Introduction
The Lepidopteran fossil record is notoriously incomplete because of its paucity and biases (Kapoor, Reference Kapoor1981; Kristensen and Skalski, Reference Kristensen, Skalski and Kristensen1999; Sohn et al., Reference Sohn, Labandeira and Davis2015). The body fossils of microlepidopterans assignable to a family level with relevant evidence are critically scarce (Sohn et al., Reference Sohn, Labandeira and Davis2015; de Jong, Reference de Jong2017). In such cases, their characteristic feeding damage, e.g., leaf mines preserved on fossilized host plants, acts as an important supplement. Moreover, such fossils provide direct evidence of the interactions between lepidopterans and plants (Labandeira et al., Reference Labandeira, Dilcher, Davis and Wagner1994; Lopez-Vaamonde et al., Reference Lopez-Vaamonde, Wilkström, Labandeira, Godfray, Goodman and Cook2006; Donovan et al., Reference Donovan, Wilf, Labandeira, Johnson and Peppe2014).
Nepticulidae is one of the major leaf-mining insect groups, currently including 22 genera and 862 species worldwide (van Nieukerken et al., Reference van Nieukerken, Doorenweerd, Hoare and Davis2016). The adult moths are the smallest in Lepidoptera, hence are commonly known as pygmy moths, and represent some of the earliest diverging lineages in lepidopteran phylogeny (Regier et al., Reference Regier, Mitter, Kristensen, Davis, van Nieukerken, Rota, Simonsen, Mitter, Kawahara, Yen, Cummings and Zwick2015). The larvae exhibit various forms of endophagous herbivory, of which the majority are leaf miners. The larval host plants are primarily dicotyledonous, arboreal angiosperms, although some lineages are associated with monocotyledons or herbaceous hosts (Davis, Reference Davis and Kristensen1999). Nepticulidae could have been among the first lepidopteran groups to start using angiosperms (Doorenweerd et al., Reference Doorenweerd, van Nieukerken and Menken2016). Currently, there are 79 records of nepticulid fossils, as both body and trace fossils, known in the world (Sohn et al., Reference Sohn, Labandeira, Davis and Mitter2012; Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015). The earliest confirmed fossil occurrence of Nepticulidae was 102 million years ago; since then, there has been a fairly regular occurrence of such mines in the fossil record (Labandeira et al., Reference Labandeira, Dilcher, Davis and Wagner1994; Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015).
In their review, Doorenweerd et al. (Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015) described the diagnostic features of nepticulid leaf mines as: (1) the presence of an egg-case or its trace involving the oviposition by a female at the beginning of the leaf mine, although this can be secondarily lost; (2) the path of the leaf mine varying species by species, but often limited by the midvein or other prominent veins on a leaf; (3) the patterns of frass deposition variable, but a frass trail in meniscate arcs is exclusively seen in some nepticulid leaf mines; and (4) the presence of a semicircular slit representing the exit hole for pupation near the end of the leaf mine.
This study describes a leaf-mine fossil from the early Miocene Geumgwangdong Formation in Pohang Basin, South Korea, identifies the mine-maker (Nepticulidae), and reviews the occurrence of the nepticulid leaf mines in the Miocene.
Geological setting
Stratigraphic information
The fossil specimen in this study was discovered from the Geumgwangdong Formation in Pohang City, Republic of Korea (Fig. 1.1). This formation belongs to the Janggi Group, which is a representative fossil bed for Neogene plant leaves on the Korean Peninsula (Yoon, Reference Yoon2010). The paleoflora of the Janggi Group are comparable to the Aniai-type flora typified by cool-temperate vegetation during the early Miocene (Huzioka, Reference Huzioka1972). Coinciding with this, Ar-Ar dating estimated the age of the rocks in the Janggi Basin as belonging to the early Miocene, ca. 21.5–14.6 Ma (Paik et al., Reference Paik, Kang, Kim, Lee, Kim and Jeong2010). The Geumgwangdong Formation (Figs. 1.2, 1.3) is ~ 70 m thick and dominated by lacustrine deposits consisting of paper shale, shaly mudstone, laminated silty mudstone, and mudstone (Paik et al., Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012).
Figure 1. Maps and photographs of the study area: (1) geological map of Pohang Basin (after Paik et al., Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012), with black arrow indicating study area; (2) fossil locality along a mountain slope, with arrow pointing to fossiliferous horizon; (3) a pile of paper shales taken off the mother beds nearby.
Paleobiota
The Geungwangdong Formation has yielded a large number of the compression/impression fossils of leaves and fruits of various plant species (Kanehara, Reference Kanehara1936; Huzioka, Reference Huzioka1972; Chun, Reference Chun1982; Kim and Choi, Reference Kim and Choi2008; Jeong et al., Reference Jeong, Kim, Uemura, Paik and Kim2010). Paik et al. (Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012) listed 64 taxa of fossil plants belonging to 27 families and 43 genera in the formation. The dominant groups included Fagus Linnaeus, Reference Linnaeus1753, Quercus Linnaeus, Reference Linnaeus1753 (Fagaceae), Betula Linnaeus, Reference Linnaeus1753 (Betulaceae), Zelkova Spach, Reference Spach1841 (Ulmaceae), Acer Linnaeus, Reference Linnaeus1753 (Sapindaceae), Leguminosae, and Metasequoia Hu and Cheng, Reference Hu and Cheng1948 (Cupressaceae), which represented tall tree vegetation from mountain slopes and valleys (Paik et al., Reference Paik, Kang, Kim, Lee, Kim and Jeong2010). In coincidence with this, Mollusca and Estheria, two groups commonly occurring in freshwater lake sediments, have not been found in the formation (Paik et al., Reference Paik, Kang, Kim, Lee, Kim and Jeong2010). Several animal remains have been discovered from the Geumgwangdong Formation (unpublished data, K-S Nam). However, only a few of these, including fish bones (Yoon, Reference Yoon1992) and insects (Kim and Lee, Reference Kim and Lee1975; Kim, Reference Kim1976), have been studied.
Materials and methods
Observation
The fossil leaf was examined using a stereomicroscope (Leica EZ4) and photographed using a digital camera (Nikon D40) in a general light box or with an adapter connected to a dissecting microscope (Leica MZ6).
Terminology
Terms for leaf and leaf-mine morphologies follow Ellis et al. (Reference Ellis, Daly, Hickey, Johnson, Mitchell, Wilf and Wing2009) and Hering (Reference Hering and Hering1951), respectively. Sequential numbers of the secondary (2°) veins in the leaf were counted successively from the most basal vein.
Repository and institutional abbreviation
The type specimen examined in this study is deposited in the collection of the Daejeon Science High School for the Gifted, Daejeon, South Korea (DSHS).
Systematic paleontology
Order Lepidoptera Linnaeus, Reference Linnaeus1758
Suborder Glossata Fabricius, Reference Fabricius1775
Superfamily Nepticuloidea Stainton, Reference Stainton1854
Family Nepticulidae Stainton, Reference Stainton1854
Genus Stigmellites Kernbach, Reference Kernbach1967
1967 Stigmellites Kernbach, p. 104.
Type species
Stigmellites heringi Kernbach, Reference Kernbach1967.
Remarks
Stigmellites was originally established as an ichnogenus but later changed to a collective-group name (for a full history, see Sohn and Lamas, Reference Sohn and Lamas2013). According to this change, Stigmellites included the nepticulid fossils for which a generic association could not be determined due to lack of evidence. Robledo et al. (Reference Robledo, Sarzetti and Anzótegui2016) rejected such a change and amended the diagnoses of Stigmellites based on only mine morphologies, not references to trace makers. We prefer to use Stigmellites as a collective-group name in two reasons. First, such usage retains information on taxonomic affinities of trace fossils, often more important than their morphological features. Bertling et al. (Reference Bertling, Braddy, Bromley, Demathieu, Genise, Mikuláš, Nielsen, Nielsen, Rindsberg, Schlirf and Uchman2006, p. 279; also cited by Robledo et al., Reference Robledo, Sarzetti and Anzótegui2016), in fact, argued that producer-based ichnotaxon names are inappropriate, when the assignments are ambiguous. The ichnogenus Stigmellites includes trace fossils whose producers are reliably identifiable at the family level (Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015). Secondly, the amendment of Stigmellites by Robledo et al. (Reference Robledo, Sarzetti and Anzótegui2016) necessitates several nomenclatural changes, including designation of many singleton morphotaxa and a new collective-group name. We do not find such actions more advantageous than simply retaining Stigmellites as a collective-genus name. Stigmellites currently comprises one adult moth in amber and 16 species based on trace fossils, spanning the Late Cretaceous to the late Pliocene (Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015).
Stigmellites janggii Sohn and Nam, new ichnospecies
Figures 2.1, 2.3, 3
Figure 2. Photographs of leaf mines: (1) a whole leaf fossil of Fagus antipofi with leaf-mine trace (holotype of Stigmellites janggii n. ichnosp., DSHS-2017201); (2) a leaf mine of the extant nepticulid, Stigmella hemargyrella on Fagus sylvatica (photo by György Csóka); (3) detail of Figure 2.1, with white and black arrows indicating traces of an egg case and an exit slit, respectively. Scale bars = 10 mm (1, 2), 5 mm (3).
Figure 3. Drawing and photographs of Stigmellites janggii n. ichnosp.: (1) line drawing of the leaf mine in Figure 2.3, with gray area indicating a frass trail, and white and black arrows indicating traces of an egg case and an exit slit, respectively; (2) detail of a void corresponding to an egg case (white arrow in Fig. 3.1); (3) detail of a part of the frass trail corresponding to dashed box in Figure 3.1. Scale bar = 5 mm.
Holotype
DSHS-2017201 (only original part present).
Diagnosis
This ichnospecies differs from the late Pliocene Stigmellites pliotityrella Kernbach, Reference Kernbach1967 associated with Fagus sylvatica Linnaeus, Reference Linnaeus1753 in Germany in having a serpentine mine path (rather straight in S. pliotityrella) and a clear centric frass trail. Among the Miocene nepticulid fossils, S. janggi n. ichnosp. is similar to a nepticulid leaf mine on Quercus virginiana Miller, Reference Miller1768, found in the USA (Opler, Reference Opler1973), but differs from the latter in having a narrower and longer mine path.
Description
The mine trace (Fig. 2.3) commences near the distal one-fifth of the fifth 2° vein and proceeds between the third and the seventh 2° veins, roughly toward the midvein. There is a small, elliptical void possibly corresponding to a lost egg case. From the starting point, the mine runs toward the fourth 2° vein for a short distance, recurves, continues beyond the fifth 2° vein, recurves again near the sixth 2° vein, keeps running in slightly sinuous way, approaches the leaf margin at the fourth 2° vein, and then turns toward the third 2° vein. There, the mine runs along the third 2° vein toward the midvein, serpentines between the third and fourth 2° veins, and then runs along the fourth 2° vein toward the midvein for a moderate distance. After that, the mine obliquely proceeds beyond the fifth 2° vein, reaches the branching point of the midvein and the sixth 2° vein, and then runs along the basal two-fifths of the sixth 2° vein. Immediately after crossing the sixth 2° vein, the mine becomes an elliptical blotch parallel to the sixth 2° vein, enlarged toward the midvein. The frass trail starts in the beginning of the mine and ends before the blotch part of the mine. The frass patterns are present as a central line of randomly-distributed pellets, gradually broadening as the mine proceeds. The frass trail fills the width of the mine in the proximal third and takes up half of the mine width in the distal two-thirds. In the blotch part of the mine, the frass granules are irregularly aligned along the circumference. A semicircular slit is present near a branch of the midvein and the sixth 2° vein.
Etymology
This ichnospecies is named after the sediment where the fossil was discovered.
Classification
This mine can be classified as a full-depth, partially serpentine, linear-blotch leaf mine. In the morphotyping system by Labandeira et al. (Reference Labandeira, Wilf, Johnson and Marsh2007), Stigmellites janggii n. ichnosp. could fall into DT90 (DT = damage type) but differs from the latter in having a more sinuous mine path and a continuous frass trail.
Modern analogs
Among the extant species of Nepticulidae feeding on Fagus, the fossil leaf mine is similar to those (Fig. 2.2) made by a European species, Stigmella hemargyrella (Kollar, Reference Kollar1832), but differs from the latter in having a less-dispersed frass trail and in lacking the frass pattern in meniscate arcs. There is also an undescribed species of Stigmella that feeds on Fagus in Taiwan, and undescribed species of Stigmella and Ectoedemia Busck, Reference Busck1907 in Japan that feed on Fagus and produce similar leaf mines (unpublished data, C. Doorenweerd). The leaf-mine type is not very distinctive and various other modern analog species feeding on deciduous woody plants could be indicated.
Preservation
The whole block is a pale brown, subtriangular shaly mudstone, 17.8 cm in maximum length, 17 cm in maximum width, and ~ 5 cm in maximum thickness. The fossil leaf (Fig. 2.1) is a compression situated on a margin of the block with its apical quarter taken off. Parts of teeth and petiole have been preserved in the specimen. The face opposite the leaf mine bears three plant fragments: two monocotyledonous leaves and one dicotyledonous leaf.
Measurements
The total length of the linear mine is ~ 140 mm; the mine path is limited to a space of 24 mm x 21 mm; path lengths between turning points range 3–10 mm; the blotch-mine section is 9.5 mm in length and 3–5 mm in width; the leaf after reconstruction is ~ 850 mm in length and ~ 45 mm in maximum width.
Host plants
This fossil leaf can be assigned to Fagus (Fagaceae) based on a diagnostic combination of three characteristics of the genus (Leng, Reference Leng1999; Manchester and Dillhoff, Reference Manchester and Dillhoff2004; Xu et al., Reference Xu, Su, Zhang, Deng and Zhou2016): the absence of spines on the teeth, the lack of fimbrial veins, and the presence of fagoid teeth. It also shares the common features of Fagus, including short petioles, lamina with evenly spaced, parallel secondary veins and closely spaced percurrent tertiary veins, the lack of intersecondary and pectinal veins, and a serrate margin with nonglabular teeth arranged one per secondary vein (Manchester and Dillhoff, Reference Manchester and Dillhoff2004). Fagus antipofi Heer, Reference Heer1858 has been recorded from the Geumgwangdong Formation (Paik et al., Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012). The overall shape of the fossil leaf on our specimen is consistent with those of the small leaves of F. antipofi, illustrated by Tanai (Reference Tanai1961).
Remarks
Association of this fossil with Nepticulidae is substantiated by a combination of similarities with extant nepticulid leaf-mines: (1) a narrow linear mine path not crossing a primary vein; (2) the presence of traces possibly corresponding to an egg case and an exit slit; and (3) a frass trail filling the width of the mine in the early phase and not greatly broadening along the mine path. Based on the host plant and the shape of the mine, the mine maker could belong to the large extant genus Stigmella Schrank, Reference Schrank1802. However, linear-blotch mines are common among nepticulid genera and our knowledge of Fagus-feeding nepticulids is very limited. Although most nepticulid species are conservative to the plant family they feed on, there are exceptions and common plants are used by multiple genera (Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015). Thus, assignment of this fossil to the genus Stigmella is pending.
Discussion
The Miocene fossils of nepticulid leaf mines have been found in several sediments worldwide covering the entire epoch. The host plants for these fossils include Berberidaceae (Liebhold et al., Reference Liebhold, Voney and Schorn1982), Betulaceae (Kuroko, Reference Kuroko1987), Fagaceae (Opler, Reference Opler1973), Lauraceae (Peñalver, Reference Peñalver1997), Schisandraceae (Knor et al., Reference Knor, Prokop, Kvaček, Janovský and Wappler2012) and Symplocaceae (Martins-Neto, Reference Martins-Neto1989). In Fagaceae, Miocene leaf mines of nepticulids have previously been found exclusively associated with Quercus leaves (Opler, Reference Opler1973), whereas none were known on beeches (Fagus). The earliest and only nepticulid leaf-mine fossil on beeches is from the Pliocene (Kernbach, Reference Kernbach1967). Paik et al. (Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012) recorded a leaf-mine fossil on Fagus from the Geumgwangdong Formation (early Miocene), but did not identify the mine maker. Therefore, Stigmellites janggi n. ichnosp. is the first confirmed nepticulid leaf mine associated with the fossil leaves of Fagus in the Miocene.
Beech trees are one of the representative components of temperate deciduous broad-leaf forests of the Northern Hemisphere (Denk, Reference Denk2003). They generally occupy a climatic space spanning cool temperate and warm temperate zones (Fang and Lechowicz, Reference Fang and Lechowicz2006). Leaves of Fagus have been found in large numbers in the early Miocene Geumgwangdong Formation (Paik et al., Reference Paik, Kang, Kim, Lee, Kim and Jeong2010, Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012). The paleoflora of this formation has been compared to the Aniai-type flora in Japan, which indicates cool-temperate vegetation. Paik et al. (Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012) examined 83 fossil leaves of Fagus from the Geumgwangdong Formation and found that ~ 60% was damaged by insects. Among the insect damage types, external feeding was responsible for the greatest proportion, whereas mining was only responsible for 3% of the total damaged leaves.
This is the first study to identify the makers of the leaf-mine fossils from the Geumgwangdong Formation. Paik et al. (Reference Paik, Kim, Kim, Jeong, Kang, Lee and Uemura2012) illustrated two exemplar leaf-mine fossils, one on Fagus and the other on Ulmus Linnaeus, Reference Linnaeus1753. Both leaf mines appeared to have a nepticulid origin. The mine trace on the fossil Fagus leaf exhibited a serpentine path with a centric frass trail that enlarged as it proceeded. The mine path was limited to between the secondary veins, indicating that the maker was very small, as are nepticulids. The mine fossil on Ulmus exhibited frass patterns in meniscate arcs that are often observed in nepticulid mines. Stigmellites janggi n. ichnosp. is distinguished from those two leaf mines in the shapes of mine path and frass trail and in the presence of the terminal blotch phase, indicating a different species of Nepticulidae was involved. Herbivorous insect fauna from the Geumgwangdong Formation has not been thoroughly investigated to date. There was a jewel beetle (Buprestidae) recorded from the formation (Kim, Reference Kim1976), but no lepidopterans.
Extant Nepticulidae feeding on Fagus also appears to be little studied, with undescribed species known from Taiwan and Japan (unpublished data, C. Doorenweerd). There are two commonly observed species of Stigmella in Europe that feed on Fagus sylvatica: S. hemargyrella and S. tityrella (Stainton, Reference Stainton1854) (Ellis et al., Reference Ellis2018). The latter produces leaf mines similar to the Pliocene species, Stigmellitis pliotytirella, which was found in Germany (Kernbach, Reference Kernbach1967). Apart from Stigmella, only Ectoedemia species are currently found feeding on Fagus. Both Stigmella and Ectoedemia are diverse on oaks (Quercus), both in extant and fossil fauna (Doorenweerd et al., Reference Doorenweerd, van Nieukerken, Sohn and Labandeira2015, Reference Doorenweerd, van Nieukerken and Menken2016). It is noteworthy that leaf-mine examples on Fagus from the Geumgwangdong Formation indicate two nepticulid associations on the same fossil plant species. Because Fagus trees were a more dominant part of forests in Neogene Asia than in present day (Peters, Reference Peters1997), it is likely that there was a more diverse fauna of leaf miners associated with beeches than there is today.
Acknowledgments
We would like to thank Q. Wang (Chinese Academy of Sciences, Beijing, China) for his help in host-plant identification and G. Csóka (NARIC Forest Research Institute, Mátrafüred, Hungary) for allowing us to use his photos of leaf mines. This work was supported by the Korea Research Fellowship program funded by the Ministry of Science, ICT and Future Planning through the National Research Foundation of Korea (fund NRF-2015H1D3A1035581) and by ‘Research Under Protection,’ a subprogram of the Individual Research program in Basic Science and Engineering funded by the National Research Foundation of Korea (fund NRF-2017R1D1A2B05028793).
