2026 年 30 巻 p. 9-16
In this paper, we described an ant fossil discovered in the Middle to Late Miocene Fuzawa Formation, Aizu region, Fukushima, Japan. This fossil was described as a new species of the Subfamily Myrmicinae Lepeletier de Saint-Fargeau, 1835, Paraphaenogaster tanemurai sp. nov., having long, straight third cubital vein reaching apical margin, and long hindfemur. The genus Paraphaenogaster is extinct, with 22 species reported worldwide, but its distribution is centered in Europe. In Asia, only four species are known from China and one from Japan. This study adds a second new species from Japan, suggesting that this genus also exhibited diversity in East Asia. This is the first discovery of insect fossil from the Fuzawa Formation and the second report of a new ant fossil from Japan.
ZooBank registration: urn:lsid:zoobank.org:pub:A0A65E16-22FF-4D89-A088-8179063C8A25

Ants are eusocial insects belonging to the family Formicidae and are classified within the order Hymenoptera along with bees. They are distributed across all continents except Antarctica and are considered the most successful insects (Hölldobler and Wilson, 1990). According to AntCat (Bolton, 2025), there are 16 subfamilies, 38 tribes, 346 genera, and 14,343 extant species in the world. In addition, fossils from six subfamilies, six tribes, 177 genera, and 837 species have been reported.
The fossil record of ants in Japan comprises at least 20 species (Aiba and Terayama, 2020). The oldest reports are from the Upper Cretaceous Tamayama Formation in Iwaki, Fukushima (Ogata et al., 2005) and the Upper Cretaceous Kunitan Formation in Kuji, Iwate (Kubota and Kubota, 2012). Both were found in amber but remain undescribed. Eight ant fossils have been reported from the Miocene of Nagasaki (Fujiyama, 1970), Fukui (Yasuno, 1979), Niigata (Fujiyama, 1985), and Yamagata (Fujiyama, 1985) and from the Pliocene of Hyogo (Inoue, 1986) and Gunma (Tanaka and Mano, 2017). However, only one species, Paraphaenogaster avita (Fujiyama, 1970) from the Middle Miocene in Iki, Nagasaki, has been described as a new species among these fossils, and others were not diagnostic. There are reports of ant fossils from the Quaternary deposits of Gifu and Tochigi, but their alpha taxonomy remains unstudied (Aiba and Terayama, 2020; Takahashi and Aiba, 2023).
The Middle to Upper Miocene Fuzawa Formation is distributed in the Aizu region of Fukushima, Japan, where a variety of animal and plant fossils have been discovered (Uemura, 2004). However, no insect fossil has been reported to date. Here, we report and describe the first occurrence of an insect from the Fuzawa Formation, a fossil ant discovered in 2014.
The Fuzawa Formation is a marine deposit from the Middle to Late Miocene that is distributed in Fuzawa in Tadami Town, Showa Village, Komadome Pass in Tajima Town, and the vicinity of Nanatsugatake, Fukushima (Figure 1). The Fuzawa Formation is approximately 200 m thick (Yamamoto and Komazawa, 2004), and it is informally divided into the lowest, lower, middle, and upper parts (Kitamura et al., 1968). The lower part consists mainly of rhyolitic volcaniclastic rocks, whereas the upper part consists mainly of dark grey mudstones with well-developed plate-like bedding (Yamamoto and Komazawa, 2004). Numerous plants, sea urchins, polychaetes, and fish fossils have been found in these mudstones (Taketani, 2001; Uemura, 2004; Kikuchi and Uemura, 2006; Inose and Jimi, 2024). Yamaguchi (1986) reported a K–Ar age of 14.0 ± 0.7 Ma from a biotite rhyolite tuff in the lower part. Yamamoto and Komazawa (2004) estimated the age of the Fuzawa Formation to be 15–10 Ma (Middle to Late Miocene) based on the microfossil biostratigraphy of the Ninosawa and Yuzuritoge formations, which are comparable facies to the Fuzawa Formation within the Aizu Basin. The fossiliferous upper part of the Fuzawa Formation correlates with the Yuzuritoge Formation, which is considered to be 13–10 Ma based on calcareous nannofossil biostratigraphy (Aita et al., 1998). This constrains the age of the ant fossil reported in this study to the late Middle to early Late Miocene.

The ant fossil was discovered by Ryunosuke Tanemura in September 2014 from the mudstone of the upper part of the Middle to Late Miocene Fuzawa Formation. The studied specimen is housed at the Fukushima Museum under the repository number FM-N201900005. The institutional abbreviation “FM” represents the Fukushima Museum.
The specimen was examined under a Leica M205 C stereo microscope (Leica Corporation, Wetzlar, Germany). Photographs and measurements were obtained using a Leica MC170 HD microscope camera with Leica Application Suite Version 4.1.3. The images were sharpened and adjusted for contrast and tonality using Adobe Photoshop version CS6 (Adobe Systems Inc., San Jose, CA, USA). The ant forewing vein nomenclature used herein follows Perfilieva et al. (2017) and Jessen (2020) and is consistent with the established terminology for Hymenoptera (Rasnitsyn, 1980).
Abbreviations for wing veins and cells are as follows: A = anal vein, C = costa, Cu = cubital vein, M = median vein, R = radial vein, RS = radial sector, cu-a = cross vein between Cu and A, r-rs = cross vein between R and RS, and m-cu = cross vein between M and Cu; cua = anterior cubital cell, mcu = mediocubital cell, and r = radial cell.
Order Hymenoptera Linnaeus, 1758
Family Formicidae Latreille, 1809
Subfamily Myrmicinae Lepeletier de Saint-Fargeau, 1835
Tribe Stenammini Ashmead, 1905
Genus Paraphaenogaster Dlussky, 1981
Paraphaenogaster tanemurai sp. nov.
[New Japanese name: Tanemura-mukashi-ashinaga-ari]
ZooBank lsid: urn:lsid:zoobank.org:act:0E27F058-6858-4810-A9D7-EF84B7A7B018


Holotype.—FM-N201900005, consisting of concave dorsal part (FM-N201900005a: Figure 2A) and convex ventral counterpart (FM-N201900005b: Figure 2B) portions of a male individual preserving the mesosoma, petiole, postpetiole, right and left forewings, right and left hindwings, part of left foreleg, part of right and left midlegs, part of right and left hindlegs, and gaster. The head is not preserved.
Type locality and horizon.—The holotype was collected from a mudstone horizon in the upper part of the Fuzawa Formation, which is distributed along a stream approximately 600 m east of the village Nonosawa in Yanatori, Tadami Town, Minamiaizu District, Fukushima, Japan (Figure 1).
Etymology.—After Ryunosuke Tanemura, the fossil collector who discovered the specimen.
Diagnosis.—Sclerotized black body with two isolated segments (petiole and postpetiole). Vein rs-m absent. Vein 3Cu long and straight, parallel to vein 4M, and reach apical margin of wing. Hindfemur long, longer than mesosoma. Gaster small, half the length of mesosoma.
Description.—Mesosoma and gaster black, sclerotized. Both wings dark in color. Estimated total body length 7 mm.
Mesosoma (Figures 2A, 3A): Sclerotized black, hairs and punctures not preserved dorsally, promesonotal suture and metanotal groove faintly visible dorsally, 3.14 mm long and 2.08 mm wide; pronotum swollen dorsally without teeth, 1.21 mm long, scutum not visible; mesonotum without teeth, 1.49 mm long and 1.74 mm wide in middle; propodeum without spines or teeth, 0.47 mm long and 0.98 mm wide in middle.
Petiole long, without spines, and with high, distinct node. Postpetiole partially preserved, however, details unclear due to poor preservation.
Forewing (Figure 3A–C): Right and left forewings almost completely preserved. Length 8.40 mm (right)/7.97 mm (left), maximum width 2.63 mm (right)/2.81 mm (left). Cells mcu and 1+2r closed. Cells 3r and cua open. Shape of cell mcu trapezoid and medium-sized. Cell rm not developed, because vein rs-m absent. Pterostigma well developed. Crossvein 2r-rs directly vertical to pterostigma bottom margin and relatively short. Veins 5RS and 4M long and straight, do not reach the apical margin of wing. Veins 2RS+M and 2Cu short and almost same length. Vein 3Cu long and straight, parallel to vein 4M, and reach posterior margin of wing.
Hindwing (Figure 3A): Right and left hindwings preserved, but right one overlaps forewing in part. Length 5.64 mm (right)/7.65 mm (left), maximum width 1.58 mm (right)/2.77 mm (left). Veins poorly preserved, and only part of vein M and vein 2Cu visible. Measurements of forewing venation length in mm (right wing/left wing): 1A (2.15/1.97), 2A (1.90/2.01), 1Cu (0.81/0.78), 2Cu (0.31/0.30), 3Cu (2.48/2.54), cu-a (0.32/0.36), 1M (0.53/0.58), 4M (3.21/3.30), m-cu (0.59/0.54), 1M+Cu (2.08/1.86), 2M+Cu (0.82/0.72), R (3.42/3.20), 1RS (0.44/0.40), 1RS+M (0.39/0.41), 2RS+M (0.46/0.38), 3RS (0.95/0.91), 5RS (2.91/2.11), 2r-rs (0.40/0.36). Measurements of hindwing venation length in mm (right wing/left wing): 1A (1.06/1.13), cu-a (0.25/0.24), 1M+Cu (1.15/1.09), 2M+Cu (0.76/0.74).
Legs (Figures 2A, 3A): Left foreleg, left midleg, left hindleg, right midleg, and right hindleg preserved. All long and thin, in particular, hindfemur long, longer than mesosoma, slightly sclerotized, brown in color and lighter than body. Only femur and tibia preserved, tibial spurs and tarsus not preserved. Measurement of leg in mm (length/width): Left forefemur (2.02/0.17), left foretibia (1.78/0.14), left midfemur (2.41/0.17), left midtibia (2.47/0.13), left hindfemur (3.36/0.32), left hindtibia (3.08/0.16), right midfemur (> 1.67/0.16), right midtibia (> 1.73/0.11), right hindfemur (> 2.78/0.27), right hindtibia (> 2.53/0.12).
Gaster (Figures 2A, B; 3A): Small, half-length of mesosoma. Sclerotized black, near apex brown, almost circular, 1.68 mm long, widest at posterior third, 1.69 mm wide. Boundary between first and second gastral segments unclear.
The following diagnostic characteristics of the genus Paraphaenogaster (Dlussky, 1981; Perfilieva et al., 2017; Jessen, 2020) are present in FM-N201900005—closed cells mcu and 1+2r, vein 5RS does not reach R, cell 3r remains open, cell rm and vein rs-m are absent, crossvein 2r-rs is directly vertical to the pterostigma bottom margin and relatively short, cell mcu trapezoid and medium-sized, second cubitoanal cell not formed, postpetiole can be wider than petiole. Therefore, FM-N201900005 can be unambiguously referred to this genus.
Paraphaenogaster was described by Dlussky (1981) based on males from Miocene deposits of the Vishnevaya Balka locality in Stavropol, Russia. It is similar in body size and wing morphology to the genus Aphaenogaster Mayr, 1853 (Radchenko and Perkovsky, 2016). The only distinguishing feature between Paraphaenogaster and Aphaenogaster is the lack of or a strong reduction in cell rm. In the past, several ant fossils from the Oligocene to the Miocene have been described as belonging to the extant genus Aphaenogaster (e.g. A. avita Fujiyama, 1970; A. maculata, Théobald, 1937; A. maculipes Théobald, 1937 and A. pannonicus Bachmayer, 1960). These were transferred to Paraphaenogaster because they did not have vein rs-m (Radchenko and Perkovsky, 2016). Recently, six species from the upper Oligocene of Germany (Enspel) were described: P. bizeri Jessen, 2020; P. freihauti Jessen, 2020; P. loosi Jessen, 2020; P. schindleri Jessen, 2020; P. wettlauferi Jessen, 2020 and P. wuttkei Jessen, 2020. Currently, 22 named species of fossil ants belonging to Paraphaenogaster have been identified worldwide (Antwiki, 2024) (Table 1).
| Taxon | 3Cu | 4M | Hindfemur | Gaster | Age (Ma) | Epoch | Locality | Country |
|---|---|---|---|---|---|---|---|---|
| Paraphaenogaster pannonica (Bachmayer, 1960) | ++ | ++ | – | – | 7.2–5.3 | Late Miocene | Brunn-Vösendorf | Austria |
| P. macrocephala (Heer, 1849) | – | – | – | ++ | 12.7–11.6 | Middle to Late Miocene | Oeningen | Switzerland |
| P. tanemurai sp. nov. | +++ | ++ | +++ | + | 13.0–10.0 | Middle to Late Miocene | Tadami Town, Fukushima | Japan |
| P. aemula (Heer, 1849) | +? | – | – | – | 13.7–12.7 | Middle Miocene | Parschlug | Austria |
| P. dumetorum (Lin, 1982) | ++ | +++ | – | ++ | 16.0–11.6 | Middle Miocene | Shanwang | China |
| P. lapidescens (Zhang, 1989) | ++ | ++ | + | +++ | 16.0–11.6 | Middle Miocene | Shanwang | China |
| P. microphthalma Dlussky, 1981 | + | ++ | – | – | 16.0–11.6 | Middle Miocene | Vishnevaya Balka Creek, Stavropol | Russia |
| P. paludosa (Zhang, 1989) | ++ | +++ | + | + | 16.0–11.6 | Middle Miocene | Shanwang | China |
| P. shanwangensis Hong, 1984 | ++ | +++ | + | +++ | 16.0–11.6 | Middle Miocene | Shanwang | China |
| P. avita (Fujiyama, 1970) | ++? | +++ | + | – | 16.0–13.7 | Middle Miocene | Chōjabaru, Iki Island | Japan |
| P. bohemica (Novák, 1877) | + | + | ++ | – | 16.9–16.0 | Early Miocene | Mokrina (Krottensee) | Czechia |
| P. jurinei (Heer, 1849) | – | – | – | +++ | 20.4–16.0 | Early Miocene | Radoboj | Croatia |
| P. tertiaria (Heer, 1849) | + | + | – | +++ | 20.4–16.0 | Early Miocene | Radoboj | Croatia |
| P. bizeri Jessen, 2020 | – | – | + | +++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. freihauti Jessen, 2020 | – | – | + | +++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. loosi Jessen, 2020 | – | – | + | ++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. schindleri Jessen, 2020 | + | ++ | + | ++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. wettlauferi Jessen, 2020 | – | – | + | +++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. wuttkei Jessen, 2020 | + | + | +++ | ++ | 24.8–24.6 | late Oligocene | Enspel Oilshale | Germany |
| P. ussuriensis Perfilieva, 2022 | + | ++ | – | – | 28.1–23.0 | Oligocene | Velikaya Kema, Primorye | Russia |
| P. maculata (Théobald, 1937) | + | + | + | ++ | 28.4–23.0 | Oligocene | Aix-en-Provence | France |
| P. maculipes (Théobald, 1937) | +? | +? | – | ++ | 37.3–33.9 | Eocene | Kleinkems | Germany |
| P. hooleyana Dlussky and Perfilieva, 2014 | ++ | ++ | – | – | 38.0–28.0 | late Eocene to early Oligocene | Isle of Wight | UK |
FM-N201900005 can be clearly distinguished from other species by the following points: vein 3Cu is long and straight, reaches the posterior margin of the wing; the hindfemur is long, longer than the mesosoma; and the gaster is small, half the length of the mesosoma (Table 1).
There is no species in which vein 3Cu reaches the posterior margin of the wing. The only possible example is P. avita (Fujiyama, 1970), which was described from the Middle Miocene of Iki Island in Nagasaki, Japan. In FM-N201900005, the vein 3Cu is long and preserved near the posterior margin. However, the posterior margin is missing in P. avita (see Fujiyama, 1970, fig. 3), so it is unclear whether it reaches the margin. Nevertheless, FM-N201900005 can be clearly distinguished by the upward curve of the vein 3Cu, the vein 4M reaching the apical margin and the short hindfemur. Many fossils show that vein 3Cu is short and curved (e.g. the type species P. microphthalma). The following six species have relatively long 3Cu veins: P. dumetorum (Lin, 1982), P. hooleyana Dlussky and Perfilieva, 2014, P. lapidescens (Zhang, 1989), P. paludosa (Zhang, 1989), P. pannonica (Bachmayer, 1960) and P. shanwangensis Hong, 1984. However, 3Cu does not extend to the posterior margin of the wing in any of those species. Another characteristic of FM-N201900005 is that the hindfemur is longer than the mesosoma. The only other fossil species with a hindfemur clearly longer than the mesosoma is P. wuttkei Jessen, 2020, but it can be distinguished from FM-N201900005 by its shorter 3Cu and 4M veins. Another diagnostic characteristic of FM-N201900005 is that gaster is smaller than the mesosoma. In most fossil ants, the gaster is approximately the same size as or larger than the mesosoma. The only exception is P. paludosa (Zhang, 1989), which has a small gaster, but this species can be distinguished from FM-N201900005 by the morphology of the 3Cu and 4M veins and its short hindfemur. The gaster is often damaged in fossil ants, and its length can vary because of telescoping variability (Tschinkel, 2013). The boundary between the first and second gastral segments of FM-N201900005 is not visible. If only the first gastral segment is visible, the actual gaster length may be close to the length of the mesosoma. However, even considering damage, the gaster of FM-N201900005 would not be longer than the mesosoma. Therefore, FM-N201900005 can be distinguished from any currently known species and is regarded as a new extinct species.
The oldest record of the genus Paraphaenogaster is P. hooleyana (Dlussky and Perfilieva, 2014), which was discovered from the upper Eocene deposits on the Isle of Wight, UK. Subsequently, it was discovered to be widely distributed in Oligocene to Upper Miocene deposits throughout Europe (Table 1). On the other hand, reports from East Asia include only four species from China and two from Japan, all restricted to the Middle to Upper Miocene (Table 1). All of these species have long 3Cu and 4M veins and are morphologically similar (Table 1). This suggests that these species may be closely related. The discovery of a new species of Paraphaenogaster provides valuable data for considering the diversity and evolution of this genus in East Asia.
We would like to express our gratitude to Ryutarou Tanemura for discovering this specimen. We would also like to thank Harufumi Nishida (Chuo University) for providing us with the opportunity to conduct this research. Additionally, we are grateful to the editors and the two anonymous reviewers for their critical comments on this manuscript.
HA initiated the study, drafted the manuscript, and compiled all the figures. HI revised the manuscript and provided geological input. Both authors contributed to the writing of the manuscript.