A new enigmatic lineage of Dascillidae (Coleoptera: Elateriformia) from Eocene Baltic amber described using X-ray microtomography, with notes on Karumiinae morphology and classification

Dascillidae are a species-poor beetle group with a scarce fossil record. Here, we describe Baltodascillus serraticornis gen. et sp. nov. based on a well-preserved specimen from Eocene Baltic amber. It differs from all known Dascillidae by its reduced mandibles. After studying the specimen using light microscopy and X-ray microtomography, we tentatively place this genus in the poorly defined subfamily Karumiinae based on the large eyes, serrate antennae, and lack of prosternal process. This is the first representative of the Dascillidae formally described from Baltic amber and the first described fossil member of the subfamily Karumiinae. We briefly discuss the problematic higher classification of Dascillidae, along with the morphology and biogeography of the group.


Introduction
Dascillidae are a small beetle family with about 100 extant species classified in 11 genera distributed in all zoogeographic realms (Ivie and Barclay, 2011;Jin et al., 2013b;Lawrence, 2016;Johnston and Gimmel, 2020). Together with the still-smaller family Rhipiceridae, they form the superfamily Dascilloidea within the series Elateriformia (Lawrence, 2016;Kundrata et al., 2017). Dascillidae are further divided into two subfamilies with unclear limits, i.e., Dascillinae (six genera, about 80 species) and Karumiinae (five genera, about 20 species) (Lawrence and Newton, 1995;Ivie and Barclay, 2011;Lawrence, 2016). There has been no worldwide taxonomic revision of Dascillidae. However, the subfamily Dascillinae has received particular attention during the last decade. The worldwide genera and Asian species of Dascillinae were revised by Jin et al. (2013b), with additional species described in several subsequent papers (Jin et al., 2015(Jin et al., , 2016Li et al., 2017;Fang et al., 2020;Wang et al., 2020a, b). The only Australian genus of the family, Notodascillus Carter, 1935, was revised by Jin et al. (2013a), and Terzani et al. (2017) revised the Western Palearctic species of Dascillus Latreille, 1797. Most recently, Johnston and Gimmel (2020) reviewed the North American Dascillidae of both subfamilies and also provided a species checklist for the New World. Although Karumiinae were included in the latter study, issues of higher classification did not receive much attention. Ivie and Barclay (2011) revised the status of certain genera associated with Karumia Escalera, 1913, but a comprehensive revision of Karumiinae in the modern sense is missing.
The fossil record of Dascillidae is rather scarce and was critically reviewed by Jin et al. (2013c). Two Australian Upper Triassic genera, Apheloodes Dunstan, 1923 andLeioodes Dunstan, 1923, described based on a single elytron each, were removed from Dascillidae and placed in Coleoptera incertae sedis, and the North American Miocene Protacnaeus Wickham, 1914 andMiocyphon Wickham, 1914 were transferred to Psephenidae and Scirtidae, respectively. The monotypic genus Mesodascilla Martynov, 1926 from the Jurassic Karatau locality in Kazakhstan, which was originally assigned to Dascillidae (Martynov, 1926), has been re-cently treated either as a member of Eulichadidae (Kirejtshuk and Azar, 2013) or Lasiosynidae (Yan et al., 2014). Jin et al. (2013c) kept only three genera which contain fossil species in Dascillidae. The genus Lyprodascillus Zhang, 1989, with two described species from the Miocene of China, has remained tentatively in Dascillidae, although authors could not confirm its systematic placement based on the available descriptions and illustrations (Jin et al., 2013c). The only Mesozoic dascillid genus is the monotypic Cretodascillus Jin, Slipinski, Pang & Ren, 2013 from the Cretaceous Yixian Formation in China (Jin et al., 2013c). Most fossil species are classified in the extant genus Dascillus, which currently contains six species from the Miocene Shandong Formation of China (Zhang, 1989;Zhang et al., 1994), one from the Miocene Latah Formation of the USA (Lewis, 1973), and one from the Eocene Florissant Formation of the USA (Wickham, 1911). However, at least some of these species probably do not belong to Dascillidae and are kept there only tentatively until the type material is examined in detail (Jin et al., 2013c). Some fossil taxa attributed to Dascillidae, and usually to Dascillus (sometimes as its synonym Atopa Paykull, 1799), remain unnamed, as for example the specimens found in the Oligocene Aix-en-Provence in southern France (Hope, 1847) or specimens reported from Baltic amber (Helm, 1886(Helm, , 1896Klebs, 1910;Larsson, 1978;Spahr, 1981).
In this study, we describe an enigmatic new genus based on a well-preserved specimen from Eocene Baltic amber. This is the first described representative of Dascillidae from amber. We discuss its systematic placement as well as the higher classification and biogeography of Dascillidae.

Material and methods
The amber piece was polished by hand, allowing improved views of the included specimen, and was not subjected to any additional treatment. For the purpose of light microscopic image capture, the amber specimen was fixed at a suitable angle of view to a Petri dish with gray plasticine modeling clay (Pelikan, Germany, model number 601492). It was photographed submersed in glycerol to prevent reflections and to reduce visibility of small scratches on the surface of the amber piece. Images were taken with a Leica MC 190 HD camera attached to a motorized Leica M205 C stereo microscope equipped with the flexible dome Leica LED5000 HDI or the conventional ring light Leica LED5000 RL-80/40 as an illuminator, applying the software Leica Application Suite X (version 3.7.2.22383, Leica Microsystems, Switzerland). Stacks of photographs were combined with the software Helicon Focus Pro (version 7.6.4, Kharkiv, Ukraine), applying the rendering method "depth map" or "weighted average".
The X-ray micro-CT (µCT) observations were conducted at Daugavpils University, Daugavpils, Latvia (DU), using a Zeiss Xradia 510 Versa system. In order to achieve the best results possible, three scans were conducted: habitus, head/prothorax and abdomen. For prothorax/head and abdomen scans, the parameters were mostly identical with the exception of exposure time. For both scans, sample-detector and source-sample distances were set to 37.4 and 38 mm, respectively, the source X-ray beam energy was set to 30 kV and power of 2 W. Tomographic slices were generated from 2401 slices through 360-degree rotation, using 4× objective. Achieved voxel size during scan was 3.3 µm. Exposure times were set to 8 s for the abdomen scan and 9 s for the head/prothorax scan. The overall scan was carried out at Xray beam energy of 40 kV and power of 3 W, source-sample distance was set to 24.4 mm and sample-detector distance was set to 117.7 mm. Tomographic slices were generated from 1601 slices through 360-degree rotation using a 0.4× objective and exposure set to 17 s. Achieved voxel size during overall scan was 11.7 µm. All three scans had binning set to 2 times as well as variable exposure time set to 2 at the thickest part of the sample. Prior to each scan a warmup scan was conducted that lasted 25 min. Acquired images were imported into Dragonfly PRO (version 2020.2) software platform for interactive segmentation and 3D visualization. Final image plates were assembled using Adobe Photoshop CC (version 2019-20.0.5).
Body length of the examined specimen was measured from the clypeus to apex of elytra, body width at the widest part of the body, pronotal length at midline, and pronotal width at the widest part. Morphological terminology follows Jin et al. (2013b) and Johnston and Gimmel (2020). The holotype is deposited in the collection of the Department of Palaeontology of the National Museum, Prague, Czech Republic (NMPC). The ZooBank LSID number for this publication is urn:lsid:zoobank.org:pub:F55CA75B-AF7C-4F1F-BD65-DCBEAAC75410 (12 March 2021).

Etymology
Derived from the words "Baltic" (referring to Baltic amber) and "Dascillus" (a genus name in Dascillidae). Gender: masculine.

Diagnosis
Baltodascillus gen. nov. can be recognized among other genera of Dascillidae by the reduced mandibles (Figs. 2a,3d). Additionally, the following combination of characters serves to distinguish it from all other genera: strongly serrate antennae, a fusiform terminal maxillary palpomere, large eyes, pronotum widest posteriorly and with a strongly developed lateral carina, lack of a developed prosternal process, complete elytra, confused elytral punctation, a weakly developed elytral epipleuron, and abdomen with five ventrites (Figs. 1-4).

Composition and distribution
Baltodascillus gen. nov. is a monotypic genus and is known exclusively from Eocene Baltic amber.

Etymology
The specific epithet "serraticornis" is a Latin adjective referring to the shape of the antennae.

Diagnosis
As for the genus (vide supra).

Description
Adult male. Body (Fig. 1) about 7.5 mm long and 2.5 mm wide, narrowly elongate, about 3 times as long as wide, weakly convex; dorsally moderately densely setose.

Remark
We conclude that the examined specimen is a male based on the shape of the body, serrate antennae, a long, narrowly rounded abdominal ventrite 5, and aedeagal structure (parameres) detected in X-ray micro-CT imaging.

Discussion
In order to understand the placement of the newly discovered fossil genus within Dascillidae, it is important to acknowledge the currently problematic situation with the subfamilial classification of the group. As currently delimited, Dascillidae contain two vaguely defined subfamilies with a divergent taxonomic history (Lawrence and Newton, 1995;Lawrence, 2016). Free-living and non-modified Dascillinae occur in non-arid areas of western North America, the Greater Antilles, Palearctic and Oriental regions, western Africa, and Australia. Karumiinae, which contain morphologically modified lineages of which at least some are associated with subterranean termites, are mostly distributed in arid and semiarid regions of western North America, northern Africa, central Asia and southern South America (Lawrence and Newton, 1995;Jin et al., 2013b;Lawrence, 2016). The latter group had long been treated as a separate family, Karumiidae, associated with the families of former Cantharoidea (i.e., roughly the soft-bodied Elateroidea) and originally included only the morphologically highly modified groups near Karumia Escalera, 1913 with an apparently soft cuticle, variously reduced elytra and unknown females (Crowson, 1955;Arnett, 1964;Paulus, 1972). Crowson (1971) first associated Karumiidae with Dascillidae based on a number of morphological characters, although he still kept the family status for both groups. He also transferred the genera Anorus LeConte, 1859, Pleolobus Philippi and Philippi, 1864 (Fig. 5), Genecerus Walker, 1871 and Emmita Escalera, 1914 from Dascillidae to Karumiidae and provided updated diagnoses for both groups. Based on his key, the Karumiidae contained taxa with a simple galea, the ligula with two short lobes, the ventral tarsal lobes absent or not basally articulated, the corpotentorium very broad, the male antenna pectinate or flabellate, and females wingless (apparently referring to known females of Anorus sp. with shortened elytra and reduced hind wings). However, Crowson's characters for separating the two subfamilies have been largely problematic since they do not apply to all included genera, and especially certain intermediate groups like Genecerus violate the diagnosis of Karumiinae (Lawrence and Newton, 1995;Lawrence, 2016). For example, both subfamilies share a very broad corpotentorium, the ventral lobes on tarsomeres of Genecerus and Anorus are similar to those in Das- cillinae, at least females of Genecerus have well-developed hind wings (Lawrence, 2016), and serrate to pectinate antennae are known only in Emmita and Genecerus but not in any other karumiine genus. Additionally, Grebennikov and Scholtz (2003) concluded that the larval morphology of Pleolobus is highly similar to that of Dascillinae, with the primary difference being that the latter possess molar ridges on the mandible. Unfortunately, the relationships among the dascillid genera have never been tested using a focused molecular approach. The only available molecular phylogenetic analyses, although either preliminary or focused on broader issues, have nonetheless called into question the current subfamilial classification (Kundrata et al., 2017;Johnston and Gimmel, 2020).
Representatives of Dascillidae display a graded series of morphological modifications connected with possible neoteny and soft-bodiedness, similar to the situation in various Elateroidea (Cicero, 1988;Kundrata and Bocak, 2019). In Dascillidae, the modifications represent a continuum from the well-sclerotized groups with adults of both sexes fully developed (although at least some females might have slightly less developed hind wings than their counterparts) (traditional Dascillinae), through more or less soft-bodied groups with males with complete elytra and females with variously shortened elytra and reduced hind wings (Anorus spp., Pleolobus, probably also Emmita) (Fig. 5) to the soft-bodied forms with males with greatly shortened elytra and females unknown but probably even more dramatically modified, termitiform (most Karumia) (Paulus, 1972;Solervicens, 1991;Ivie and Barclay, 2011;Johnston and Gimmel, 2020). Such modifications also affect the morphology of the thorax, such that the hard-bodied groups have a well-developed, spinelike prosternal process, whereas the soft-bodied forms have the prosternal process dramatically reduced to form a short triangle-shaped denticle. It is probable that the modifications of morphology connected with the independent evolution of soft-bodiedness have influenced the formal classification of Dascillidae, similar to many cases in the Elateroidea (Kundrata and Bocak, 2019). This hypothesis, however, needs to be tested using a well-sampled molecular phylogenetic analysis.
Under the current state of knowledge, we tentatively place the newly described Baltodascillus gen. nov. in Karumiinae based on the large eyes, serrate antennae, and lack of a spine-like prosternal process. This fossil genus is morphologically similar to the New World genera Anorus and Pleolobus (Fig. 5) or the African/Middle Eastern genus Genecerus but differs from these in the strongly serrate (but not pectinate) antennae, crenulate hind margin of the pronotum, and confused elytral punctation. It differs from all known Dascillidae by the highly reduced mandibles which do not overlap and are apparently non-functional. Given the unusual nature of this character, further specimens of this genus will be critical in determining whether this is a stable character or the result of a single teratological specimen. This is the first representative of the Dascillidae formally described from Baltic amber and also the first described fossil member of the subfamily Karumiinae since Semenoviola obliquotruncata Martynov, 1925, originally described in that group, was determined to be a dermapteran (Bolívar y Pieltain, 1926). Also, Baltodascillus gen. nov. is the first karumiine known from Europe and, therefore, sheds light not only on the paleodi- versity and systematics of the group but also on its biogeography.
Data availability. The holotype of Baltodascillus serraticornis gen. et sp. nov. is deposited in the collection of the Department of Palaeontology of the NMPC. Volume renderings of X-ray microtomography of habitus, head/prothorax, hind tarsus, and abdomen are available as Video supplements 1-4, respectively.
Author contributions. RK conceived and designed the study. MLG and RK carried out the morphological investigation. AB conducted the micro-CT scanning. SMB prepared microphotographs. GP and RK prepared figure plates. RK and MLG wrote the initial manuscript with help of GP, AB, and SMB. All authors performed the literature search, discussed the results, and edited, reviewed, and approved the manuscript.
Review statement. This paper was edited by Florian Witzmann and reviewed by Shuhei Yamamoto and one anonymous referee.