The numbers prefixed with acronyms, e.g. “USNMENT” or “OSUC”, are unique identifiers for the individual specimens (note the blank space after some acronyms). Details on the data associated with these specimens may be accessed at the following link: purl.oclc.org/NET/hymenoptera/hol and entering the identifier in the form. Persistent URIs for each taxonomic concept were minted by xBio:D in accordance with best practices recommend by Hagedorn et al. (2013). Morphological terms were matched to concepts in the Hymenoptera Anatomy Ontology (Yoder et al. 2010) using the text analyzer function. A table of morphological terms and URI links is provided in Appendix 1. Taxonomic synopses and matrix-based descriptions were generated from the Hymenoptera Online Database (hol.osu.edu) and the online program vSysLab (vsyslab.osu.edu) in the format of character: state. Characters that were not visible because of preservation or orientation of the insects are coded as “not visible”.

The amber pieces were cut and polished to optimize the viewing and photography of specimens for taxonomic study. Direct examination of the specimens was made with a Zeiss V8 stereomicroscope and an Olympus BX51 compound microscope.

Photographs were captured with multiple imaging systems: a Z16 Leica lens with a JVC KY-F75U digital camera using Cartograph and Automontage software; an Olympus BX51 compound microscope with a Canon EOS 70D digital SLR camera; and a Leica DM2500 compound microscope with a Leica DFC425 camera. Illumination was achieved with a lighting dome or with LED gooseneck lamps and mylar light dispersers. Images were rendered from z-stacks with Automontage, Helicon Focus or Zerene Stacker. In some cases, multiple montage images were stitched together in Photoshop to produce larger images at high resolution and magnification. Full resolution images are archived at the image database at The Ohio State University (specimage.osu.edu).

Dissections for scanning electron microscopy were performed with a minuten probe and forceps and body parts were mounted to a 12 mm slotted aluminum mounting stub (EMS Cat. #75220) using a carbon adhesive tab (EMS Cat. #77825-12) and sputter coated with approximately 70 nm of gold/palladium using a Cressington 108 auto sputtercoater. Micrographs were captured using a Hitachi TM3000 Tabletop Microscope at 15 keV.

EJT: photography, scanning electron microscopy, taxon concepts, manuscript preparation; NFJ: manuscript preparation; CKS: manuscript preparation, provision of Burmese amber; DR: manuscript preparation, provision of Burmese amber.

The amber specimens of Proterosceliopsis studied here were collected from Kachin (Hukawng Valley) of northern Myanmar, which was dated at 98.79+0.62 Ma (Cruickshank and Ko 2003, Shi et al. 2012), equivalent to the earliest Cenomanian and approximately 1 Myr within the boundary between the Early and Late Cretaceous (Walker et al. 2012). This deposit yielded many well-preserved insect fossils (Chen et al. 2018a–b, Li et al. 2018, Wang et al. 2016, Zhang et al. 2018).

Specimens on which this work is based are deposited in the following repositories with abbreviations used in the text:

CCHH Hoffeins Collection, Hamburg, Germany

CNCI Canadian National Collection of Insects, Ottawa, Canada

CNU Key Lab of Insect Evolution and Environmental Changes, Capital Normal University, Beijing, China

FSCA Florida State Collection of Arthropods, Gainesville, FL, USA

KUNH Kansas University Natural History Museum, Lawrence, KS, USA

OPPC Ovidiu Popovici, personal collection, “A.I. Cuza” University, Faculty of Biology, Iasi, Romania

OSUC C.A. Triplehorn Insect Collection, The Ohio State University, Columbus, OH, USA

USNM National Museum of Natural History, Washington, DC, USAA

1Rs (Fig. 15) 1^st radial sector vein

2Rs (Fig. 15) 2^nd radial sector vein

b (Fig. 15) bulla

C, C+R (Fig. 15) marginal vein

Cu (Fig. 15) cubital vein

fas (Fig. 10) facial striae

ff (Figs 43–44) felt field

M (Fig. 15) medial vein

mas (Fig. 10) malar striae

mees (Figs 16, 20) mesepimeral sulcus

ms (Figs 7–8, 10) malar sulcus

net (Figs 16–17, 19) netrion

p (Fig. 22) cuticular pores

pp (Fig. 17) mesopleural pit

prcs (Figs 16, 19, 21) pronotal cervical sulcus

ps (Figs 1–6) papillary (basiconic) sensilla

r (Fig. 15) stigmal vein

R1 (Fig. 15) postmarginal vein

Rs+M (Fig. 15) basal vein

sk (Figs 11–13) skaphion

Sc+R (Fig. 15) submarginal vein

S6 (Fig. 47) metasomal sternites 6

T6–T8 (Figs 45–49, 55, 61) metasomal tergites 6–8

tel (Figs 16, 19, 21) transepisternal line

Most of the diagnostic characters of Proterosceliopsis can be found in extant platygastroids but are present in a unique combination in this genus. Additionally, Proterosceliopsis exhibits significant differences from each of the taxa with which it shares characters. We present discussions of these characters and their distribution among platygastroid taxa as a prelude to the generic treatment.

Antenna

Bin (1981) defined the antennal clava in Telenominae (Scelionidae) based on the presence of papillary sensilla on the ventral side of the distal antennomeres (previously referred to as basiconic sensilla). The presence of these sensilla on female antennae is a synapomorphy for Platygastroidea (Austin et al. 2005), and Bin’s definition of the clava has since been applied to all members of the superfamily. We continue to define the clava on the basis of papillary sensilla instead of relative antennomere size for multiple reasons: antennomeres can increase in size gradually (Figs 2–4); the size among antennomeres with papillary sensilla can be variable (Fig. 5); distinctly enlarged antennal segments may not have papillary sensilla, as in the clubbed antennae found in most males of Helava Masner & Huggert (Masner and Huggert 1989, Talamas and Masner 2016). The claval formula is often of taxonomic value, particularly at the species-level, and its use requires that specimens be preserved well enough to observe these structures. Small bubbles in the amber are sometimes visible at the tips of the sensilla (Fig. 6) and can be useful for identifying their presence.

A transverse orientation of papillary sensilla is known only in Nixonia (Fig. 4) (Johnson and Masner 2006) whereas all other extant platygastroids have a longitudinal arrangement (Fig. 1). We found that in some species of Proterosceliopsis (Figs 3, 5) and Cretaceous Electroteleia Brues (Fig. 2) the sensilla are arranged at an oblique angle. This was found in multiple specimens in which the general shape of the antennomeres is intact, and thus we do not attribute this to taphonomic deformation. Instead, we consider that different arrangements of these sensilla were possible earlier in the history of Platygastroidea.

Malar sulcus

The malar sulcus found in Proterosceliopsis is unaccompanied by facial or malar striae (Fig. 7), a state which can be found in some Scelionidae (Fig. 8) and some Sparasionini (Electroteleia). In Platygastridae, the malar sulcus is present in only two genera: Metaclisis Förster and Orseta Masner & Huggert, each of which have both facial and malar striae (Fig. 10) (Masner and Huggert 1989), and Orwellium, which does not have facial or malar striae. The malar sulcus is entirely absent in Nixonia Masner (Fig. 9).

Palpal formula

Ortega-Blanco et al. (2014) recorded the palpal formula of P. masner to be 5:3. We observed 5-merous, cylindrical maxillary palps in P. nigon, P. torquata, and P. wingerathi; and at least 4 maxillary palpomeres in P. plurima. Two labial palpomeres are visible in P. torquata. Five-merous palpomeres were retrieved as the plesiomorphic condition for Platygastroidea in the phylogenetic analysis by Popovici et al. (2017) and are found in Sparasionini, Nixonia, and Archaeoteleia. This indicates that Proterosceliopsis is also a basal lineage well outside of Scelionidae in which the palpal formula is 4:2 or less (Popovici et al. 2017), and Platygastridae in which the palpal formula is 2:1 or less (Popovici, personal communication).

Skaphion

The anterior mesoscutum in most species of Proterosceliopsis features a smooth, transverse structure to which we apply the term skaphion (Figs 11, 50, 58, 62). We do not assert that the skaphion of Proterosceliopsis is homologous with that found in scelionid genera (Figs 12–13) and note that it can be found in other Cretaceous taxa that do not belong to Proterosceliopsis or Scelionidae, e.g. Electroteleia (Fig. 14).

Costal vein

In Proterosceliopsis, the costal vein is present anterior to the fusion with R and extends proximally beyond the bulla (Fig. 15). This form is known from other Cretaceous platygastroids (Fig. 14), but not in extant taxa.

Setation of the pronotal cervical sulcus

The pronotal cervical sulcus can take many forms, including a distinct line of foveae (Fig. 18), a well-defined smooth groove (Fig. 19), or a weakly defined furrow along the lateral pronotal rim (Fig. 17). In Nixonia, Johnson and Masner (2006) reported an area of dense setation along the anterior margin of the lateral pronotum that is often associated with solidified exudate (Fig. 20). Setation along the pronotal cervical sulcus is found in Proterosceliopsis (Fig. 16, 52, 59) and many platygastrids (Figs 19, 26), and often contains what appears to be solidified exudate. The unpublished phylogenetic analysis of Platygastridae (Blaimer et al.) indicates that stem lineages of this family have a setose pronotal cervical sulcus. Figures 65–66 illustrate the oldest known platygastrid, in Burmese amber, that exhibits this character, consistent with our treatment of it as a plesiomorphy for Platygastridae that varies significantly in some derived platygastrid lineages (Figs 21–26). Figures 21–22 illustrate the unusual form found in Sacespalus in which setation is absent and the pronotal cervical sulcus is broadly expanded. Pores can be seen in the ventral portion of the pronotal cervical sulcus of Sacespalus (Fig. 22), which is congruent with the hypothesis that this character is associated with glandular secretions. We do not know of any scelionids with a setal patch along the anterior portion of the lateral pronotum but note that the antespiracular setal patch (sensu Yoder et al. 2012) in Scelionidae is associated with cuticular pores (Fig. 28). Based on the pronotal position of these setal patches, we suspect that they have similar function.

Netrion

A netrion is clearly present in 3 of the 5 species that we here describe. Mikó et al. (2007) reported that in Nixonia the trachea associated with the anterior thoracic spiracle extends ventrally between the netrion sulcus and the posterior pronotal inflection. Externally this results in the netrion sulcus dorsally terminating anterior to the anterior thoracic spiracle (Fig. 20). This character can be used to separate Nixonia from nearly all other platygastroids, in which the netrion sulcus, when present, terminates posterior or ventral to the anterior thoracic spiracle (Figs 16–18, 20).

Transepisternal line

Among extant platygastroids, the transepisternal line is found exclusively in Platygastridae (Fig. 19). We treat it as a plesiomorphy for Platygastridae based on its presence in the Cretaceous specimen illustrated in Figs 65–66, and because it is found in stem lineages of the family based on a preliminary analysis of molecular data (Blaimer et al.). The internal anatomy associated with this structure has yet to be examined in detail and will likely shed light on its function and evolution within Platygastridae. The transepisternal line in Proterosceliopsis is clearly present in all specimens from Burmese amber (Figs 16, 53) suggesting a close relationship between Proterosceliopsidae and Platygastridae.

Mesepimeral sulcus

The presence of a fully developed mesepimeral ridge was retrieved by Vilhelmsen et al. (2010) as a potential autapomorpy for Proctotrupomorpha s.s. This ridge corresponds externally to the mesepimeral sulcus, which is found in all platygastroid families except Platygastridae (Figs 16–20). We consider the loss of the mesepimeral sulcus to be an apomorphy for Platygastridae, but whether this internally corresponds to loss of the mesepimeral ridge has yet to be investigated. Some examples of an absent or weakly indicated mesepimeral sulcus can be found in Scelionidae. In some cases, these are clearly secondary derivations that occurred at the species level (Fig. 17). In the unusual Doddiella Kieffer (Scelionidae) the mesepimeral sulcus is not indicated externally but the mesepimeral ridge can be seen through the semitransparent exoskeleton (Fig. 27). The presence of the mesepimeral sulcus in Proterosceliopsidae provides a reliable mesosomal character to separate it from Platygastridae.

Setation and sulci of the metasoma

Proterosceliopsis from Burmese amber and Nixonia share the presence of transverse, depressions along the anterior margins of T1–T5 (Figs 29, 32) and S1–S5 (Figs 34, 37). In Nixonia, these depressions have visible setation. In Proterosceliopsis they appear to have very fine setation, but the cloudy exudate prevents an unobscured view. In P. torquata these depressions are not clearly visible on T6, and in P. plurima they are not clearly present on S6. In the other Burmese species of Proterosceliopsis, and in all species of Nixonia, these depressions are present on T6 and S6. We suspect that they are present on T6 and S6 in all species of Proterosceliopsis, but the preservation of these specimens prevents us from reaching a confident conclusion.

The two components of this character, setation/glandular secretion and sulci along the anterior tergites and sternites, vary independently among platygastroid families. Sparasionini and Archaeoteleia feature transverse sulci across the anterior margins of the external tergites and sternites (Figs 33, 38, 44), which may sometimes be reduced on T6 and S6 in females of Electroteleia (Sparasionini) and in Archaeoteleia. These sulci are found in all specimens that we examined in Lebanese amber, which constitute the oldest platygastroid fossils, and we consider it likely that this is the plesiomorphic condition for Platygastroidea. The metasoma in Sparasionini and Archaeoteleia lacks setation on the anterior tergites and sternites beyond T1 and S1, but there may be numerous felt fields (Fig. 44).

In Platygastridae and Scelionidae, the anterior portions of T1–T2 and S1–S2 typically feature a transverse line of foveae or pits, which may or may not be clearly defined (Figs 30–31, 35–36). The transverse sulci in scelionids are not associated with setation, and patches of dense setae tend to be located on the lateral portion of the tergite, sometimes in well-defined pits (Figs 39–40). Some teleasines (Scelionidae) exhibit a transverse sulcus along anterior T3 and S3 (Fig. 30, 35) and setal patches with pores on T4–T5 (Fig. 42). Platygastridae exhibits the greatest variation in the size, location, and shape of setal pits on metasomal segments 1 and 2. They may be present laterally or medially, as a broad transverse patch that spans the width of the tergite or sternite (Fig. 43), in well-defined pits (Fig. 31), or they may be absent entirely (e.g. Orwellium Johnson, Masner & Musetti). Felt fields, which are absent in Nixonia and Proterosceliopsis, are found only on S2 in Platygastridae and Scelionidae, with a few exceptions (e.g. Heptascelio Kieffer (Johnson et al. 2008b)). Neuroscelio doddi Galloway, Austin & Masner exhibits an unusual and noteworthy form of sternal setation in which dense tufts are located medially on S1–S2 (Fig. 41).

Setal patches on the metasoma and pronotum, which are often associated with pores in the integument, clearly warrant further examination. A survey of platygastroid morphology with a scanning electron microscope has revealed an array of additional pore locations and forms on the head and mesosoma. Our present analysis of these characters, which provides only a broad overview as it relates to family level classification of Proterosceliopsis, will be developed further in future projects. The presence of cuticular pores and solidified exudate in the same locations point to the activity of internal glands.

Ovipositor system

Ortega-Blanco et al. (2014) interpreted T7+8 to be the ovipositor, which they described as short and broad. We here clarify that these sclerites are not the ovipositor, but they could be considered part of the ovipositor system sensu Austin and Field (1997). The known diversity of ovipositor systems in Platygastroidea has expanded in recent years, with at least two derivations of a telescoping ovipositor system found in Platygastridae (Talamas et al. 2017b). Extension of the ovipositor system in Proterosceliopsis appears to operate via telescoping membrane between T6 and T7, with T7 and T8 clearly present as separate sclerites in some specimens (Fig. 45). The unpublished phylogenetic analyses suggest that conjunctival expansion between T6 and T7 occurred independently in Scelionidae and Archaeoteleia. The placement of Proterosceliopsis well outside of these families indicates that expansion of conjunctiva between T6 and T7 into a telescoping ovipositor system has occurred at least three times: in Proterosceliopsis, Archaeoteleia, and Scelionidae. We also have found evidence of conjunctival expansion between T6 and T7 in the platygastrid genus Fidiobia Ashmead. It does not appear to be telescoping but provides clear evidence of yet another independently derived elongation of conjunctiva between T6 and T7 (Figs 46–47), lending evidence to the plasticity of this morphological system.