Christelenkidae, a new extinct family based on a new taxon from Eocene Baltic amber (Diptera: Acalyptratae), with X-ray synchrotron microtomography imaging of its structures

A new family of Diptera Acalyptratae, Christelenkidae Roháček fam. nov. , is established for Christelenka multiplex Roháček gen. et sp. nov. , an unusual extinct taxon described from a unique male specimen preserved in Baltic amber (Mid-late Eocene, ca 48–34 Ma). Apart from detailed examination by light microscopy and photography, the holotype of the new species has also been studied by means of X-ray synchrotron microtomography with the aim of obtaining additional morphological data for consideration of its relationships. Because of a very peculiar combination of morphological characters, the new family is tentatively considered a separate lineage of Acalyptratae having no apparent sister-group relationship with any of the known families. Its probable relationships to some families of Opomyzoidea and Ephydroidea are discussed.


Introduction
Schizophora is a section of true flies (Diptera) with about 80 member families. This monophyletic lineage comprises a huge number of species (over 50, 000), is very diverse and relatively recent in the geological record. Within the Schizophora, two formal groups of families are recognized: the paraphyletic Acalyptratae and the monophyletic and younger Calyptratae (Wiegmann and Yeates 2017). The oldest reliable acalyptrate (and apparently also schizophoran) fossil records are known from Cambay amber from India (Early Eocene, ca 52 Ma, see Rust et al. 2010;Grimaldi and Singh 2012) and, particularly, from Baltic amber (also including Bitterfeld and Rovno amber) (Mid-late Eocene, 48-34 Ma, cf. Hennig 1965, 1966, 1967, 1969, 1971, 1972Tschirnhaus and Hoffeins 2009;Weitschat and Wichard 2010;Perkovsky et al. 2010). The oldest reliable record of Calyptratae, belonging to the family Anthomyiidae, also originates from Baltic amber (Michelsen 2000). It is peculiar that no representative of Acalyptratae has hitherto been found in older Tertiary amber from France: Oise (53 Ma, Nel and Brasero 2010), or in any of the late Cretaceous ambers (see e.g. McKellar and Wolfe 2010; Grimaldi and Nascimbene 2010). Given the unexpectedly large diversity of acalyptrate families in Baltic amber, which is comparable to that of recent times or even higher (Tschirnhaus and Hoffeins 2009), it may be assumed that there was an "explosive" radiation of Acalyptratae in the Mid to Late Eocene. Within this ~10-15 Myr long period, representatives of almost all currently known acalyptrate families appeared in the so-called Baltic amber forest. This paleohabitat was characterized by authors including Weitschat and Wichard (2002), Weitschat (2008), Kvaček (2010), Szwedo (2012), Słodkowska et al. (2013) and Sadowski et al. (2020). The rapid diversification of higher flies was most likely connected with development of luxuriant vegetation in the Baltic area during the Early Eocene Climatic Optimum. The vegetation in the Baltic area had a tropical character in the early Eocene warming maximum (ca 49 Ma) that was subsequently changing to subtropical or warm-temperate as a result of gradual cooling towards the end of the Eocene (Szwedo 2012;Słodkowska et al. 2013;Sadowski et al. 2020). Based on habitat and trophic demands of beetles (Coleoptera) recorded from Baltic amber, Alekseev and Alekseev (2016) characterized the Baltic amber forest as a thermophilic, humid-mixed forest similar to contemporary subtropical forests of eastern and southeastern Asia, but there were certainly also other forest types in the area formed in a warm-temperate climate (Sadowski et al. 2020).
This study is aimed at description of a very peculiar Eocene (Baltic amber) taxon of "opomyzoid" appearance distinguished by an unusual combination of adult morphological characters that prevents its clear association with any of the known families of Acalyptratae Diptera. Because some important structures of the ventral side of the postabdomen of this fly inclusion are obscured and, hence, invisible with an optical light microscope (cf. Fig. 1), X-ray synchrotron microtomography imaging techniques were used to reveal the morphology of some postabdominal sclerites and structures of the male terminalia.
Comparison of the resultant set of mophological characters of this new fossil taxon with those of other acalyptrate families confirmed that it cannot be affiliated with any of them, and, therefore, it is described as a new genus and species belonging to a new family (see below).

Material
A single amber sample from the collection of Ch. & H.W. Hoffeins (Hamburg, Germany) has been examined (Figs 1,2). It is now deposited in the Senckenberg Deutsches Entomologisches Institut, Müncheberg, Germany; SDEI.

Preparation of amber sample
The amber with the fly inclusion was prepared by H.W. Hoffeins following the methods described by Hoffeins (2001) and von Tschirnhaus and Hoffeins (2009). It was cut, ground and polished as close and as parallel as possible to the frontal, dorsal and lateral sides of the fly; subsequently the preparatum was embedded in artificial resin (also ground and polished) in order to facilitate its stereoscopic study.

Techniques of microscopic investigation, photography and measurements
The amber specimen was observed, drawn and measured by means of two types of binocular stereoscopic microscopes (Reichert, Olympus). Drawings of legs were prepared on squared paper using a Reichert binocular microscope with an ocular screen. The whole specimen and its parts were photographed using a Canon EOS 5D Mark III digital camera, a Nikon CFI Plan 4× /0.10NA 30 mm WD or Nikon CFI Plan 10x/0.25NA 10.5 mm WD objective attached to a Canon EF 70-200 mm f/4L USM telephoto zoom lens. During photography, the specimen was repositioned upwards between each exposure using a Cognisys StackShot Macro Rail and the final photograph was compiled from 35 layers using Helicon Focus Pro 7.0.2. The final images (including also those obtained from synchrotron microtomography, see below) were edited in Adobe Photoshop CS6. Some illustrations were drawn using the resultant macrophotographs in which details were inked based on direct observation at higher magnification using a binocular microscope. Measurements: Six characteristics were measured -body length (measured from anterior margin of head to end of cercus, thus excluding the antenna), wing length (from wing base to wing tip), wing width (maximum width), index Cs 3 : Cs 4 (= ratio of length of 3rd costal sector : length of 4th costal sector), index r-m\dm-cu : dm-cu (= ratio of length of section between r-m and dm-cu on cell dm : length of dm-cu).

Techniques of X-ray synchrotron microtomography imaging
The specimen was scanned with the Imaging Beamline P05 Wilde et al. 2016) operated by the Helmholtz-Zentrum Hereon at the PETRA III storage ring (Deutsches Elektronen Synchrotron -DESY, Hamburg, Germany), using a photon energy of 18 keV and a sample-to-detector distance of 100 mm. A series of projections was recorded with a custom developed 20 MP CMOS camera system with an effective pixel size of 1.28 µm (Lytaev et al. 2014). For each tomographic scan 3601 projections at equal intervals between 0 and π were recorded. Tomographic reconstruction was conducted by applying a transport of intensity phase retrieval approach and using the filtered back projection algorithm (FBP) carried out in a custom reconstruction pipeline using Matlab (Math-Works) and the Astra Toolbox (Moosmann et al. 2014;van Aarle et al. 2015van Aarle et al. , 2016. Raw projections were binned twice for further processing, resulting in an effective pixel size of the reconstructed volume (voxel) of 2.56 µm. We have conducted a reconstruction of the scanned volumes with Drishti ver. 2.6.6 (Limaye 2012). To downscale the strain on the RAM and video card of the computer used, we have scaled down all the tiff images to 50% using the Fiji "scale" function (Schindelin et al. 2012). Downscaled stacks were then rendered into 3D volume in Drishti ver. 2.6.6 (Limaye 2012).

Morphological terminology
Terminology of morphological characters follows Ro háček (2013) and Roháček and Hoffeins (2021) to be in continuation with recent studies of the senior author on fossil Anthomyzidae and Clusiomitidae. Terms of structures for the male terminalia are largely based on the "hinge" hypothesis for the origin of the eremoneuran hypopygium, re-discovered and documented by Zatwarnicki (1996) and, therefore, the terms derived from other hypotheses are listed below as synonyms to avoid any confusion.
Diagnosis. Same as above for Christelenkidae fam. nov.
Etymology. The name of the genus is an abbreviated compound of the first names of two ladies, viz. Christe[l] + Lenka, playing important roles in the scientific career of the first author. It is dedicated to Christel Hoffeins  (Hamburg, Germany) who discovered this amazing fly in Baltic amber, and to Lenka Roháčková, the wife of the first author, for her lifelong support and patience with his research in dipterology. Roháček, sp. nov.

Figs 1-35
Description. Male. Total body length 3.1 mm; general colour brown to dark brown, with some parts of head and abdomen lighter coloured (see below); thorax and abdominal terga probably subshiny to partly shiny (Figs 1-4). -Head: (Figs 5-10) about 1.3 times as high as long, dorsally very slightly wider than thorax. Dorsal part of occiput distinctly concave (Fig. 7). Head distinctly bicolourous, dorsally and posteriorly largely dark brown, anteriorly and ventrally orange ochreous to whitish yellow. Frons (Figs 6, 9) moderately broad, hardly tapered anteriorly, largely bare, dark brown behind apex of ocellar triangle and ors, reddish brown to orange ochreous in middle part and yellow to whitish yellow on crescent-shaped anterior margin surrounding bases of antennae. Orbital plate hardly delimited, slightly visible only at base of ors, dark brown behind latter, reddish brown to orange ochreous anteriorly, up to level of ptilinal fissure. Frontal triangle not developed. Ocellar triangle blackish brown, small, somewhat elongate, slightly protruded among ocelli. Ocelli small (Fig. 9). Lunule absent (or concealed within dorsal medial part of face). Face (prae-frons) slightly concave, yellow to yellowish white. Parafacialia and gena yellowish white, obviously whitish microtomentose; gena narrowly yellow-margined ventrally.
Comparative remarks. As remarked above, Christelenka multiplex sp. nov. at first glance looks like an opomyzoid fly, in external appearance most resembling some species of Opomyzidae, Anthomyzidae, Stenomicridae or Aulacigastridae (cf. images in Marshall 2012). However, it differs from any member of these families by a combination of its most diagnostic characters (see above under Christelenkidae) and can be immediately recognized by the bare forefrons, absence of the frontal triangle and lunule, a single posteriorly positioned ors, the extremely dorsobasal arista, the wing pointed at apex of R 4+5 , a strong seta on fore coxa, peculiar chaetotaxy of the fore (with dorsal seta and an anteroventral row of small spines) and hind (with a dorsal and a longer anteroventral  Roháček det. 2020' (red label). The specimen is embedded in polyester resin, size of preparatum 10.4 × 6.6 × 5.9 mm, size of cut amber 7.5 × 4.5 × 3 mm (Fig. 1)

Affiliation of the new family
The placement of this new family in the current classification system of Schizophora proved to be rather difficult although Christelenkidae undoubtedly belongs to this group. As was found for the extinct families Hoffeinsmyiidae (Michelsen 2009) and Yantaromyiidae (Barták 2019), Christelenkidae also seems not to have an obvious sister-group family among the Diptera Acalyptratae.
The situation is also complicated by the fact that there is no accepted modern classification of families of Acalyptratae. The only comprehensive system of Acalyptratae Diptera (based on a phylogenetic hypothesis resulting from a manual analysis of morphological characters) remains that by McAlpine (1989). Although the most recent molecular and phylogenomic studies indicate that some groups (superfamilies and suprafamilies) recognized by McAlpine (1989) may not be monophyletic (see Winkler et al. 2010;Wiegmann et al. 2011;Bayless et al. 2021), grouping of the acalyptrate families in these studies is not consistent. Also, the attempt at producing a morphological hypothesis of the phylogeny of Diptera families by Lambkin et al. (2013) did not resolve the relationships within Acalyptratae sufficiently to support a new classification. However, the monophyly of several (more or less distinct) superfamilies have been confirmed recently, re-classified and (sometimes) also re-diagnosed by morphological characters. Nerioidea and Diopsoidea belong to these groups, having their monophyly supported and taxonomic limits clarified by Lonsdale (2020) on the basis of a thorough phylogenetic morphological analysis. Also, Tephritoidea are (long recognized, see McAlpine (1989)) recognized as a monophyletic group, which was confirmed morphologically; e.g., by Korneyev (1999) and molecularly by Han & Ro (2016). In addition, the monophyly and taxonomic limits of the Sciomyzoidea have been long recognized (see Tóthová et al. 2013), but the recent inclusion of Chamaemyiidae, Lauxaniidae and even Conopidae on the basis of phylogenomic data (Bayless et al. 2021) has put into question the morphological delimitation of this superfamily. Thus, Conopoidea and Lauxanioidea of McAlpine (1989) have now been included in the expanded concept of Sciomyzoidea by Bayless et al. (2021). Similarly, the monophyly of Carnoidea, analysed morphologically and re-defined by Buck (2006), has recently also been rejected by Bayless et al. (2021). Sphaeroceroidea (sensu McAlpine 1989) was also recognized as monophyletic (Bayless et al. 2021)  studies (Bayless et al. 2021;Winkler et al. 2022), the phylogenetic hypotheses proposed in these two studies differ significantly in recognition of the sister-group of Ephydroidea. From the above review, it is apparent that also among the superfamilies long considered monophyletic, the placement and relationships of certain families remain unstable and often recognized differently in hypotheses based on morphological and molecular data. While the classification and relationships of families affiliated by McAlpine (1989) in Sphaeroceroidea (= Heleomyzoidea) have been (partly) supported by Bayless et al. (2021), those of families formerly classified in Opomyzoidea, seem to be totally confusing when comparing results of the recent morphological (Lambkin et al. 2013;Lonsdale 2020) and molecular (Winkler et al. 2010;Wiegmann et al. 2011;Bayless et al. 2021) phylogenetic hypotheses.
The only consensus of all these studies is that this superfamily is not monophyletic. We assume that the precipitous radiation of acalyptrate families during a relatively short period in the Mid-late Eocene (see also Lonsdale 2020: 4, for more detail) could be the major reason for the difficulties in clarifying their phylogenetic relationships (both using morphological and molecular methods) and, subsequently, their unresolved systematic classification.
Comparing diagnostic features of Christelenkidae with sets of apomorphic characters defining superfamilies of Acalyptratae according to McAlpine (1989) with refinements added by more recent morphological studies (e.g. Korneyev 1999;Buck 2006;Tóthová et al. 2013;Lonsdale 2013Lonsdale , 2020Lonsdale et al. 2010), we have reached the conclusion that this new fossil family is most probably allied with some families of the (polyphyletic) Opomyzoidea or the (monophyletic) Ephydroidea.
The comparison of Christelenkidae with Opomy zoidea (as delimited by McAlpine 1989) is more problematic because it is a very heterogeneous group, obviously polyphyletic in origin (first dismantled by Winkler et al. 2010), see above. Some of the characters used by McAlpine (1989) as synapomorphies to demonstrate monophyly of Opomyzoidea are obviously erroneously polarised or incorrectly selected. Therefore, the above set of putative apomorphies of Christelenkidae is compared rather with some families (recognized as most similar to Christelenka) placed historically in Opomyzoidea; of course, without those already excluded from this superfamily, such as Acartophthalmidae (now in Carnoidea, see Buck 2006) and Fergusoninidae (now in Nerioidea, see Lonsdale 2020). Also, several other families of Opomyzoidea markedly dissimilar to Christelenkidae are excluded from the considerations below. These include Clusiidae (recognized as the most basal lineage of Acalyptratae in a hypothesis by Lonsdale (2020)), the fossil Clusiomitidae (see Roháček & Hoffeins 2021), Agromyzidae (see Lonsdale 2021a) and Odiniidae. The latter two families were recognized as separate lineages distant from the Opomyzidae-Anthomyzidae pair which Lonsdale (2020) supported as the only families remaining in Opomyzoidea. However, it must be noted that also the relationship of this pair was doubted practically in all molecular studies treating both these families (Winkler at al. 2010;Wiegmann et al. 2011;Bayless et al. 2021). Further distinctly dissimilar and distantly related families hitherto placed tentatively in Opomyzoidea are Marginidae, Neminidae, Teratomyzidae, Asteiidae and Xenasteiidae. All these groups share with Christelenka only a few of the apomorphies listed above (and all these are widely homoplasious) but are characterized by a number of apomorphies specific to them but lacking in Christelenkidae.
The family Neurochaetidae seems to share with Christelenkidae these apomorphies: 1, 2, 9, 10 (1 or 2 dc), 12, 14, 15 (1 or 2 stpl), 18, 27 (f 3 with only dorsal seta), 28. However, Neurochaetidae are distinguished by the peculiarly modified antenna with enlarged cap-like pedicel encompassing base of small 1st flagellomere, discrete orbital plate, enlarged frontal triangle, rich cephalic chaetotaxy (with proclinate anterior ors as in Aulacigastridae and strong setosity of anterior portion of gena most resembling that in Periscelididae), highly modified dorsoventrally flattened thorax, prosternum narrowed (linear), subscutellum reduced, wing with cells bm, cup and also alula atrophied, postabdomen asymmetrical, epandrium flat and band-like, and distiphallus of aedeagus very long and at least partially coiled (McAlpine 1988;Lonsdale 2021b), thus markedly different from those of Christelenka. On the other hand, the broad sterna and reduced pleural membrane of preabdominal segments (treated as an autapomorphy of Neurochaetidae by McAlpine (1989) but not by Lonsdale (2020)) are reminiscent of those of Christelenka but preabdomen of the latter is not flattened, not to mention its peculiarly prolonged 6th segment.
The family Periscelididae is treated here in the narrowed concept (= Periscelidinae of Rung & Mathis 2021b), thus excluding genera affiliated to the family Stenomicridae (for review see Roháček 2011). Perisce lididae seems to share more apomorphies with Christelenkidae than Stenomicridae and, therefore, only this family is compared with Christelenka here. The shared apomorphies are as follows: 1, 2, 5, 7, 11, 12, 14 (but often with setulae at the posterior margin of mesopleuron), 17, 31. Although the secondarily symmetrical postabdomen can be seen in both groups, that of Periscelididae differs by a large T6 and large, entirely symmetrical dorsal pregenital synsclerite probably formed by fusion of T7, S7 and S8 (Roháček & Andrade 2017). Periscelididae is also distinguished in having pedicel with distinct dorsal seam (as in Ephydroidea), arista pectinate, oc arising outside ocellar triangle, gena anteriorly strongly setose (as in Neurochaetidae), C without breaks and ending at apex of R 4+5 , and, most significantly, by uniquely formed male genitalia having reduced to absent gonostylus, ventrally positioned elongate cerci and, particularly, the extremely expanded (larger than entire external genitalia) pocketshaped phall apodeme (see Roháček & Andrade 2017). Although the male genitalia are not precisely visible in Christelenka, it is clear that they are wholly different from those of Periscelididae. Species of the family Aulacigastridae resemble externally Christelenkidae but the shared apomorphies are only a few: 5,8,9,11,12,18,21 (sometimes forming a kink on R 1 ), 22, 31. As above, almost all these apomorphies are widely homoplasious in Acalyptratae. We would like to remark on character 8 (anterior half of frons bare). The frons of Aulacigastridae is bare (without setulae) anteriorly but the strong proclinate-inclinate (anterior) ors usually arises in anterior half of frons. This distinctive ors seta, the strongly reduced or absent oc, the absent pvt and true vibrissa are the most distinct differences in cephalic chaetotaxy against Christelenkidae. The symmetrical postabdomen (character 31) is shared by both families but in Aulacigastridae there is a large pregenital sclerite (fusion of T6 and S6 according to Rung & Mathis (2021a: fig. 88:6), but obviously also integrating S7 and S8, cf. Rung et al. (2005)) forming almost complete (ventrally shortened, asymmetrical and narrowly open) ring. This synsclerite can be considered an autapomorphy of Aulacigastridae. The external male genitalia are also very distinctive in Aulacigastridae because a true gonostylus is lacking and the epandrium is provided with a rigid posteroventral process on each side (erroneously treated as surstylus by authors, see Rung & Mathis (2021a)), thus they are quite dissimilar to those of Christelenka.

Most significant apomorphies of
Christelenkidae (same numbering as in above list) (3) Arista extremely dorsobasal. The ancestral condition is surely an apical arista on a porrect antenna (as known in Phoroidea). In Schizophora, the dorsal or dorsobasal position of the arista should be considered a synapomorphy of this group and the (uncommonly occurring) apical arista in some of its families as a reversal. Therefore, we believe that its shifting extremely basally, just at the distal margin of the pedicel (Figs 8,9) in Christelenka, is secondary and hence an apomorphic condition, unusual in Acalyptratae.
(4) Frontal lunule absent (lost) (Figs 6, 9). The lunule is a synapomorphy of the Schizophora (McAlpine (1989: 1423) but its reduction occurs commonly among families of Acalyptratae. However, in these cases some remnant of the lunule (often depressed and sunken into a small medial fissure) are preserved. The condition found in Christelenka is considered unusual because the lunule disappeared entirely and only a ptilinal suture can be seen above bases of antennae.
(13) Laterobasal sc longer than the crossed apical sc (Fig.  16). Apical sc are often crossed but normally longer than laterobasal sc. The above combination is rare in Acalyptratae (known in some Tephritidae and in Diastatidae: Diastata species, where apical sc are, however, upright) and are unknown in families historically placed in Opomyzoidea.
(19) Wing with apex somewhat pointed at end of R 4+5 (Figs 18, 19). Apically pointed wings are known mainly in some Ephydroidea (Camillidae, Diastatidae: Campichoeta) but also in some species of other groups of Acalyptratae, viz. Fergusoninidae (see Lonsdale 2020: figs 192, 411), Teratomyzidae (cf. Papp 2011; Rodrigues et al. 2016) and also in Typhamyza bifasciata (Wood) from the family Anthomyzidae (see Roháček 1992Roháček , 2006. It can perhaps be considered a putative apomorphy indicating relationships of Christelenkidae to Ephydroidea. (24) cx 1 elongate and with 1 distinct ventral seta in middle (Figs 3, 15). Elongated cx 1 is not common in Acalyptratae but the presence of a single strong medial ventral seta is quite unusual and probably unique character.
(25, 26) f 1 with an anteroventral ctenidium-like series of small spines and 1 dorsal seta (Figs 20,21). This chaetotaxy of the fore femur seems to be unique. The row of small anteroventral spines seems to be homologous with a series of dense, thickened setae or spines called a ctenidium e.g. in Lauxaniidae (Homoneura Wulp), Diastatidae, Curtonotidae, Sepsidae (Nemopoda Robineau-Desvoidy). It was also found in Marshallya platythorax Roháček (Roháček 2018: fig. 27) and in species of Barbarista Roháček belonging to Anthomyzidae (Roháček 2021: fig. 16). These ctenidium-like spines or spiniform setae are anteroventral, and therefore, not homologous with the posteroventral "ctenidial spine" in Anthomyzidae. In all the above cases, however, the dorsal seta is absent. The presence of the dorsal seta on f 1 is considered an unusual and distinct apomorphy of Christelenka.
(27) f 3 with 1 anteroventral and 1 distinct dorsal seta (Fig.  23). This combination is also considered unusual in Acalyptratae. Both these setae can be seen in some Pallopteridae, with addition to 1 or 2 anterodorsal setae, although all are situated more distally (McAlpine 1987) than those in Christelenka. The presence of a strong anteroventral seta or setae on f 3 is more common, similarly as is the occurrence of anterodorsal setae, or both in combination. A distinct truly dorsal seta of f 3 seems to be rare, known in some representatives of Neurochaetidae (Lonsdale 2021).
(30) T5 and S5 enlarged and very elongate (Figs 24-28). This character has hitherto been unknown in the majori-ty of families of Acalyptratae. The only exception is the markedly elongated 5th abdominal segment in males of the genus Protearomyia McAlpine (see MacGowan & Rotheray 2008;MacGowan 2014) belonging to the unrelated Lonchaeidae (superfamily Tephritoidea).
(32) Male S7+S8 completely fused, with a pair of strong dorsal setae. Surprisingly, a similar pair of robust setae at the posterior margin of S8 is (only) known in Cypselosomatidae and Pseudopomyzidae (see Lonsdale 2020: figs 216, 217) belonging to Nerioidea.

Conclusion
The peculiar mixture of characters apomorphic for the Christelenkidae and, particularly, those considered almost unique or rarely occurring in Acalyptratae, indicate that although this new group seems to have an affinity with some groups of the Ephydroidea (Diastatidae in particular) and Opomyzoidea (sensu McAlpine 1989) (mainly Opomyzidae and Anthomyzidae), it cannot be related to any of them. This is also demonstrated by the fact that the apomorphic characters shared with these taxa are relatively few in number and, moreover, known to occur as scattered homoplasies not only among other families of Opomyzoidea but also in evidently unrelated groups of Acalyptratae. Therefore, it is suggested that the Christelenkidae could be tentatively considered a separate lineage of Acalyptratae possibly related to Opomyzoidea sensu Lonsdale (2020) (= Opomyzidae + Anthomyzidae) or to Ephydroidea, having no apparent sister-group relationship with any of the currently recognized families of these groups.

Competing interests
The authors have declared that no competing interests exist.

Acknowledgements
First of all, we would like to thank Mrs Christel Hoffeins (Hamburg, Germany) for the discovery of this peculiar fossil taxon in Baltic amber, for making it available for study and for valuable comments on earlier drafts of the manuscript. We are grateful to Mr. Marius Veta (Palanga, Lithuania), owner of the amber company "Ambertreasure4u" from whom the specimen was purchased, for his continuing support with Baltic amber Diptera and especially acalyptrates. Our sincere gratitude is also expressed to Miroslav Barták (Praha, Czech Republic)