Eocene drill holes in cold-seep bivalves of Hokkaido, northern Japan


Kazutaka Amano, Department of Geoscience, Joetsu University of Education, Niigata, Japan.
E-mail: amano@juen.ac.jp


Four specimens of the thyasirid Conchocele bisecta (Conrad) and one small specimen of the vesicomyid Hubertschenckia ezoensis (Yokoyama), each with a drill hole made by a naticid gastropod, were found at a cold-seep site in the upper Eocene Poronai Formation of Hokkaido. This is apparently the oldest record not only of drill holes, but also of predation scars, in a cold-seep fauna. In addition, drilled vesicomyids are known from several Miocene cold-seep sites in Japan. We suggest that the Eocene and Miocene chemoautotrophic bivalves were drilled only in the shallow-water settings preferred by most naticids. The lack of drill holes in Oligocene chemoautotrophic bivalves in the northwestern USA suggests that this innovation, which allowed naticids to prey upon highly toxic bivalves, first appeared in the western Pacific during the Eocene.


Predation is one of the most important agents in evolution, as expressed in the hypothesis of escalation (Vermeij 1987). Drill holes are good indicators of predation, because they are relatively easy to distinguish from other durophagous features (Harper et al. 1998).

The temporal pattern of drilling predation has been discussed by many authors. Except for Walker & Brett (2002), most studies have shown that the frequency of drilled species among local assemblages increased from Cretaceous to Eocene (see also Appendix), reflecting predation escalation (Vermeij 1987, 2002; Kowalewski et al. 1998; Kelley & Hansen 2003). These findings were based mainly on studies of shallow-water faunas, as only two cases of fossil drilling predation on deep-water faunas are known. Hansen & Kelley (1995) observed an unexpectedly higher drilling frequency (21.2%) in lower-shelf deposits of the Eocene Yazoo Formation than in coeval shallow-water faunas. Walker (2001) also noted a high frequency of drilling predation in a late Pliocene bathyal fauna in Ecuador. These studies are intriguing, but more data are needed to elucidate the temporal pattern of deep-sea drilling-predation intensity.

Large and thriving chemosynthetic metazoan assemblages have been found in reducing deep-sea environments around hydrothermal vents, cold seeps, whale falls, sunken wood and kelp falls (e.g., Van Dover 2000; Smith et al. 2005). These faunas rely upon chemosynthetic bacteria, such as methane- and sulfur-oxidizing bacteria, for their major food source (Van Dover 2000). Their large biomass would be highly attractive for carnivores and scavengers. The main obstacle to accessing this abundant source of prey is the high concentration of reducing compounds in both the environment and the prey tissues. These reducing compounds, especially sulfides, are extremely toxic for metazoans because they bind to cytochrome-c oxidase and obstruct cellular aerobic respiration (Arp et al. 1984). Most Cenozoic-type vent and seep macrofaunal taxa appeared in the late Mesozoic (Van Dover et al. 2002; Little & Vrijenhoek 2003), when the Mesozoic Marine Revolution (Vermeij 1977, 1987) occurred. In order to migrate to such reducing environments, the predators had to acquire a physiological system for detoxifying sulfides.

The fossil record suggests that drilling predators expanded into such environments to prey upon chemosynthetic communities in the Miocene. Amano (2003a) reported many specimens of Calyptogena pacifica Dall and Conchocele bisecta (Conrad) drilled by naticids and muricids in the cold-seep association of the upper Miocene Morai Formation in western Hokkaido, Japan. In a report on Calyptogena from the Joetsu District in Niigata Prefecture, central Honshu, Amano & Kanno (2005) illustrated another upper Miocene, incompletely drilled specimen of C. pacifica from the Nodani Formation. While surveying the literature, we saw that Kamada (1962, pl. 1, Fig. 4) illustrated a single drilled specimen (reillustrated as Fig. 2C herein) of Calyptogena chitanii (Kanehara) as Adulomya chitanii Kanehara from the lower Miocene Taira Formation (Honya mudstone member) in Iwaki City, northeastern Honshu. Kamada (1962) overlooked its significance, and the specimen has apparently been ignored by subsequent workers. Thus, the oldest record of drilling predation remains in the early Miocene. It would be interesting to know whether naticid drilling predation on cold-seep bivalves was confined to Miocene deposits of Japan, and which predator species adapted to this environment prior to the Miocene.

Figure 2.

 Drilled cold-seep species. (A, D–F) Conchocele bisecta (Conrad); A: L = 18.2 mm, Joetsu University of Education (JUE) no. 15830-2; D: L = 30.6 mm, JUE no. 15830-1; E: L = 23.4 mm, JUE no. 15830-3; F: L = 19.5 mm, JUE no. 15830-4, Poronai Formation. (B) Hubertschenckia ezoensis (Yokoyama), L = 11.8 mm, JUE no. 15831. (C) Calyptogena chitanii (Kanehara), L = 38.2 mm, IGPS (Institute of Geology and Paleontology, Tohoku University) no. 87339, Taira Formation. White arrows show the drill holes.

As it turns out, we found drilled bivalves typical of chemoautotrophic communities in late Eocene cold-seep associations in Japan. In this study, these Eocene drill holes are described and discussed from an ecological point of view.

Material and Methods

The cold-seep deposits are located in a high cliff along the Ikushunbetsu River to the west of Yayoi Town in Mikasa City, central Hokkaido (Fig. 1). Four drilled specimens of Conchocele bisecta (Fig. 2A,D–F) and one small drilled specimen of Hubertschenckia ezoensis (Yokoyama) (Fig. 2B) were collected from a large columnar calcareous concretion (2 m in diameter) included in dark gray mudstone of the upper Eocene Poronai Formation (B zone of Teshima 1955; Fig. 3). We identified four mollusk species within this carbonate body, and 10 species in the surrounding mudstone, with only one species appearing in both lithologies (Table 1).

Figure 1.

 Locality with drilling cold-seep species.

Figure 3.

 Outcrop of seep carbonate in the Poronai Formation at the study locality. A large calcareous concretion (2 m in diameter) with chemosynthetic species is present in the dark gray mudstone. White line is 1 m scale.

Table 1.   Mollusks from the Poronai Formation at the study locality.
species nameCaMb
  1. Number in the table shows number of shells while number in parenthesis shows that of articulated bivalves.

  2. aFrom the carbonate rocks included in the mudstone.

  3. bFrom the mudstone surrounding the carbonates.

Yoldia sobrina Takeda 2 (2)
Yoldia sp.1 
Conchocele bisecta (Conrad)45 (42)3 (3)
Hubertschenckia ezoensis (Yokoyama)17 (16) 
Bathyancistrolepis sp.2 
Acila (Truncacila) picturata (Yokoyama) 5 (5)
Malletia poronaica (Yokoyama) 6 (6)
Portlandia (Portlandella) watasei (Kanehara) 6 (6)
Cyclocardia tokudai (Takeda) 17 (17)
Turritella? sp. 2
Orectospira wadana (Yokoyama) 8
Euspira? sp.11
Colus cf. fujimotoi Hirayama 1

In order to examine how the formation of the carbonate body was influenced by methane, the carbon and oxygen isotopic compositions of the carbonate rocks with Hubertschenckia and Conchocele were analyzed using micro-drilled powdered samples. The powdered samples, about 10 mg each, were reacted with 100% phosphoric acid in vacuo at 25 °C for 24 h. The evolved CO2 gas was analyzed in a dual inlet attached to a Finnigan MAT 252 stable isotope ratio mass spectrometer at the University of Tokyo. External precision of the standard was less than 0.03‰ for both carbon and oxygen isotopes.

In describing drilling predation, drilling intensity is calculated by using the formula D/(0.5 DV + AV) (D, valves with completely drilled holes; DV, disarticulated valves; AV, articulated valves) (see Hoffmeister & Kowalewski 2001; Kowalewski 2002). Naticids make characteristic parabolic holes on their prey (Kitchell et al. 1981; Kabat 1990). Moreover, Kitchell et al. (1986) noted that nonfunctional naticid holes are those in which the ratio of inner borehole diameter (IBD) to outer borehole diameter (OBD) is less than 0.5. Thus, for examining hole morphology, we measured both diameters.

The Yayoi fauna consists of extinct species, so its paleobathymetry was inferred by comparison with modern analogs (Higo et al. 1999) at the generic level. The shell length, drilled site, and OBD and IBD were measured and examined on drilled prey. All collected specimens are housed in the Joetsu University of Education (JUE).

Occurrence and paleobathymetric depth

The carbonate concretion at Yayoi site has δ13C values (relative to PDB; belemnite from the Pee Dee Formation in US) of +5.8‰ to −44.7‰ (R.G. Jenkins, unpublished data). Highly positive δ13C values imply methane formation rather than oxidation, whereas highly negative δ13C values imply anaerobic oxidation of methane (Peckmann & Thiel 2004). The values at the Yayoi site strongly suggest that the precipitation of the concretion was influenced by methane.

The methane seep-related carbonate at this locality yielded articulated valves of three bivalve species and a single shell of two gastropods (Table 1). Among the collected bivalves, the thyasirid C. bisecta and the vesicomyid H. ezoensis are characteristic of chemosynthetic assemblages. Hubertschenkia was described by Takeda (1953), with the type species Tapes ezoensis (Yokoyama 1890), and was placed in the Veneridae and not compared with vesicomyids. Kanno & Teshima (1994) pointed out that the morphology of Hubertschenkia is very similar to Calyptogena Dall. Conchocele bisecta is represented mainly by small specimens, which generally are smaller than 30.6 mm in shell length. The buccinid gastropod Bathyancistrolepis sp. is most similar to the Recent Bathyancistrolepis trochoideus (Dall) and it is the first record of this gastropod from the Poronai Formation. Other than the aforementioned species, the fauna at Yayoi site consists of 10 heterotrophic species including numerous taxodont bivalves, the carditid Cyclocardia tokudai (Takeda) and the naticid Euspira? sp., but their shells other than Euspira? sp. were found only in the mudstone surrounding the carbonate body (Table 1, Fig. 4).

Figure 4.

 Some characteristic non-drilled species from the calcareous concretion (Co) and the surrounding mudstone (M). (A) Orectospira wadana, H = 14.6 mm+, M; (B) Portlandia (Portlandella) watasei (Kanehara), L = 30.4 mm, M; (C) Malletia poronaica (Yokoyama), L = 19.7 mm, M; (D) Cyclocardia tokudai (Takeda), L = 19.0 mm, M; (E) Euspira? sp., H = 26.7 mm, M; (F) Acila (Truncacila) picturata (Yokoyama), L = 25.5 mm, M; (G) Hubertschenckia ezoensis (Yokoyama), L = 74.8 mm, Co.

The Recent species of Cyclocardia, Orectospira and Bathyancistrolepis are known from depths between 100 and 400 m and we infer such a depth for our locality during precipitation of the carbonate body. The faunal composition of the molluskan assemblage in the surrounding mudstone is similar to the lower sublittoral to upper bathyal Malletia poronaicaC. tokudai assemblage of Suzuki (2000) from the lower part of the Poronai Formation. Kaiho (1984) has inferred that the maximum depth of the Poronai Formation was 350 m, based on benthic foraminifers.

Details of drilled holes

Among 45 specimens (42 articulated) of C. bisecta, two right and two left valves were drilled. The shell material of the smallest specimen was dissolved (Fig. 2A), therefore the outer diameter of the boreholes was measured on the other three specimens. These boreholes are parabolic in shape, with outer diameters ranging from 2.2–3.0 mm (IBD/OBD = 0.52–0.67) (Table 2). These holes were drilled in the upper central parts of the valves (Fig. 2A,D–F) and they closely resemble the position of the holes which Amano (2003a) observed on younger specimens in the Miocene Morai Formation. It is difficult to compare drilling intensity between the new Eocene material of C. bisecta and the Miocene Morai sample, because of the low number of specimens. However, the apparent drilling intensity is similarly low in both the Eocene C. bisecta (0.09) and the Miocene Morai samples (0.09).

Table 2.   Measurements of drilled specimens.
speciesnumbervalvelength [mm]OBD [mm]IBD [mm]IBD/ OBD
  1. JUE = Joetsu University of Education; OBD = outer borehole diameter; IBD = inner borehole diameter.

Conchocele bisectaJUE no. 15830-1right30.
JUE no. 15830-2right18.22.0
JUE no. 15830-3left23.
JUE no. 15830-4left19.
Hubertschenckia ezoensisJUE no. 15831left11.

Among 17 specimens of H. ezoensis (16 articulated), one small left valve (shell length = 11.8 mm) was drilled. The borehole is parabolic in shape, with an outer diameter of 1.0 mm (IBD/OBD = 0.70) (Table 2). This hole was drilled near the postero-dorsal margin (Fig. 2B) and its position resembles that of the holes observed by Amano (2003a) in Calyptogena pacifica from the Miocene Morai Formation.


Kelley & Hansen (1993) recorded drilled Lucina parva from the Upper Cretaceous Ripley Formation in Georgia, southeastern USA, which may be the oldest record of drilled chemoautotrophic bivalves. Most lucinids harbor endosymbiotic bacteria in their gills, while some lucinids are heterotrophic (Fisher 1990; Taylor & Glover, 2000). Moreover, the Ripley fauna did not live at a cold seep.

Kiel (2006) described healed injuries caused by crabs on the shell surfaces of two gastropod genera, Provanna and Turrinosyrinx, from the Oligocene Lincoln Creek Formation in Washington, USA. This is the oldest preexisting record of predation in a cold-seep fauna. However, as shown in the present study, Eocene Conchocele and Hubertschenckia in the Poronai Formation were possibly bored by Euspira? sp. Based on our results, the evidence for predation in the cold-seep fauna goes back to the late Eocene. Moreover, naticids are not likely to have accidentally wandered into a cold-seep site, but evidently have been predatory members of cold-seep faunas since the late Eocene to the late Miocene in the northwestern Pacific.

The high frequency of Cenozoic drilled shells is apparently related to the enhanced radiation of naticids in the Eocene (Kase & Ishikawa 2003), and also from this epoch we note the oldest record of predation on cold-seep bivalves. We infer that naticids adapted to prey in a reducing environment (i.e., cold seeps) as one aspect of their pronounced radiation in the Eocene.

The traces of naticid drilling activity are better known from the shells of vesicomyid bivalves (Calyptogena chitanii and Calyptogena pacifica) in the Miocene deposits of Japan (Kamada 1962; Amano 2003a; Amano & Kanno 2005). The drilling intensity (0.20) on C. pacifica from the Morai Formation in Hokkaido corresponds to that (0.10–0.17) for the shallow-water boreal species Glycymeris yessoensis (Sowerby) in lower Pleistocene deposits of Hokkaido (Amano 2003b).

In contrast, no drilled specimens and few naticids have been recorded from Recent vent and seep faunas. When summarizing the gastropod occurrences in the Recent vents and seeps, Warén & Bouchet (2001) listed no naticid species other than Naticidae sp. from Aleutian Trench seeps at 4774–4947 m depth. Sasaki et al. (2005) reported no naticids in Recent cold seep and hydrothermal vent faunas around Japan.

This phenomenon might be due to the significant depths of most Recent vents and seeps examined by Warén & Bouchet (2001) and Sasaki et al. (2005). With the exception of a few Euspira species, most naticids in the Japanese Islands live in shallow water (Higo et al. 1999). The paleobathymetry of the fossil chemosynthetic communities with drilled bivalves discussed here ranges from the lower sublittoral to the uppermost bathyal zone. This suggests that the fossil seep mollusks were drilled by naticids in relatively shallow-water chemosynthetic communities (see also Majima et al., 2005).

Several specimens of Natica sp., drilled most probably by others of the same species, are known from the late Oligocene seep community of the Lincoln Creek Formation in the northwestern USA (Peckmann et al. 2002), but surprisingly there is no record of predation on chemosynthetic species at that site. Some other factor evidently discouraged naticids from drilling chemosynthetic species. When Kicklighter et al. (2004) fed the tissue of chemosynthetic species, including Calyptogena magnifica, to shallow-water fishes and crabs, some of the tissue was unpalatable to the predators. These workers claimed that a bacteria-produced chemical deterrent other than hydrogen sulfide was responsible for deterring predators. With this in mind, it is possible that the Lincoln Creek Formation naticids could not cope with the chemical deterrent in chemoautotrophic bivalves, even though their Japanese counterparts could.


We are very grateful to Geerat J. Vermeij (UC Davis), Kazushige Tanabe (University of Tokyo), Lisa Levin (Scripps Institution of Oceanography), Jim Barry (Monterey Bay Aquarium Research Institute), Gregory Dietl (North Carolina State University), Patricia Kelley (University of North Carolina-Welmington), Andrzej Kaim (Institute of Paleobiology, Polish Academy of Sciences, Warsaw), and Louie Marincovich (California Academy of Sciences) for their reviews of our manuscript and their useful information. We thank Masanori Shimamoto (Tohoku University) for his help in examining the fossil specimens. We also thank Maurice E. Jenkins for checking our English. We are also indebted to the late Hiroshi Hayakawa for facilitating fieldwork and for his hospitality. This research was partly supported by a Grant-in-aid for Scientific Research from the Japan Society for the Promotion of Science (C, 15540450, 2003–2005). Our study was also carried out as part of the Japanese Scientific Research Fund of the 21st Century Center of Excellence Program at the University of Tokyo (G3, leader T. Yamagata).


Table Appendix..   Geologic time scale from Cretaceous to Neogene (Gradstein et al. 2004).
  1. Ma = 106 years before present.

  Holocene (0.01–0 Ma)
  Pleistocene (1.81–0.01 Ma)
  late (3.60–1.81 Ma)
  early (5.33–3.60 Ma)
  late (11.61–5.33 Ma)
  middle (15.97–11.61 Ma)
  early (23.03–15.97 Ma)
  late (28.45–23.03 Ma)
  early (33.90–28.45 Ma)
  late (37.20–33.90 Ma)
  middle (40.40–37.20 Ma)
  early (55.80–40.40 Ma)
  late (58.70–55.80 Ma)
  middle (61.70–58.70 Ma)
  early (65.50–61.70 Ma)
 Cretaceous (145.5–65.50 Ma)