details and diversity of the family in the Oligocene of northern Kyushu remain poorly understood.
In the study of plotopterids, the tarsometatarsus is a particularly well-documented bone element. Among the 13 species from 10 genera that have been identified, tarsometatarsi have been documented in six species from six genera (Copepteryx hexeris, Hokkaidornis abashiriensis, Klallamornis? clarki, Olympidytes thieli, Phocavis maritimus, and Tonsala hildegardae). Phocavis and K.? clarki were described solely based on isolated tarsometatarsi (Goedert, 1988; Olson and Hasegawa, 1996; Sakurai et al., 2008; Mayr and Goedert, 2016; Mayr and Goedert, 2018; Mayr and Goedert, 2022). Consequently, the morphology of tarsometatarsus is frequently referenced in discussions of plotopterid phylogeny (Mayr, 2004; Mayr and Goedert, 2016, 2018), and any new plotopterid tarsometatarsus could provide valuable insights into their evolution and diversity.
Here, I describe a new plotopterid tarsometatarsus specimen from the Oligocene of northern Kyushu, Japan and compare its intertrochlear structure with those of contemporary avian species.
Institution abbreviations.—AMP, Ashoro Museum of Paleontology, Ashoro, Hokkaido, Japan; KMNH, Kitakyushu Museum of Natural History and Human History, Kitakyushu, Fukuoka, Japan; LACM, Natural History Museum of Los Angeles County, Los Angeles, California, U.S.A.; OMNH, Osaka Museum of Natural History, Osaka, Osaka, Japan; SINH, Shikoku Institute of Natural History, Susaki, Kochi, Japan; SMF, Senckenberg Research Institute Frankfurt, Frankfurt, Germany; UWBM, Burke Museum of Natural History and Culture, University of Washington, Seattle, Washington, U.S.A.; YIO, Yamashina Institute for Ornithology, Abiko, Chiba, Japan.
The new plotopterid tarsometatarsus specimen KMNH VP 60017 was collected from the island Ainoshima in Kitakyushu, Fukuoka, northern Kyushu, western Japan by Kiichiro Hachiya in May 1992. Although the exact location of the specimen remains unknown, it is reasonably certain that it originated from the Yamaga Formation, which is exposed across the entire island (Nakae et al., 1998). This formation comprises siltstones to coarse-grained sandstones deposited in shelf to near-shore environments (Tomita and Oji, 2010). Several studies have investigated the age of the Yamaga Formation. Fission track ages of 30.3 ± 1.2 Ma and 31.7 ± 2.3 Ma have been determined from tuff horizons in the upper and lowest parts of the formation, respectively (Murakami et al., 1989; Ozaki and Hamasaki, 1991). The calcareous nannoflora from the upper member of the formation has been assigned to subzone CP19a (Okada, 1992), which corresponds to the “mid” Oligocene (29.62–26.84 Ma; Vandenberghe et al., 2012). Given the approximate concordance of these ages, the Yamaga Formation is inferred to date to the late Rupelian.
I compared the new specimen with published figures of other plotopterid tarsometatarsus specimens (Goedert, 1988; Olson and Hasegawa, 1996; Sakurai et al., 2008; Dyke et al., 2011; Mayr and Goedert, 2016, 2018; see taxa and specimens listed in the Table 1) and directly with tarsometatarsi of extant Aequornithes (the waterbird clade that includes cormorants, gulls, penguins, and others: see Burleigh et al., 2015). In this study, I directly observed tarsometatarsus specimens from 56 individuals representing 21 species across 15 genera of waterbirds, encompassing 11 families, including Phalacrocoracidae (cormorants) and Spheniscidae (penguins) (Table 2).
| Taxon | Tarsometatarsus length (A: mm) | Tarsometatarsus shaft width (B: mm) | Length/shaft width (B/A) | Tarsometatarsus shaft depth (mm) | Length of trochlea metatarsi IV (C: mm) | Lateral angle (D: °) | Tarsometatarsus proximal width (mm) | Tarsometatarsus distal width (mm) | Relative length of the tarsometarsi IV (C/A) | trochlear metatarsi II splayed |
|---|---|---|---|---|---|---|---|---|---|---|
| Copepteryx hexeris (KMNH-VP 200,001) | 104e | 26.5e | 3.7 | 14e | 17.4g* | 134°g* | 52e | 60.3e | 0.17 | Yes |
| Hokkaidornis abashiriensis (AMP 44)d | 96.7 | 27.7 | 3.4 | 15.2 | 18.7** | 134°** | 51.3 | 59.4 | 0.19 | Yes |
| ?Kllalamornis clarki (LACM 129405) | 85.2f | 24.0c** | 3.6 | – | 10.5f | 135°c** | ~43c | ~53c | 0.12 | Yes |
| KMNH VP 600017 | ~60 | 12.3 | ~4.6 | 6.7 | 10.8 | 144°* | – | ~26 | ~0.18 | Yes |
| Olympidytes theili (SMF Av 609)c | ~53 | 15.2** | >3.5 | – | 6.4** | – | 27.4 | – | <0.12 | – |
| Phocavis maritimus (LACM 123897) | 59.6b | 13.9b | 4.4 | – | 6.9a** | 154°a** | 23.6b | 25.7b | 0.12 | No |
| Tonsala hildegardae (SMF Av 645)a | 49.3 | 13.6** | 3.6 | – | 8.9** | 135°** | – | 26.3 | 0.18 | No |
| Family | Binomial name | Common name | number of specimens observed |
|---|---|---|---|
| Ardeidae | Ardea cinerea | Grey heron | 6 |
| Ardea intermedia | Medium egret | 1 | |
| Egretta garzetta | Little egret | 5 | |
| Nycticorax nycticorax | Black-crowned night heron | 7 | |
| Charadriidae | Pluvialis fulva | Pacific golden plover | 1 |
| Ciconiidae | Ciconia boyciana | Oriental stork | 2 |
| Diomedeidae | Diomedea exulans | Snowy albatross | 1 |
| Phoebastria albatrus | Shrot-tailed albatross | 2 | |
| Phoebastria immutabilis | Laysan albatross | 3 | |
| Fregatidae | Fregata ariel | Lesser frigatebird | 1 |
| Gaviidae | Gavia arctica | Black-throated loon | 2 |
| Gavia pacifica | Pacific loon | 2 | |
| Gavia stellate | Red-throated loon | 2 | |
| Hydrobatidae | Oceanodroma monorhis | Swinhoe’s storm petrel | 2 |
| Pelecanidae | Pelecanus onocrotalus | Great white pelican | 2 |
| Phalacrocoracidae | Phalacrocorax capillatus | Japanese cormorant | 1 |
| Phalacrocorax carbo | Great cormorant | 8 | |
| Phalacrocorax pelagicus | Pelagic cormorant | 1 | |
| Procellariidae | Ardenna grisea | Sooty shearwater | 2 |
| Calonectris leucomelas | Streaked shearwater | 4 | |
| Spheniscidae | Spheniscus humboldti | Humboldt penguin | 1 |
I followed the descriptive terminology of Baumel and Witmer (1993), except for the intertrochlear fossa/canal. The angles between the medial edges of the trochlea metatarsi II and the shaft (Figure 1D) were measured on published or provided images using the line-drawing tool of the note-taking software Curio 30 (Zengobi, North Carolina, United States). Mayr and Goedert (2022) regarded the relative length of the trochlea metatarsi IV as an informative character that distinguishes Copepteryx and Hokkaidornis from North American plotopterids, including Klallamornis? clarki. However, this character has not been clearly defined. In this study, this length was defined as the trochlea metatarsi IV length divided by the tarsometatarsus length. In addition, tarsometatarsal gracility was evaluated by determining the proportion between shaft width and length (see length/shaft width (B/A) in Table 1).
Linear measurements of the new specimen were taken with a generic caliper to the nearest 0.1 mm. For other compared specimens, previously reported lengths were used, or linear measurements were obtained directly or indirectly from published or provided figures (see Table 1) using Image J (Schneider et al., 2012).
Class Aves Linnaeus, 1758
Order Suliformes Sharpe, 1891
Family Plotopteridae Howard, 1969
Plotopteridae gen. et sp. indet.
Referred specimen.—KMNH VP 60017, an incomplete left tarsometatarsus.
Locality.—Ainoshima, Kitakyushu, Fukuoka, western Japan.
Horizon and age.—Yamaga Formation, Ashiya Group; late Rupelian (early Oligocene).
Measurements.—Length, 59.7 mm (as preserved); shaft width, 12.3 mm; shaft depth, 6.7 mm (shaft width and depth were measured at the narrowest point); distal width, 26.2 mm (as preserved) (Table 1).
Description and comparison.—The tarsometatarsus lacks its proximomedial part, but a small portion of the cotyla lateralis (which enables a length measurement), the lateral surface, and the trochleae metatarsorum II to IV are mostly preserved. KMNH VP 60017 is from a medium-sized individual, comparable to Phocavis in size, and distinctly smaller than Copepteryx and Hokkaidornis (Table 1). It is also gracile, even when compared with other specimens of similar or smaller sizes, such as Tonsala (fig. 2A in Mayr and Goedert, 2018) and Olympidytes (fig. 3H–K in Mayr and Goedert, 2016). Indeed, the trochlea metatarsi II of KMNH VP 60017 forms an angle of 144° relative to the lateral edge of the shaft, and the length/shaft width ratio is 4.6, significantly larger than the same angle and ratio in the trochlea metatarsi II of other plotopterids (134 to 135°, 3.4–3.7, Table 1), with the exception of Phocavis (154°, 4.4; see figs. 2, 3B in Goedert, 1988), which contributes to its noticeable constriction at the mid shaft. This large angle is primarily achieved by the less expanded proximal half, with the trochlea metatarsi II being more medially splayed than in Phocavis. In this aspect, it resembles other plotopterids more closely than Phocavis. The lateral foramen vasculare proximale is relatively large compared to Copepteryx and Hokkaidornis. The plantar surface of the shaft is convex and lacks a sulcus. The potential fossa metatarsi I on the medioplantar surface is obscured by a crack. An intertrochlear fossa is present between the trochleae of metatarsi III and IV, bordered distodorsally by low ridges on the lateral and medial walls of the trochlea (Figure 3E, F). A similar intertrochlear fossa is present in Copepteryx (Figure 3H), Hokkaidornis, and in some individuals of extant Aequornithes (Figure 3I, J). The trochlea metatarsi IV is relatively longer than in Phocavis and Olympidytes (Table 1). In lateral view, the dorsal profile of the trochlea metatarsi III is less prominent than in Phocavis, more closely resembling Hokkaidornis and ?Klallamornis in this respect. The trochlea grooves are weak. The absence of a growth plate along with the fusion of metatarsals without any discernible sutures and the well-ossified articular surfaces of the of the trochleae indicate that the specimen is not from a small juvenile individual but had reached a size comparable to that of an adult (see Watanabe and Matsuoka, 2013; Watanabe, 2017).
Remarks.—KMNH VP 60017 was discovered from Ainoshima, where specimens of Empeirodytes were also collected. The coracoid of Empeirodytes is approximately 60–70% the size of that Stenornis kanmonensis, which is comparable in size to Copepteryx hexeris (fig. 6 in Ohashi and Hasegawa, 2020). Similarly, KMNH VP 60017 is about 60% of the size of the metatarsal of Copepteryx hexeris (Table 1). Consequently, KMNH VP 60017 would roughly match the size of Empeirodytes. For these reasons, KMNH VP 60017 is potentially assignable to Empeirodytes okazakii. Nevertheless, Empeirodytes was based on two isolated coracoids, and no other parts of the taxon have been reported. Given that other possibilities cannot be ruled out, it is prudent to withhold a definitive classification until additional, more complete specimens are available.
The examination of KMNH VP 60017 revealed that this specimen exhibits a unique combination of characteristics found in various plotopterids. While these features offer promising avenues for future phylogenetic studies, the current lack of comprehensive data prevents a robust analysis at this time.
Of particular interest is the intertrochlear fossa observed in KMNH VP 60017 and related structures in other specimens. Most Aequornithes possess a foramen vasculare distale and an intertrochlear canal, which originates from this foramen and extends between the distal portions of the third and fourth metatarsals (Figure 4A, B). However, in two black-crowned night heron specimens (Nycticorax nycticorax, SINH-Av-23, SINH-AV-7) and one juvenile little egret (Egretta garzetta, OMNH-A-6011), the bony roof covering the dorsal intertrochlear canal is incomplete, forming an intertrochlear fossa (Figure 4C, D). The presence of this trait in individuals from a different genus suggests that this structure represents a minor variation within Aequornithes. Consequently, the intertrochlear fossa observed in plotopterids is likely developmentally equivalent to the intertrochlear canal or fossa observed in Aequornithes.
The intertrochlear structures observed in plotopterids are atypical among Aequornithes. Plotopterids exhibit variability in these structures: Phocavis only possesses the foramen vasculare distale (Mayr, 2004; also see Figure 4E, F), others solely have the intertrochlear fossa (Copepteryx, Hokkaidornis, and KMNH VP 60017; Figure 4G, H), whereas others lack both structures entirely. Given that both the foramen vasculare distale and intertrochlear canal are typically present in Aequornithes, even in suliform specimens, despite their accepted phylogenetic proximity to plotopterids, these features likely represent an ancestral or plesiomorphic condition in the family Plotopteridae. Although their phylogeny remains inconclusive, it is plausible that distinct evolutionary modifications, potentially involving independent loss of the foramen vascular distale and intertrochlear fossa, occurred within various plotopterid lineages. Notably, certain penguin specimens (king penguin, Aptenodytes patagonicus: OMNH 97-66; Humboldt penguin, Spheniscus humboldti: YIO-72480) also possess an intertrochlear fossa, albeit without ridges. It is conceivable that their ecological adaptations, akin to those of plotopterids, contributed to the development of this structure.
The intertrochlear canal or fossa serves as a conduit for a musculus extensor brevis digit IV, as well as blood vessels into the lateral intertrochlear incisure (Baumel and Witmer, 1993), which likely served for the interdigital web. The absence of the intertrochlear canal or fossa in certain plotopterids may be indicative of the loss or reduction of the web, as a result of diminished reliance on the hindlimbs for propulsion. The variable dependencies on the hindlimb in swimming among plotopterids are also suggested by Mori and Miyata (2021).
Regarding the intertrochlear fossa of KMNH VP 60017, an alternative interpretation is also feasible. Although fundamentally similar to that of Copepteryx and Hokkaidornis, the intertrochlear fossa of KMNH VP 60017 exhibits low ridges that resemble the structure observed in the snowy albatross (Diomedea exulans: OMNH-A161), which also possesses an intertrochlear canal in addition to this trait. This highlights the possibility that the fossa in KMNH VP 60017 may not be developmentally equivalent to those found in Copepteryx and Hokkaidornis. Should this be the case, it would indicate an even greater diversity in the intertrochlear structure of plotopterids.
The tarsometatarsus of KMNH VP 60017 differs from those of Copepteryx and Hokkaidornis, underscoring high morphological diversity among plotopterids in Japan. Recent studies (Matsuoka et al., 2014; Ohashi and Hasegawa, 2020) have suggested a greater diversity of plotopterids in Kyushu, which is supported by this study. Moreover, numerous earliest Oligocene? plotopterid specimens have been unearthed in Iwaki, Fukushima in northeastern Japan (Koda et al., 1991). Further research on these specimens may elucidate the diversity of plotopterids in Japan.
I thank Kiichiro Hachiya (Tokai Fossil Society, Nagoya) for donating the specimen that he found during his fieldwork. I also thank Gerald Mayr (Senckenberg Research Institute and Natural History Museum, Frankfurt) for his critical comments and for sending me a photo of Phocavis, and Tomoyuki Ohashi (KMNH) for his advice and for providing me with the photos of Copepteryx. K. Iwami (Yamashina Institute for Ornithology, Abiko), T. Wada (Osaka Museum of Natural History, Osaka), and S. Yachimori (Shikoku Institute of Natural History, Susaki) provided access to specimens under their care. S. A. McLeod (Natural History Museum of Los Angeles County, Los Angeles) kindly supplied the measurements for LACM 129405. I deeply appreciate the assistance of J. L. Goedert (Burke Museum of Natural History and Culture, Seattle) for providing me with photos of Phocavis and Klallamornis? clarki, critical comments, and help editing and personal observations of plotopterid specimens from Washington. I also thank anonymous reviewers for their critical comments.
