†Cretolixon – a remarkable new genus of rhopalosomatid wasps (Hymenoptera: Vespoidea: Rhopalosomatidae) from chemically tested, mid-Cretaceous Burmese (Kachin) amber supports the monophyly of Rhopalosomatinae

Rhopalosomatidae, currently considered the sister group of the Vespidae, are an enigmatic family of aculeate wasps that originated in the Late Jurassic or Early Cretaceous. Despite their considerable age, very few fossils of the family have been reported – all of them in amber (Miocene Dominican, Miocene Mexican, and mid-Cretaceous Burmese ambers). Here we report a new mid-Cretaceous rhopalosomatid wasp, Cretolixon alatum Lohrmann, gen. et sp. nov., from Burmese (Kachin) amber. This new genus has a unique mixture of characters, some of which are only known from the recent brachypterous genus Olixon and others of which are known only from the recent macropterous genera. Thus, Cretolixon Lohrmann, gen. nov. not only provides further evidence for the monophyly of the family but also contributes evidence for the monophyly of the Rhopalosomatinae. Key characters of the family are discussed, and an updated checklist of the world genera and fossil species and occurrences of Rhopalosomatidae is provided. Additionally, a chemical analysis was performed for three of the newly reported fossils as well as for the amber piece containing the rhopalosomatid larva described by Lohrmann and Engel (2017) to ascertain their amber vs. copal nature and their affinities with each other and previously described Burmese amber.

Within the family, two distinct morphological forms exist, which prompted Engel (2008) to split the family into two subfamilies -the Olixoninae Engel and Rhopalosomatinae Ashmead. The Olixoninae are solely represented by the highly apomorphic, brachypterous genus Olixon Cameron (1887), while the Rhopalosomatinae include the three remaining recent macropterous genera (Liosphex Townes, 1977;Paniscomima Enderlein, 1904;and Rhopalosoma Cresson, 1865). Engel also included the fossil genus Propalosoma in his Rhopalosomatinae, but this genus is now considered a formicoid wasp (Archibald et al., 2018). Unfortunately, Engel's subfamily division was not based on a phylogenetic analysis, and most diagnostic characters listed by him for the Rhopalosomatinae are most likely plesiomorphies for the family rather than derived characters for the subfamily. Nevertheless, Engel's classification is supported by earlier morphological and unpublished molecular analyses suggesting Olixon to be the sister to a monophylum comprising the recent macropterous genera (Brothers, 1999;Guidotti, 1999;Blaschke et al., unpublished results). Guidotti (1999) listed 22 potential synapomorphies for the Rhopalosomatinae, but for most of these it is not clear whether they have evolved in the stem of the family or the subfamily mainly due to (a) the highly derived morphology of Olixon, which is linked to the reduction of wings, and (b) the lack of stem group fossils elucidating the early evolution of the family.
Here, we describe and illustrate an enigmatic new genus of rhopalosomatid wasps from mid-Cretaceous Burmese (Kachin) amber using four recently discovered specimens with a remarkable mix of characters. Based on our analyses, we discuss key characters of the family and provide an updated checklist of the world genera and fossil species of Rhopalosomatidae. In order to validate the origin of the spec-imens, we performed chemical analyses of three of the four amber pieces containing the investigated fossils.
2 Material and methods

Material studied
The new fossils have been compared with type material of different species representing all genera currently assigned to the family Rhopalosomatidae, i.e., Eorhopalosoma, Liosphex, Olixon, Paniscomima, and Rhopalosoma. Furthermore, more than 30 additional rhopalosomatid wasps in Burmese amber have been studied to investigate the morphological variation in these Cretaceous fossils. The material used and described in this study is deposited in the following institutions: NIGP -Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences (CAS), Nanjing, China SNSB -Bayerische Staatssammlung für Paläontologie und Geologie, Staatliche Naturwissenschaftliche Sammlungen Bayerns, Munich, Germany UMB -Übersee-Museum Bremen, Bremen, Germany ZMB -Museum für Naturkunde, Berlin, Germany.

Origin and age of the amber
The four newly reported specimens are each preserved in a clear, yellowish piece of amber. Burmese amber originates from several mining sites in Kachin State ("Kachin amber", ca. 98 Ma), Sagaing Region ("Hkamti amber", ca. 110 Ma), and Magway Region ("Tilin amber", ca. 72 Ma), but distinction between these deposits is rarely documented on the amber market (Zheng et al., 2018;Xing and Qiu, 2020). Miocene amber and even recent copal also apparently exist in the country, although with no known precise locality (Cockerell, 1922). Some pieces of these younger resins are sometimes offered by traders among bags of so-called "mid-Cretaceous Kachin amber" (Vincent Perrichot, personal observation, 2019), making their actual origin difficult to distinguish. The holotype MB.I 6500 studied herein was originally accessed by Sieghard Ellenberger (Kassel, Germany) in the early 2010s, directly from the Noije Bum site of the Hukawng Valley in Kachin State. It was then sold to the late Jens-Wilhelm Janzen (Seevetal, Germany) and eventually purchased by the ZMB. The three other fossils investigated herein were accessed no later than 2017 and supposedly originate from mines in the Hukawng Valley as well. Since one of our samples was obtained from an online auction (VL-Bu-04), we performed a chemical analysis to ascertain its amber vs. copal nature and its affinities with two of the other amber pieces. Chemical analyses confirmed the samples to be similar to the referenced Kachin amber (see results in Sect. 3.1 and the discussion below), thus from mines Engel (2017) a Three additional fossil taxa, originally described within the family, have been transferred to other groups, i.e., Mesorhopalosoma Darling (in Darling and Sharkey, 1990) (mid-Cretaceous) and Paleorhopalosoma Nel et al., 2010 (Paleocene) have been transferred to sphecoid wasps (Osten, 2007;Archibald et al., 2018) and Propalosoma Dlussky and Rasnitsyn, 1999 (middle Eocene) to myrmeciine ants (Archibald et al., 2018). Furthermore, Grimaldi and Engel (2005) and Engel (2008) make mention of an additional fossil, Lithoserix williamsi Brown, 1986, which purportedly had been assigned to the family by Nel (1991). But in his publication Nel assigned the fossil as "Hymenoptera Vespina Vespomorpha (?), famille incertae sedis, nov. stat." and makes no mention of the family Rhopalosomatidae. Actually, the species was placed within Ichneumonidae by Kasparyan and Rasnitsyn (1992). b The larva may or may not belong to one of the rhopalosomatid species already described from Burmese amber.
in the Hukawng Valley of Kachin State, northern Myanmar. The age of Kachin amber has been established as approximately 98 Ma based on radiometric dating of zircons (Shi et al., 2012), and bivalve borings and an ammonite found in an inclusion support an age dating to around the Albian-Cenomanian boundary (Smith and Ross, 2017;Yu et al., 2019). This period is generally referred to informally as the "mid-Cretaceous".

Thermochemolysis-gas chromatography-mass spectrometry
Amber pieces including three of the four new specimens (MB.I 6500, SNSB-BSPG 2020 XCIII 25, VL-Bu-04) and the rhopalosomatid larva published by Lohrmann and Engel (2017; here VL-Bu-10) were analyzed by thermally as-sisted hydrolysis and methylation (THM) also called thermochemolysis. A yellowish Baltic amber was analyzed as a reference to determine their chemical class. An additional amber piece containing a female paratype of the new species (NIGP173865) could not be analyzed for safety reasons (the amber surface was very close to the fossil inclusion, so the risk of damage during amber sampling was too high), but since the fossil it contains is conspecific with the three others we assume that all four pieces are of the same age and NIGP173865 cannot originate from the older Hkamti site or younger Tilin site. Small portions of amber were crushed manually in an agate mortar. The powders (∼ 0.1 mg) were heated with TMAH (tetramethylammonium hydroxide) at 400 • C. THM was carried out using a Frontier Lab PY-2020iD pyrolyzer coupled with a Shimadzu GCMS-QP2010 Plus system oper-218 V. Lohrmann et al.: †Cretolixon -a remarkable new genus of rhopalosomatid wasps ating with a split ratio of 100. Separation was achieved using a capillary column SLB-5MS (60 m × 0.25 mm inner diameter, 0.25 µm film thickness) with the carrier gas He with a flow of 1.1 mL min −1 . The operating conditions were as follows: initial temperature held at 50 • C for 2 min and increased to 310 • C at a rate of 4 • C min −1 for 3 min. Individual compounds were identified based on comparisons of (i) mass spectrometry (MS) data with the NIST 2014 library and literature (Anderson, 1994(Anderson, , 1995Dutta et al., 2011;Fischer et al., 2017) and (ii) retention time with the Baltic amber.
The relative distribution of identified compounds was determined by measuring the area of a specific fragment, denoted m/z integration in Table A1. The peak area of the selected m/z for each compound was integrated and corrected by a mass spectra factor (Table A1) calculated as the reciprocal of the proportion of the fragment used for the integration and the entire fragmentogram. The molecular ratios were calculated using those corrected areas, which allows an approximation of areas on the total ion chromatogram while preventing the integration of co-eluting compounds.

Preparation and study of fossils
All amber pieces were manually polished using a Buehler Metaserv 3000 polisher and Buehler CarbiMet silicon carbide papers to obtain flat surfaces for optimal observation and imaging of the insect inclusions.
The inclusions were studied with a Leica MZ12 and a Zeiss SteREO Discovery.V20 stereomicroscope. Measurements were taken with an ocular micrometer. However, due to the preservation in amber not all characters could be measured in a direct vertical view. Thus, measurements are approximations rather than exact numbers.
Extended-focus images were taken with a Leica DFC490 digital camera with a Leica Z16 APO system or with the BK PLUS Lab System (Dun, Inc.) with a Canon MP-E 65 mm lens mounted on a Canon 6D camera. Pictures were aligned and stacked using the software package Auto-Montage Essentials by Syncroscopy (Version 5.03.0061 ES) or Zerene Stacker under the PMax value.

Terminology
The general morphological terminology is adopted from Huber and Sharkey (1993) with additions from Mason (1986Mason ( , 1990, and the following morphological abbreviations and modifications are used in the text: Cu 2 is the section of fore wing cubitus separating the 2Cu and 2M (not closed) cells (Fig. 7). FI, FII, etc. are flagellomere I, flagellomere II, etc. IOD is the interocellar distance or the shortest distance between the lateral ocelli. LOD is the maximum diameter of a lateral ocellus. MOD is the maximum diameter of the median ocellus. M 1 is the section of fore wing media separating the R and 1M cells (Fig. 7). M 2 is the section of fore wing media separating the 1M and 1Rs cells (Fig. 7). OOD is the ocellocular distance or the shortest distance between the lateral ocellus and eye. Rs 1 is the section of the fore wing radial sector separating the 1R1 and 1Rs cells (Fig. 7). Rs * is the section of the hind wing radial sector separating the R and R1 cells (Fig. 7). Tarsal fenestrae are sublateral, slightly transparent, sclerotized, less setose areas at the apex of the female tarsomeres in many macropterous Rhopalosomatidae. TL is the temple length or shortest distance between the posterior eye margin and occipital carina.

Nomenclatural acts
This published work and the nomenclatural acts it contains have been registered in ZooBank (http://www.zoobank.org/, last access: 3 December 2020, date of registry: 5 December 2019), with the following LSID (reference): urn:lsid:zoobank.org:pub:3650519D-0470-42E2-97D7-A267658C0B4F. The electronic edition of this work has been archived and is available from the following digital repositories: Deutsche Nationalbibliothek, US Library of Congress, Portico, and CLOCKSS.

Chemical analysis
The total ion chromatograms of the thermochemolysis-gas chromatography-mass spectrometry (THM-GC-MS) analysis of the Burmese resins are illustrated in Figs. 1 and B2 (see also Fig. 2 for a comparison of the reconstructed chromatograms of Baltic and Burmese amber). The identified compounds, their fragments used for integration, and their mass spectra factors are listed in Table A1. The relative proportions of the (i) different categories of compounds and (ii) labdanoid diterpenes and compositional ratios are given in Table 2.

Systematic paleontology
Order Hymenoptera Linnaeus, 1758 Family Rhopalosomatidae Ashmead, 1896 Rhopalosomidae Ashmead 1896: 303. Type genus: Rhopalosoma Cresson, 1865. (Ashmead preferred his form of the family name to Rhopalosomatidae (see footnote on page 303). In doing so, he used an incorrect spelling for the stem of his proposed new family name. Since the correct combining stem is "Rhopalosomat-", the family name was consequently corrected to Rhopalosomatidae by Brues (1922)).
With only eight enclosed fore wing cells (C, R, 1Cu, 1R1, 2R1, 1Rs, 1M, and 2Cu), the wing venation of fully winged rhopalosomatid species may superficially resemble those of some other aculeates, e.g., extinct Formiciinae, Sphecomyrminae, and some Myrmeciinae, whereas members of the genus Olixon are brachypterous or apterous. However, Rhopalosomatidae can be distinguished from them by the very narrow fore wing costal cell, the relatively large and oblique fore wing cell 1M, and the position and form of the hind wing veins Rs * (short, straight, and reclivous but angled or curved in most Rhopalosoma) and cu-a (basal or opposite of the branching of veins M and Cu).

Sexual dimorphism
Apart from the genital morphology, males and females can be differentiated by the following: the number of flagellomeres -females have 10, males have 11; the number of visible metasomal tergites -females have 6, males have 7; the form of the tarsomeres II-IV -cylindrical for males and dorsoventrally flattened for females; the absence or presence of tarsal plantulae -absent in females, present in males; and the form of the apex of the pretarsal claws -simple in females, bifid in males.
Genus Additionally, females of Cretolixon can be easily identified by the six apical bristles on the first flagellomere (Figs. 5b, 6e, 8a) -the maximum number of apical bristles on flagellomere I in any other known rhopalosomatid wasp is two (as in Fig. 3a, d).
Mesosoma. Pronotum rather long, anterad strongly narrowing in dorsal view, dorsal posterior margin strongly concave. Notaulus apparently absent. Parapsidal sulcus straight, slightly longer than half of mesoscutal length, reaching posterior mesoscutal margin. Mesosternal lobes present, separated from mesosternum by slight depression (but not constricted as in Paniscomima and Rhopalosoma). Metapleuralpropodeal suture carinate (Fig. 4c). Propodeum with median slightly longitudinal tubercle on anterior section. Dorsal face of propodeum in lateral view convexly rounded.
Legs. Fore leg with one tibial spur. Middle and hind legs with two tibial spurs each, inner spur of hind leg modified as calcar. Tarsomeres I-IV each with plantar lobe. Pretarsal claws bifid. Arolium present.
Fore wing. With eight enclosed cells: C, R, 1Cu, 1R1, 2R1, 1Rs, 1M, and 2Cu (Figs. 4e, 7). Costal cell very narrow, over complete distance narrower than bordering veins. Pterostigma narrow. Cell 2R1 (marginal cell) short, ending well before apex of wing. Base of vein M + Cu neither tubular nor present as remnant. Vein cu-a slightly curved, distad of M 1 by 1.0× length of cu-a. Vein M 1 curved, intersecting vein Cu at an angle of almost 90 • . Vein 1m-cu straight. Faint remnant of vein 2m-cu received by distal section of second submarginal cell (1Rs). Vein Rs 1 almost straight, about 1.7× length of M 2 . Cell 1Rs 2.1× as long as high. Vein CuA distinct, CuP barely present but very short and inconspicuous. Anal lobe without adventitious vein. Wing membrane covered and ventral margin fringed with short, evenly distributed setae.
Hind wing. With two complete cells: R and Cu (Fig. 7). Costal cell only present as membranous area anterior to Sc + R, not enclosed by C. Jugal lobe absent. Anal vein tubular and pigmented, ending at cu-a. Sc + R tubular and pigmented, at basal two-thirds fused with C. Vein R tubular not reaching further than distal hamuli. Vein Rs tubular, straight and slightly reclivous at basal section (Rs * ), distal section not tubular but weakly pigmented; pigmentation distally fading. Vein M + Cu tubular and slightly pigmented. Vein Cu not tubular but markedly pigmented, not reaching distal half of wing membrane. Vein M basally tubular but not pigmented, diverging from vein M + Cu by its length beyond cu-a, ending even before connecting with r-m. Vein cua oblique, tubular but not pigmented. Five to six basal hamuli straight, their length continuously increasing distad. Five to six distal hamuli. Distal hamuli curved, hook-shaped; similar in size and form. Wing membrane covered and ventral margin fringed with short, evenly distributed setae.
Metasoma. Segment I less than 1.5× as long as wide. Posterior half of tergite I with median depression broadening posterad. Cercus paddle-shaped and with apical setae. Parameres spine-like, upcurved.
Pilosity. Body completely covered with fine pubescence. Pubescence on metasoma slightly longer than on rest of body.

Etymology
The new genus group name is a combination of "cret-" (referring to Cretaceous) and the name of the recent genus Olixon. The name is neuter.
Paratypes. Female (Fig. 5), VL-Bu-04, deposited in the Übersee-Museum Bremen, Bremen, Germany (ex coll. Volker Lohrmann), mid-Cretaceous (Albian-Cenomanian) amber from northern Myanmar, Kachin State, Hukawng Valley. The specimen is almost completely preserved without any significant distortions. The left side of the mesosoma and base of the wings is hidden under an air bubble. Tarsomeres II-V of the right hind leg are missing. In order to access crucial characters of the fossil the amber piece has been cut into two fragments. The following syninclusions have been observed: one Coleoptera (Staphylinidae), one Cicada (nymph), one Blattodea (nymph), plant fibers, and stellate hairs. Female (Fig. 6b, e), NIGP173865, deposited in the amber collection of the Nanjing Institute of Geology and Palaeontology, Nanjing, China, mid-Cretaceous (Albian-Cenomanian) amber from northern Myanmar, Kachin State, Hukawng Valley. The specimen is incompletely preserved but without any significant distortions. The posterior half of the second metasomal segment and remainder of the metasoma, as well as apical sections of fore wings and hind wings, are missing. Additionally, a fracture is hiding large parts of the anterodorsal mesosoma.
Male (Fig. 6a, c, d), SNSB-BSPG 2020 XCIII 25, deposited in the Staatliche Naturwissenschaftliche Sammlungen Bayerns, Munich, Germany (ex coll. Patrick Müller, BUB 3227), mid-Cretaceous (Albian-Cenomanian) amber from northern Myanmar, Kachin State, Hukawng Valley. The specimen is almost completely preserved without any significant distortions. The following parts of the legs are missing: tarsomeres II-V of right fore leg, apical half of tarsomere II and tarsomeres III-V of left middle and right hind legs, and tarsus and tibial spurs of left hind leg. Additionally, several air bubbles hide parts of the left lower mesosoma and parts of the wing. The following syninclusions have been observed: parts of a filiform antenna and leg of an unidentified insect, stellate hairs, and plant fibers.

Diagnosis
As for the genus (vide supra).

Description
As for the genus, with the following additions (n.m. = not measurable, due to preservation).

Etymology
The specific epithet alatum (ala (L.) = wing) refers to the presence of fully developed wings.

Chemical analysis and the origin and botanical source of the amber
Our thermochemolysis of the investigated fossils revealed chromatograms that are dominated by bicyclic products. These products represented 41 % (MB.I 6500) to 60 % (VL-Bu-10) of the identified compounds, which is characteristic of class I ambers. The retention times of the methyl esters of C 15 labdanoids were similar in Burmese ambers and in the reference Baltic amber, which indicated that these acids were derived from polymerized communic acid and a regular configuration of labdanoid diterpenes, typical of class Ia and Ib ambers (Fig. 2). The absence of succinic acid classified Burmese ambers as class Ib resins (Anderson, 1994). Among labdanoid diterpenes, the main products were 1-methylbicyclic hydrocarbons that are supposed to derive from A-ring defunctionalization of the polylabdanoid structure (Anderson, 1995). These compounds represented 57 % (VL-Bu-04) to 77 % (SNSB-BSPG 2020 XCIII 25) of labdanoid diterpenes. Their proportion is assumed to increase with maturation (Clifford et al., 1999). Compared with other Cretaceous and Triassic class Ib ambers, the large predominance of 1-methylbicyclic hydrocarbons among labdanoid diterpenes appears as a characteristic element of the Burmese class Ib ambers since it has already been described for a Cretaceous amber from Myanmar (= mid-Cretaceous Kachin amber; Dutta et al., 2011). The chemical composition of these ambers clearly differs (i) from Raritan (Anderson, 2006), Austrian (Triassic;Fischer et al., 2017), and Canadian (Poulin and Helwig, 2015) ambers whose pyrograms are dominated by the methyl ester of C 15 labdanoids and (ii) from the probable pinaceous resin from the Isle of Wight whose pyrogram is dominated by abietane diterpenes and whose main compound is dehydroabietic acid methyl ester (Bray and Anderson, 2008). However, it cannot be used as a diagnostic chemical composition since a Cretaceous French amber from the Alps mountains also exhibits this predominance (Nohra et al., 2015; see also Table 2).
Compositional ratios calculated on labdanoid diterpenes distributions and based on the relative proportions of C 14 over C 15 bicyclic products ranged from 0.2 (SNSB-BSPG 2020 XCIII 25) to 0.3 (VL-Bu-04), which highlights the predominance of C 15 compounds. This type of ratio has been suggested to decrease with maturation (Anderson et al., 1992;Clifford et al., 1999). Such values for these compositional ratios classified these ambers in the lower range of previously described Tertiary class I resinites (Clifford et al., 1999).
The second most abundant category of compounds produced by the thermochemolysis of Burmese ambers comprise methyl(hydro)naphthalene and methylphenanthrene. This category of compounds included sesquiterpenoids 5,6,7,8-tetrahydrocadalene (22) and norcadalene (29) and diagenetic products of sesquiterpenoids and diterpenoids. These molecules represented 24 % (VL-Bu-10) to 40 % (MB.I 6500 and SNSB-BSPG 2020 XCIII 25) of the identified compounds, and their proportion is assumed to increase with maturation (Clifford and Hatcher, 1995). Together with the high proportions of 1-methylbicyclic hydrocarbons and the low compositional ratios for labdanoid diterpenes, this relatively high proportion of methyl(hydro)naphthalene and methylphenanthrene could indicate that Burmese ambers are characterized by a high degree of maturation.
In conclusion, the results of THM-GC-MS highlight the amber rather than copal nature of these materials. Their chemical compositions correspond to class Ib ambers that would have undergone a relatively high degree of maturation. They are in accordance with previously published analyses of Burmese Cretaceous ambers from Kachin (e.g., Dutta et al., 2011) and are clearly distinct from Campanian Tilin amber (Zheng et al., 2018). So far, Hkamti amber has not been analyzed by THM-GC-MS, but it is significantly older than Kachin amber (Xing and Qiu, 2020) and should arguably be distinct as well. The attested origin of the holotype MB.I 6500, from Noije Bum Hill, supports the same origin as this or another site of the Hukawng Valley for the three other specimens.

Cretolixon and Rhopalosomatidae
Cretolixon is an enigmatic new genus, which can be confidently assigned to the crown group rhopalosomatids. Specifically, the presence of diagnostic characters such as a very narrow fore wing costal cell, distinct apical bristles on the basal flagellomeres, posterad paired mesosternal lobes covering the bases of the mesocoxae, laterally widened female tarsomeres, upcurved parameres, and an upcurved sting support the placement of the genus in the family. On the other hand, a placement among the two recent subfamilies is far more challenging and without a phylogenetic analysis rather difficult. Thus, we choose a conservative approach and will test the placement of Cretolixon in a future study. However, in the following, we will discuss key characters and their implications for a potential subfamily affinity in more detail (see also Figs. 9 and 10).
Cretolixon -a common ancestor of Rhopalosomatinae and Olixoninae?
As it would "undoubtedly" render the Rhopalosomatinae paraphyletic, Krogmann et al. (2009) argued against Engel's subfamily classification, thus ignoring earlier morphological studies supporting a sister group relationship between the macropterous rhopalosomatids and the brachypterous Olixon, i.e., Brothers and Carpenter (1993), Brothers (1999), and Guidotti (1999). Brothers and Carpenter (1993) and Brothers (1999), based on their phylogenetic analyses, list nine potential synapomorphies for the Rhopalosomatinae. However, since most of these characters are not applicable to the morphologically aberrant Olixon it remains unclear whether these character states are novelties for the subfamily or whether they have evolved in the stem of the family with a secondary loss or derivation in the Olixoninae. The same might be true for the 22 characters received by Guidotti (1999) in her analysis of the generic relationships of rhopalosomatid wasps.
For the first flagellomere, however, this line of arguments needs a slight modification. The single long bristle on the apex of flagellomere I of some Olixon species is most likely homologous to one of those two apical bristles present in the macropterous genera, and these again are most likely homologous to two of the six found in female Cretolixon (Figs. 5b, 6e, 8a).
Another important character of Cretolixon is the position of the hind wing cross vein cu-a (Fig. 7), which is, similarly to its position in recent genera, clearly basal to the branching of the veins M and Cu (similar to Fig. 9b, c). In Eorhopalosoma and most of the unpublished Burmese amber fossils of the family the cross vein cu-a is opposite or very near to the branching of the veins M and Cu (similar to Fig. 9a). Brothers (1975) and Brothers and Carpenter (1993) regarded the basal position of the cross vein cu-a as the ancestral charac-ter state among aculeates and a distal position as the derived state but mentioned that several reversals back to the ancestral state occurred throughout the aculeates, and this might be true for Cretolixon and the recent rhopalosomatid genera as well.
Cretolixon and its implication for the biology of early rhopalosomatids Lohrmann and Engel (2017) showed that the ectoparasitoid larval habit of rhopalosomatids has not changed since the Cretaceous, and the discovery of enlarged ocelli in Eorhopalosoma (Engel, 2008;Boudinot and Dungey, 2020) indicates that early members of the family may have had crepuscular or nocturnal habits similar to those of recent rhopalosomatids. In contrast to Eorhopalosoma and Figure 10. Distribution of the discussed characters among Rhopalosomatidae. The tree topology of the recent genera is based on Brothers and Carpenter (1993), Brothers (1999) (both: "rhopalosomatids" sister to Olixon), Guidotti (1999), and unpublished molecular data (Blaschke et al., unpublished results) (both: Olixon + (Liosphex + (Rhopalosoma + Paniscomima))). most species of Paniscomima and Rhopalosoma (similar to Fig. 9g), Cretolixon has moderately small ocelli (Figs. 4b, 5b, 6b); i.e., the size is similar to those found in Liosphex (Fig. 9i). Otherwise, only Paniscomima seyrigi (Berland, 1951) has ocelli of a similar size (Fig. 9h). While this is surely the plesiomorphic character state within the family, it does not allow any conclusion with respect to the biology of the species.
The two previously described Burmese rhopalosomatid taxa are known from males only. Thus, C. alatum gen. et sp. nov. represents the first Cretaceous rhopalosomatid species known from both sexes and thus allows for investigations of sexual dimorphism. As known from recent Rhopalosomatidae, Cretolixon shows only moderate sexual dimorphism. Apart from the reproductive organs, the visible differences are limited to the number of visible metasomal tergites and the morphology of the antennae and legs.

Conclusions
With its unique mixture of characters, Cretolixon gen. nov. represents a significant piece of evidence for our understanding of the early evolution of Rhopalosomatidae. Cretolixon not only provides further evidence for the monophyly of the family but also contributes evidence for the monophyly of the Rhopalosomatinae. A robust phylogenetic placement of the new genus, however, is not possible without a detailed phylogenetic study, which will be part of an ongoing collaboration including morphological and molecular data.

Taxonomic and ethical statement
The current contribution includes revised parts of a chapter used earlier in the PhD thesis of Volker Lohrmann, i.e., the description of the male holotype of Cretolixon alatum as well as Fig. 3. Because the thesis includes a statement disclaiming all relevant nomenclatural acts for taxonomic use, the new taxon is formally described here to make its name available according to the ICZN (1999).
We are well aware of the ethical problems linked to the study of Burmese amber that originates from excavation sites in Kachin State, Myanmar (e.g., Sokol, 2019). We agree with the general intentions of two recently published opinions, i.e., Haug et al. (2020) and Szwedo et al. (2020), and declare that all Kachin amber pieces included in this study are deposited in public collections and have been collected and subsequently traded before November 2017 -the time when the military troubles in the amber mine area of Kachin State started (Human Rights Council, 2019).   Figure B2. Total ion chromatograms from the THM-GC-MS analysis of Burmese ambers. Numbered peaks refer to identified compounds in Table A1 and Fig. B1.
Author contributions. VL designed the study. QZ, PMü, and VL provided fossils. VL and VP carried out the preparation of the fossil amber samples for the THM-GC-MS analysis and for microscopic studies of the fossils. LJ did the chemical analysis. VL prepared the taxonomic descriptions. JB provided unpublished molecular data integrated in Fig. 10. PMi prepared the photographs of the fossils. VL, VP, and LJ wrote the first draft of the paper. All authors contributed to the discussion and proofreading of the manuscript.
Competing interests. The authors declare that they have no conflict of interest.
Acknowledgements. We sincerely thank the late Jens-Wilhelm Janzen (Seevetal, Germany) and Barbara Odifredi (Carrara, Italy) for making two of the herein newly reported amber fossils available and David Grimaldi (American Museum of Natural History, New York, USA), Bo Wang (Nanjing Institute of Geology and Palaeontology, Nanjing, China), and Christian Neumann (Museum für Naturkunde, Berlin, Germany) for the loan of material. Furthermore, we thank Brendon Boudinot (University of California, Davis, USA) for double-checking important characters of the type of Eorhopalosoma lohrmanni, Takuma Yoshida (Hokkaido University, Sapporo, Japan) and his team for sharing their knowledge of the first record of Paniscomima from the Palearctic Region, and Sieghard Ellenberger (Kassel, Germany) for information on the holotype's origin. For their generous financial support enabling the UMB to purchase additional rhopalosomatid inclusions in Burmese amber, we would like to thank the Petra und Joachim Schaffer-Stiftung (Bremen, Germany) and Sabine Stuth (Bremen, Germany), as well as Irmgard and Nils Kadelbach (Bremen, Germany). Finally, we thank the reviewers, Denis J. Brothers (University of KwaZulu-Natal, Scottsville, South Africa) and Alexandr P. Rasnitsyn (Russian Academy of Sciences, Moscow, Russia), and the editor, Alexander Schmidt (University of Göttingen, Germany), for their valuable comments and time spent on the manuscript.
Financial support. Financial support in the initial and final stage of this project was provided by the Studienstiftung des Deutschen Volkes (to Volker Lohrmann), grant OSUR from program AO2 2018 for visiting researchers (to Vincent Perrichot, enabling Volker Lohrmann's work at the University of Rennes 1) and by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB26000000) and the Second Tibetan Plateau Scientific Expedition and Research (2019QZKK0706; both to Bo Wang and enabling the purchase of the NIGP specimen studied herein and Vincent Perrichot's work at NIGP, respectively).
Review statement. This paper was edited by Alexander Schmidt and reviewed by Denis J. Brothers and Alexandr P. Rasnitsyn.