Open Access
How to translate text using browser tools
23 October 2024 The Bee Fauna of Eocene Fushun Amber (Hymenoptera: Apoidea)
Michael S. Engel, Jiaying Xie
Author Affiliations +
Abstract

The now inaccessible amber deposits of the Fushun coalfield (Guchengzi Formation; Ypresian) represent the only diverse record of Paleogene arthropods from northeastern Asia. Among the wealth of inclusions recovered from the mines before they were closed and filled, only five specimens of bees were discovered. Meager as they are, these samples provide an important paleogeographical point of reference for piecing together the fauna of forest-dwelling bees during and after the Early Eocene Climatic Optimum. Three species in two genera are recorded, one species representing a new extinct genus and tribe of Megachilinae related to Glyptapini, Ctenoplectrellini (here including Aspidosmia Brauns), and perhaps Dioxyini, and the other two species comprising a new genus of the eusocial corbiculate tribe Melikertini (Apinae). The early-diverging tribes of Megachilinae—Glyptapini, Ctenoplectrellini, and the new tribe, all possessing a distinct metatibial scopa—are briefly reviewed. Glyptosmia Engel, n. gen., with Glyptosmia hemiaspis Engel, n. sp., is the sole member of Glyptosmiini Engel, n. tribe. Although it somewhat resembles species of the genus Ctenoplectrella Cockerell (from Baltic, Bitterfeld, Oise, and Rovno ambers), Glyptosmia also shares distinctive traits of Glyptapini (areolate propodeum) and even the cleptoparasitic Dioxyini (tuberculate metanotum). The tribe can be distinguished by the bare compound eyes, dense mesosomal punctation, tuberculate metanotum, areolate propodeum, flattened mesoscutellum with a sinuate apical margin, and characteristically thickened metatibial spurs, among other characters. Two species of the corbiculate bee tribe Melikertini are described, both of the genus Thyreomelikertes Engel, n. gen. Thyreomelikertes lacks the facial protuberances found in genera such as Aethemelikertes Engel, Haidomelikertes Engel, Amelikertotes Engel, or Succinapis Engel and is superficially similar to Melissites Engel or Mochlomelikertes Engel, Breitkreuz, and Ohl, with its long, flattened, and trapezoidal mesoscutellum somewhat reminiscent of the latter genus. The genus is also noteworthy for the putatively plesiomorphic retention of relatively developed grooves on the outer surface of the mandible and dense mesosomal pubescence. The two included species, Thyreomelikertes electrosinicus, n. sp., and T. kongi, n. sp., can be distinguished by size and the development of setae on the meso- and metatibiae. All the individuals are morphologically workers, and so, like all other melikertines, Thyreomelikertes was social and, based on the phylogenetic position of the tribe, presumably lived in anchored eusocial colonies. By contrast, G. hemiaspis was likely a free-living solitary species. The species from Fushun amber are described, figured, and compared with other species of Cenozoic and living bees. The mandibular structure of Thyreomelikertes is unique among Melikertini and permits a fuller description of the diversity of structural homologies across corbiculate bee mandibles.

INTRODUCTION

Paleogene amber rich in biotic inclusions is not uncommon, with several deposits in North America and Europe, as well as extensive exposures in northwestern India. However, northern Asia is noticeably poor in amber outcrops from the Paleocene through Oligocene, with only two outcrops bearing significant inclusions. Paleocene-Eocene amber from Sakhalin Island, Russia provides one source of data, although inclusions have been few and significant new excavation is needed before the paleobiota of this outcrop can be more extensively researched. The other, and currently the only Paleogene amber from northeastern Asia with diverse biotic inclusions, is that of the Fushun Basin in Liaoning Province, China (fig. 1A, B). The amber layers from the Fushun coalfield (fig. 1A) are of early to middle Ypresian in age (approximately 50–53 Ma), making it roughly contemporaneous with amber from the Cambay and Paris basins, and slightly older than the more extensively studied deposits across northeastern and eastern Europe (Baltic, Bitterfeld, Rovno). Given its rich biota and unique paleogeographic location, Fushun amber provides an important window into Eocene forest life in northern Eurasia and also during the Early Eocene Climatic Optimum (Wang et al., 2014).

Despite the fact that excavations of amber in the Fushun coalfield extend back to the beginning of the 20th century (fig. 1A), the study of biological inclusions from this amber is comparatively recent. For much of its history, the amber from Fushun was treated as a gemstone and used in jewelry or other materials. Ping (1931) did describe a roach nymph from Fushun amber, but a subsequent four decades would follow before the subject would be taken up again. It was not until the middle 1970s that the late You-Chong Hong (1929–2019) and colleagues began to explore systematically the insects found in Fushun amber (e.g., Hong et al., 1974; Hong, 1981, 1982, 2002a, 2002b). Unfortunately, much of the material upon which this earlier work was based was dispersed and many of the type specimens are currently untraceable. Given that the descriptions and figures of the inclusions were poor, it is challenging to be confident in the identities of the 223 species recorded in Fushun amber prior to 2002. In some cases, there is sufficient information to note gross misplacements as to family, but often it is hard to compare the historical work with material currently available. A great example of this conundrum is the abundant and diverse ants (Formicidae) from Fushun amber (Hong, 2002a), some of which cannot be placed to subfamily based on existing information. Ironically, nearly all the names proposed by Hong (2002a) are unavailable as he failed to meet the basic criteria for nomenclatural availability (ICZN, 1999). Specifically, for nearly all the species proposed by Hong (2002a) he failed to explicitly indicate a repository for his material (ICZN, 1999: Art. 16.4.2) or even to designate a holotype (e.g., in most cases he simply refers to “Material” under each species account and never identifies a specimen as the “holotype”—required by the ICZN, 1999: Art. 16.4.1). Although the names he employed can be discarded, his specimens remain of great importance if they can be restudied. Future study on Fushun amber will require detective work to locate many of Hong's samples and to newly document and describe this important fauna. That work is made all the more vital given that the mines are now closed and the exposures buried, rendering it impossible to access the amber-bearing strata, which were anyway largely removed during the excavation of the coal seams. Meanwhile, work on existing collections in China is expanding our knowledge of this biota and permitting comparisons with other Eocene faunas (e.g., Giłka et al., 2016; Stebner et al., 2016; Wang et al., 2016; Zhang et al., 2016; Azar et al., 2018; Fedotova et al., 2022).

Remarkably, during the decades of work on the fauna no bees were found or described from Fushun amber. A single meliponine fossil was attributed to Fushun amber (Engel and Michener, 2013) but subsequently the amber was found to have been Baltic amber of misrepresented provenance. Bees are uncommon as fossils but are nonetheless well represented in Eocene amber deposits, making their presumed absence from Fushun problematic. Fortunately, among existing samples several bees have now been discovered in genuine Fushun amber and these are documented here (fig. 1C). Only one other amber deposit in China is currently known to include bees, that is the significantly younger Neogene amber of Fujian Province in southeastern China (Wang et al., 2021). The Zhangpu biota includes numerous individuals of two species of resin-collecting stingless bees from the middle Miocene (Engel et al., 2021a), while the much older Fushun amber includes bees consistent with near-contemporaneous outcrops from the Eocene of Europe and India (Engel, 2001; Rust et al., 2010; Engel et al., 2013).

MATERIAL AND METHODS

For over 110 years extensive opencast mining for coal was undertaken in the Fushun Basin just south of the city of Fushun, Liaoning Province, China (fig. 1A, B). The fill in the basin consists of a series of Paleocene and Eocene sedimentary formations. Of these, several amber-bearing strata occurred in the middle and upper coal beds of the Eocene Guchengzi Formation, which largely comprises thick coal beds intercalated among carbonaceous shale (Wang et al., 2014). About 10 years ago mining ceased and the coal mines were filled in, rendering it impossible to access the amber-bearing layers, which were also largely mined out during the extraction of coal. The amber-bearing strata have been assigned to the early-middle Ypresian, 50–53 Ma, as constrained by isotopic dating, paleomagnetic data, and the fossil flora preserved in these layers (Wang et al., 2014). The flora preserved is that of a moist subtropical forest, and chemical composition of the amber points to Cupressaceae (Pinales; sensu Gadek et al., 2000) for the amber-producing tree, with the most common botanical compressions in the amber-bearing layers as well as inclusions in Fushun amber being those of Metasequoia Hu and W.C. Cheng (Sequoioideae) and other cupressaceous genera (Wang et al., 2014). Many Fushun-amber inclusions are altered through some compression and darkened by heating, the results of intrusion of magma into the coal deposits during the Oligocene (Wang et al., 2014).

FIGURE 1.

Fushun locality and diversity of pieces with bees showing colors and translucency. A. Aerial photograph of the Fushun coalfield during its period of operation. B. Map with location of Fushun coalfield indicated by red pentagon. C. Pieces of Fushun amber with bees.

img-z5-1_01.jpg

The amber pieces (fig. 1C) were trimmed and polished, as necessary, and then examined with a stereomicroscope using reflected and transmitted light. When needed, a small drop of glycerine was placed on the amber surface and then covered by a small glass coverslip to minimize optical distortion resulting from those uneven surfaces too close to the inclusions to permit polishing. Morphological terminology generally follows that of Engel (2001), Michener (2007), and Engel et al. (2017, 2021a), with terms of orientation for legs, particularly for corbiculate bees, following those of Engel et al. (2021a). Measurements were made with a Nikon SMZ 25 microscope with an attached Nikon DS-Ri2 digital camera system. As adopted in other recent works (Tran et al., 2024), Arabic numerals are used when referring to quantities (e.g., five segments, 10 flagellomeres), while Roman numerals are used to distinguish the identity of a given object (e.g., flagellomere X, for the 10th flagellomere in the series; segment V, tergum V, sternum V for the fifth segment of the metasoma and its constituent tergum and sternum).

For comparative purposes and to correct some past errors, the Fushun megachiline was compared with Eocene bees in Baltic amber of the related tribes Glyptapini and Ctenoplectrellini. A female of Glyptapis nr. fuscula Cockerell and a male of Ctenoplectrella viridiceps Cockerell were imaged using propagation phase contrast X-ray synchrotron microtomography (afterwards PPC-SRµCT) at the Advanced Photon Source in Argonne National Laboratory (Chicago, IL). The specimens were scanned following the protocols described in Soriano et al. (2010) with a monochromatic beam at an energy of 25 keV and 250 mm of distance between the camera and the sample. Scan acquisition consisted of 1800 projections over 180°, with 0.3 s of exposure time, and voxel size of 1.45 µm. After the scans, the slices were reconstructed using a filtered back-projection algorithm adapted for local tomography applications (Tomopy software, APS, Gürsoy et al., 2014). The later three-dimensional processing was made using VGStudioMax 2.2 software (Volume Graphics; Heidelberg, Germany). The scans of the two tribes allow for a more thorough characterization of these lineages, including the first data on the male terminalia for fossil Ctenoplectrellini, and a more extensive comparison with the Fushun megachiline. Unfortunately, attempts to scan two of the Fushun bees produced less than ideal results, with poor clarity obtained between the bees and the matrix. Nonetheless, some minimal conclusions could be drawn although it must be hoped that in the future improved technologies might better resolve the fine density differential between insect and Fushun resin.

SYSTEMATIC PALEONTOLOGY

Clade Anthophila Latreille

  • Within the superfamily Apoidea, Anthophila comprise the bees, with over 20,500 described extant species (Michener, 2007). The clade Tanyglossata (Engel, 2004) comprises the long-tongue bees, which in current estimations are sister to a monophyletic Cryptocoxifera (=Andrenidae, Halictidae, Colletidae + Stenotritidae). Cryptocoxifera and Tanyglossata together constitute the clade Zanthophila, “very flower loving” (from Greek ζά, “very,” ἂνθος “flower,”and φίλος, “loving”). Most bee fossils are of the Zanthophila, and particularly the Tanyglossata.

  • Clade Tanyglossata Engel
    Family Megachilidae Latreille
    Subfamily Megachilinae Latreille

  • The subfamily Megachilinae comprises over 4000 extant species of bees, including the familiar leafcutter bees, orchard bees, woolcarder bees, and resin bees, among many others (Michener, 2007). Considerable phylogenetic work has meant that the classification of Megachilinae has undergone significant revision and improvements over the past decade or so (e.g., Praz et al., 2008; Gonzalez et al., 2012, 2019; Litman et al., 2016; Trunz et al., 2016; Parizotto et al., 2022), particularly in relation to the circumscription and composition of the constituent tribes and subtribes (table 1). This is extended here to the circumscription of the earliest tribes of Megachilinae based on new data from the tribes Glyptapini (figs. 27) and Ctenoplectrellini (figs. 812). The current hierarchical classification is summarized in table 1.

  • One of the more interesting conclusions of these phylogenetic explorations is that several genera of early-diverging Megachilinae are noteworthy for the presence of a metatibial scopa (figs. 2A, 3A, C, 4A–D, 7, 8A) in addition to the typical metasomal scopa of all other Megachilidae. Most of these genera are known only from the Paleogene and represent the tribes Ctenoplectrellini and Glyptapini, while one is extant in southern Africa. In some analyses these tribes form a grade at the base of Megachilinae (Gonzalez et al., 2012), suggesting that the presence of the metatibial scopa is shared symplesiomorphically and is ancestral for the subfamily, and perhaps even for a more inclusive clade as similar metatibial scopae are present in Pararhophitinae and seemingly Protolithurgini. Alternatively, other analyses suggest that they form a monophyletic group, along with the cleptoparasitic Dioxyini, which could imply that the metatibial scopa is derived among them (Gonzalez et al., 2019), with its absence a presumed apomorphic loss associated with cleptoparasitism in Dioxyini. The latter analyses also indicate that Aspidosmiini should be placed as a junior synonym under Ctenoplectrellini (Gonzalez et al., 2019), as was also alluded to by Gonzalez et al. (2012). It is also worth noting that all these tribes, with the exception of Dioxyini, have the basal vein arched, with 1M and 1Cu orthogonal to each other at their bases, and that this is another putative apomorphy.

  • These early-diverging tribes of Megachilinae are briefly discussed here to place into context a newly discovered tribe (figs. 1319) from the Fushun coalfield. A revised key is provided to the tribes of Megachilinae currently known from the Eocene.

  • TABLE 1

    Hierarchical Suprageneric Classification of Family Megachilidae

    img-A5K_01.gif

    FIGURE 2.

    Photographs of holotype (IMGP K72) female of Glyptapis mirabilis Cockerell in the Baltic amber, collection of the Geowissenschaftliches Museum, Institut für Geologie und Paläontologie, Georg-August-Universität Göttingen. Scale bar = 1 mm. A. Ventral habitus. B. Dorsal habitus.

    img-z8-1_01.jpg

    FIGURE 3.

    Synchrotron-radiation CT-scan reconstruction of Glyptapis nr. fuscula Cockerell (note that the fine ocular setae and other thin setae or minute branches are beyond the resolution of the CT scan). A. Ventral view. B. Left lateral view. C. Dorsal view. D. Posterior view.

    img-z9-1_01.jpg

    Key to Eocene Tribes of Megachilinae

    1. Basal area and lateral surface of propodeum areolate (figs. 2B, 3B, 3D, 4B, 4D, 4E); mandible with apical tooth and large edentate preapical margin (figs. 5I–L) 2

    —. Basal area and lateral surface of propodeum imbricate to punctate, never areolate (e.g., fig. 4A); mandible multidentate (tridentate in females, bidentate to quadridentate in males) [Baltic, Bitterfeld, Rovno, Oise; five extinct genera, 1 extant genus (southern Africa)] Ctenoplectrellini Engel

    2. Mesoscutellum rounded, not projecting posteriorly over metanotum or propodeum (figs. 2B, 3B, D, 4B, D, E); compound eye hirsute (fig. 2A); metatibial spurs slender (fig. 4A, D); metanotum simple (figs. 2B, 4E); mesoscutum and mesoscutellum with large, craterlike punctures (figs. 2B, 3B–D, 4B–E) [Baltic, Bitterfeld, Rovno; one extinct genus, Glyptapis Cockerell] Glyptapini Cockerell

    —. Mesoscutellum flat, extending posteriorly as shield over metanotum and basal area of propodeum (figs. 15, 16A); compound eye bare (fig. 14A); metatibial spurs thickened (fig. 18B); metanotum bituberculate medially; mesoscutum and mesoscutellum with dense, strong, small punctures (fig. 15) [Fushun; one extinct genus, Glyptosmia, n. gen.] Glyptosmiini Engel, n. tribe

    Tribe Glyptapini Cockerell

  • The tribe Glyptapini currently comprises a single genus, Glyptapis Cockerell (figs. 2, 3). When the genus was first described Cockerell wrote (1909a: 13), “So far as I can judge, Glyptapis and Ctenoplectrella stand near the stem-form of the Megachilidae, but so remote from the modern members of that group that they at least form a distinct subfamily, Glyptapinae. Their nearest relative in the modern fauna appears to be Ctenoplectra.” Cockerell was prescient in his conclusion that Glyptapis and Ctenoplectrella were close to Megachilidae, although his comment regarding the unrelated Ctenoplectra Smith misdirected subsequent researchers for nearly a century. As we understand Ctenoplectra today, the genus is a specialized group of Apidae, until recently classified as the apine tribe Ctenoplectrini (Michener, 2007), and currently constituting with Tetrapediini the subfamily Ctenoplectrinae (Engel et al., 2021b), which is considered related to Xylocopinae (Bossert et al., 2019). Ctenoplectra has followed a tortuous path to find its current resting place next to the xylocopines, and Glyptapis and Ctenoplectrella were dragged through this journey merely by association. Cockerell (1920) established a family Ctenoplectridae, which he felt was close to Macropis Panzer, thereby associating the group with more modern concepts of Melittidae. Michener (1944) followed suit and placed Ctenoplectra as a subfamily of Melittidae, although made no mention of the putative fossils. Based on nothing more than the offhand note of an affinity between these fossil genera and Ctenoplectra, Zeuner and Manning (1976) asserted both genera to be melittids, a similar action followed by Michener and Greenberg (1980) when returning ctenoplectrines to family status. One of us (M.S.E.) similarly followed this precedent when discussing the fossil history of Megachilidae (Engel, 1999a), although this was before extensive examination of material for the genus. Eventually, Ctenoplectra was found to be nested within the Apidae (Roig-Alsina and Michener, 1993; Silveira, 1993), and the first suspicions of a misconnect between the fossils and Ctenoplectra were alluded to when Roig-Alsina and Michener (1993: 160) wrote that fossil “forms assigned to the Ctenoplectrini may be misplaced.” Engel (2001) was the first to recognize that Glyptapis and Ctenoplectrella were early diverging groups of Megachilinae, bringing the journey nearly full circle and echoing Cockerell (1909a). Engel (2001) initially placed them as independent subtribes in an expanded Osmiini before ultimately recognizing that they should be removed as distinct tribes (Engel, 2005; Engel and Perkovsky, 2006). In fact, their removal was the beginning of a progressive narrowing of the circumscription of Osmiini to achieve monophyly, eventually achieved through the later removal of Aspidosmia Brauns, Noteriades Cockerell, Ochreriades Mavromoustakis, Afroheriades Peters, and Pseudoheriades Peters (e.g., Michener, 2007; Praz et al., 2008; Gonzalez et al., 2012, 2019).

  • Perhaps the most immediately distinctive features of Glyptapis are the craterlike punctures of the mesosoma (figs. 2B, 3, 4), the areolate propodeum (figs. 2B, 3, 4), the hirsute compound eyes (fig. 2A), and the large edentate margin to the mandible (fig. 5I–J). These characters certainly make it possible to easily recognize the bees among fossil Apoidea, but there are numerous other interesting combinations of traits present in the groups, beyond just the aforementioned metatibial scopa. Continued study of these bees over the past 20 years and the availability of the scans reproduced here permit a more thorough and corrected diagnosis for the tribe.

  • Diagnosis: Integument apparently without maculation and nonmetallic (integumental coloration poorly preserved in available fossils).

  • Female Mandible with single apical tooth and long edentate margin apically composed of expanded trimmal extension (fig. 5I–J); fimbrial ridge present on mesal surface and paralleling apical edentate margin (fig. 5I); interdental laminae lacking (fig. 5I, K). Malar space linear (fig. 5A–C). Maxillary palpus tetramerous. Labial palpus tetramerous (fig. 5H); palpomeres I and II flattened, elongate, palpomere I about 0.5× length of palpomere II; palpomere II with several erect stiff setae ventrally; palpomere III projecting obliquely from axis of palpomere II; ventral (posterior) surface of glossa with longitudinal groove, forming a glossal canal with inner longitudinal ridge and bordered by annulate, ectal surfaces of glossa (fig. 5G). Clypeus somewhat flattened, extending below lower tangent of compound eyes (fig. 5A–C); clypeus apical margin covering labral base, thus labroclypeal articulation obscured (fig. 5A–C). Single subantennal sulcus directed to outer margin of antennal torulus (fig. 5A). Juxtantennal carina absent (fig. 5A). Compound eyes hirsute. Posterior margin of vertex gently concave; preoccipital area sharply angled above but not carinate, otherwise rounded (fig. 5E, F).

  • Pronotal collar virtually absent, with posterior border blending uninterrupted onto dorsal surface without transverse ridge medially (in profile slope continuous and without anterior, transverse ridge demarcating dorsal-facing surface comprising collar relative to lower anterior neck), with transverse dorsal ridge carinate laterally (effaced medially) and extending across pronotal lobe without dorsolateral angle (fig. 4B) (dorsolateral angle present in Dioxyini except Prodioxys Friese), without dorsolateral ridge extending vertically across lateral surface (fig. 4B) (present in Dioxyini).

  • Mesoscutum broadly rounded anteriorly (figs. 3C, 4C), raised above pronotal posterior margin, with deep craterlike punctures/foveae (figs. 2B, 3B, C, 4B–E) (also present on mesoscutellum and upper mesepisternum); parapsidal lines linear; preaxilla vertical and asetose (as in Dioxyini and Anthidiini); mesoscutalmesoscutellar sulcus deeply impressed, with small, medial, V-shaped notch on mesoscutellum; mesoscutellum low, rounded, posterior margin narrowly vertical with transverse row of small areolae, posterior margin not extending over metanotum; axillae simple (figs. 2B, 3D, 4E); metanotum sloping, without medial tubercles or spines (metanotum tuberculate or spinous in Dioxyini); basal area of propodeum sloping, asetose, areolate (figs. 2B, 3D, 4B, D, E); omaular ridge carinate (fig. 4B); scrobal sulcus absent; preepisternal sulcus absent; mesepisternum areolate (fig. 4B); dorsal lamella of metepisternum absent.

  • Forewing (refer to Engel, 2001) with pterostigma proximal to r-rs as long as or slightly longer than pterostigmal width; pterostigmal length 2× or more its basal width; prestigma more than 2× as long as broad; basal vein (1M) arched, thus orthogonal to 1Cu (as in Ctenoplectrellini and the new tribe described below); marginal cell apex acutely rounded on anterior wing margin; two submarginal cells; 1m-cu and 2m-cu entering second submarginal cell, i.e., 2rs-m proximal to 2rs-m; hind wing with six distal hamuli.

  • Protibial calcar (fig. 6) with anterior ridge on rachis bordering velum (fig. 6D) (as in Anthidiini), malus simple, short (fig. 6D), less than 0.5× length of velum, apical margin of velum simple (rachis and apex of velum with serrations in Anthidiini and Osmiini); tibiae not spiculate; pro- and mesotibiae with apical outer spine (fig. 6A), spine fainter on protibia; mesotibial spur long, slender, minutely ciliate (fig. 6); metabasitibial plate absent; metatibial scopa present (fig. 7) (as in Ctenoplectrellini and the new tribe described below; absent in other megachiline tribes); metatibial spurs long, slender, minutely serrate to minutely ciliate (fig. 7); pretarsal claws with short inner ramus (figs. 6, 7D); arolium present (figs. 6, 7).

  • Metasomal scopa present; sting and associated structures well developed (vestigial in Dioxyini).

  • Male. Unknown.

  • Remarks: It is worth noting here a clarification regarding the publication and date from which the name Glyptapis and that of its type species, Glyptapis mirabilis Cockerell, were first made nomenclaturally available. More than 20 years ago, one of us (M.S.E.) followed the assertion of Zeuner and Manning (1976) that the first usage of the names Glyptapis and G. mirabilis were in Cockerell's (1909b) paper in The Entomologist, and his more extensive descriptions of the genus, its type species, and the remaining species of the genus appeared subsequently in Schriften der physikalischökonomischen Gesellschaft zu Königsberg (Cockerell, 1909a). However, examination of the prefatory material to volume 50 of the Schriften reveals at the end of the table of contents an outline of the dates of publication for the various numbers constituting the volume. Cockerell's (1909a) article was published 20 September 1909, while the article that Zeuner and Manning (1976) considered as the taxon name's first usage—erroneously it turns out—was not released until December 1909. Thus, the genus Glyptapis and its type species were made available in the article Cockerell intended (Cockerell, 1909a) and his usage of the names later that year in The Entomologist was merely a subsequent mention rather than the place from which the names were made available. Why Zeuner and Manning (1976) made such a reversal of precedence for the names is unclear but wrong based on currently available evidence. The same order of precedence applies to the name Ctenoplectrella.

  • FIGURE 4.

    Synchrotron-radiation CT-scan reconstruction of Glyptapis nr. fuscula Cockerell, with wings removed (note that the fine ocular setae and other thin setae or minute branches are beyond the resolution of the CT scan). A. Ventral view. B. Left lateral view. C. Dorsal view. D. Right lateral view. E. Posterior view.

    img-z12-1_01.jpg

    FIGURE 5.

    Synchrotron-radiation CT-scan reconstruction of Glyptapis nr. fuscula Cockerell (note that the fine ocular setae and other thin setae or minute branches are beyond the resolution of the CT scan). A. Facial view (pink line = right subantennal sulcus; cyan line = right half epistomal sulcus). B. Right lateral view of head. C. Left lateral view of head. D. Ventral view of head. E. Posterior view of head. F. Dorsal view of head. G. Apical portion of glossa, ventral view. H. Labial palpi, ventral view. I. Right mandible, mesal (inner) view. J. Right mandible, dorsal view. K. Left mandible, lateral (outer) view. L. Left mandible, dorsal view.

    img-z13-1_01.jpg

    FIGURE 6.

    Synchrotron-radiation CT-scan reconstruction of fore- (A–D) and mid legs (E–H) of Glyptapis nr. fuscula Cockerell. A. Foreleg, prolateral (anterior) view. B. Foreleg, dorsal view. C. Foreleg, ventral view. D. Foreleg, retrolateral (posterior) view, with magnification detail of antenna cleaner. E. Midleg, Dorsal view. F. Midleg, prolateral (anterior) view. G. Midleg, retrolateral (posterior) view. H. Midleg, ventral view.

    img-z14-1_01.jpg

    FIGURE 7.

    Synchrotron-radiation CT-scan reconstruction of hindlegs (E–H) of Glyptapis nr. fuscula Cockerell. A. Retrolateral (posterior) view. B. Ventral view. C. Dorsal view. D. Prolateral (anterior view).

    img-z15-1_01.jpg

    FIGURE 8.

    Photographs of representative Ctenoplectrellini. A. Oblique lateral habitus of neotype (AMNH B-JH 95) female of Ctenoplectrella viridiceps Cockerell in Baltic amber (courtesy D.A. Grimaldi and A. Pierwola). B. Dorsal habitus of holotype (MNHN PA 3190 1/17) female of C . eocenica (Michez and Nel) in Oise amber (courtesy O. Béthoux).

    img-z16-1_01.jpg

    Tribe Ctenoplectrellini Engel

  • The tribe Ctenoplectrellini includes superficially osmiine-like bees, so much so that the group was originally assigned as a subtribe of Osmiini (Engel, 2001) before elevation to a distinct tribe (Engel, 2005; Engel and Perkovsky, 2006). Ctenoplectrellines have been found as amber inclusions from the Baltic (including Bitterfeld), Rovno, and Oise deposits (Cockerell, 1909a; Engel, 2001, 2008; Engel and Perkovsky, 2006; Gonzalez and Engel, 2011; Engel and Davis, 2021), and as compression fossils from Messel in Germany and the Green River Formation of Colorado, all from the Eocene (Wedmann et al., 2009; M.S.E., unpubl. data). There are also compressions of Ctenoplectrellini from the Paleocene of Menat, France (Nel and Petrulevičius, 2003; M.S.E., unpubl. data). Subsequently, the extant sub-Saharan Aspidosmia was discovered to be nested among the ctenoplectrelline genera (Gonzalez et al., 2012, 2019), confirming the prediction of Gonzalez and Engel (2011) and documenting a surviving relict of this clade hitherto known only from Eocene occurrences. The tribe Aspidosmiini was established for Aspidosmia and separate from the Ctenoplectrellini (Gonzalez et al., 2012), although its recognition renders the latter paraphyletic and obscures the significance of the relationship of Aspidosmia and the various fossil genera (Gonzalez et al., 2019). In fact, Gonzalez and Engel (2011) noted the shared characters between Aspidosmia and Ctenoplectrella, suggesting that the former could be a surviving relict of ctenoplectrellines, a conclusion later supported by the analysis of Gonzalez et al. (2019). The linkage between Aspidosmia and Ctenoplectrella is further emphasized by the male terminalia, with both genera having a characteristically trilobate apical margin to hidden sternum VII (see below). Accordingly, the tribe Aspidosmiini is considered a junior synonym of Ctenoplectrellini (new synonymy), thereby removing the paraphyly of the group.

  • Diagnosis: Integument with or without maculation (maculation present on face of males of Aspidosmia) and either nonmetallic or weakly dark metallic green (fig. 8).

  • Female Mandible tridentate; frimbrial ridge absent (known for Ctenoplectrella and Aspidosmia); interdental laminae lacking. Malar space linear (figs. 8A, 9B, C, 10A, C). Maxillary palpus tetramerous. Labial palpus tetramerous; palpomeres I and II flattened, elongate, palpomere I 0.5–0.9× length of palpomere II; palpomere III projecting obliquely from axis of palpomere II. Clypeus somewhat flattened, not or only slightly extending below lower tangent of compound eyes (figs. 8A, 9C, 10A); clypeus apical margin covering (figs. 8A, 10A) or not convering labral base (Aspidosmia), thus labroclypeal articulation either obscured or not (variable among genera, refer to key, below). Single subantennal sulcus directed to outer margin of antennal torulus (fig. 10A). Juxtantennal carina absent. Compound eyes bare (fig. 8A). Posterior margin of vertex gently concave; preoccipital ridge rounded, sometimes sharply angled dorsally.

  • Pronotal collar virtually absent, with posterior border blending uninterrupted onto dorsal surface without ridge medially (in profile slope continuous and without anterior ridge demarcating dorsal-facing surface comprising collar relative to lower anterior neck), with dorsal transverse ridge effaced or rounded, without dorsolateral angle, without dorsolateral ridge extending vertically across lateral surface.

  • Mesoscutum broadly rounded anteriorly (fig. 8B, 9A), raised above pronotal posterior margin, sculpturing variable, generally smooth to imbricate with scattered small to minute punctures (fig. 9A); parapsidal lines linear; preaxilla sloping with setae (as in Megachilini); mesoscutal-mesoscutellar sulcus impressed, without V-shaped notch on mesoscutellum; mesoscutellum low, rounded, posterior margin meeting metanotum and not extending over metanotum (fig. 9); axillae simple (figs. 8B, 9); metanotum sloping, without medial tubercles or spines; basal area of propodeum sloping, asetose, smooth to faintly imbricate (fig. 9); omaular ridge rounded (fig. 9B, D); scrobal sulcus absent; preepisternal sulcus absent; mesepisternum smooth to faintly imbricate, sometimes with scattered punctures (fig. 9B, D); dorsal lamella of metepisternum absent.

  • Forewing (fig. 8B; refer also to Engel, 2001; Gonzalez and Engel, 2011) with pterostigma proximal to r-rs about as long as or slightly longer than pterostigmal width; pterostigmal length 1.75× or more its basal width; prestigma more than 2× as long as broad; basal vein (1M) arched, thus orthogonal to 1Cu (as in Glyptapini and the new tribe described below); marginal cell apex rounded, slightly offset from anterior wing margin; two submarginal cells; 1m-cu and 2m-cu entering second submarginal cell, i.e., 2rs-m proximal to 2rs-m; hind wing with 6–13 distal hamuli (6–7 hamuli in Ctenoplectrella, 7 in Aspidosmiella, 10 in Glaesosmia, 11–13 in Aspidosmia).

  • Protibial calcar with or without anterior ridge on rachis bordering velum, malus simple, short, less than 0.5× length of velum, apical margin of velum simple; tibiae not spiculate; pro- and mesotibiae with apical outer spine, spine less prominent on protibia; mesotibial spur long, slender, minutely ciliate to serrate; metabasitibial plate absent; metatibial scopa present (fig. 8A) (as in Glyptapini and the new tribe described below; absent in other megachiline tribes); metatibial spurs long, slender (fig. 8A), minutely serrate to minutely ciliate; pretarsal claws with short inner ramus; arolium present.

  • Metasomal scopa present; sting and associated structures well developed (vestigial in Dioxyini).

  • MALE (known only for Aspidosmia and Ctenoplectrella). Mandible bidentate (fig. 9C, 10A) to quadridentate. Metasomal tergum VI without preapical carina; tergum VII exposed, directed posteriorly, without pygidial plate (fig. 11B–D); sternum VI with apical margin simple (Aspidosmia) or with lateral notches for reception of gonostyli (fig. 11F) (Ctenoplectrella); sternum VII well sclerotized, not divided medially, apical margin trilobate (fig. 11G) (medial lobe large in Ctenoplectrella: fig. 11G; medial lobe formed as short point in Aspidosmia: refer to Peters, 1972); volsella present (fig. 12) (particularly large in Ctenoplectrella), with or without differentiated digitus and cuspis (differentiated in Ctenoplectrella: fig. 12A; not differentiated in Aspidosmia).

  • Remarks: A new key is provided to the genera of living and fossil Ctenoplectrellini. The subtribe Glaesosmiina Engel (new subtribe; type genus: Glaesosmia Engel, 2001; diagnosed by combination of subantennal sulcus as long as torular diameter, interocular distance greater than compound eye length, and forewing 2Rs oblique to 2M, forming acute apical angle) is here recognized to remove paraphyly of Ctenoplectrellina relative to Aspidosmiina and allow for retaining a group for Aspidosmia and its visible labroclypeal articulation and shorter pterostigmal form relative to the fossils. In addition, Ctenoplectrella phaeton Gonzalez and Engel is removed to a separate genus given its peculiarly sparse pubescence and rather discomfited inclusion in Ctenoplectrella.

  • FIGURE 9.

    Synchrotron-radiation CT-scan reconstruction of male of Ctenoplectrella viridiceps Cockerell. A. Dorsal view. B. Left lateral view. C. Ventral view. D. Right lateral view.

    img-z18-1_01.jpg

    FIGURE 10.

    Synchrotron-radiation CT-scan reconstruction of head of male of Ctenoplectrella viridiceps Cockerell. A. Facial view. B. Dorsal view. C. Right lateral view. D. Posterior view. E. Ventral view.

    img-z19-1_01.jpg

    Key to Genera of Ctenoplectrellini

    1. Clypeus apical margin covering labral base, thus labroclypeal articulation obscured in facial view; pterostigmal length 2× or more basal width 2

    —. Clypeus apical margin not covering labral base, thus labroclypeal articulation not hidden; pterostigmal length less than 2× as long as basal width [subtribe Aspidsomiina] Aspidosmia Brauns

    2(1).Forewing 2Rs oblique to 2M, forming acute apical angle; subantennal sulcus as long as torular diameter; interocular distance greater than compound eye length [subtribe Glaesosmiina Engel, n. subtribe] 3

    —. Forewing 2Rs orthogonal to 2M; subantennal sulcus longer than torular diameter; interocular distance subequal to or shorter than compound eye length [subtribe Ctenoplectrellina] 4

    3(2).Gena broader than compound eye; pterostigma length in first submarginal cell from base to r-rs as long as prestigma Probombus Piton

    —. Gena as broad as compound eye; pterostigma length in first submarginal cell from base to r-rs longer than prestigma Glaesosmia Engel

    4(2).Pronotal collar virtually absent 5

    —. Pronotal collar short but distinct Friccomelissa Wedmann, Wappler, and Engel

    5(4).Body pubescent; mesoscutum and mesoscutellum with numerous, short, erect to suberect, fine setae, setae rarely separated by more than a setal length; metasomal scopa composed of dense bands of setae. Ctenoplectrella Cockerell

    —. Body distinctly sparsely pubescent; mesoscutum and mesoscutellum sparsely pubescent, setae, when present, largely separated by many times setal lengths; metasomal scopa formed of sparse bands of setae Aspidosmiella Engel, n. gen.

    FIGURE 11.

    Synchrotron-radiation CT-scan reconstruction of male terminalia of Ctenoplectrella viridiceps Cockerell. A. Apical view of metasoma as preserved. B. Tergum VII, dorsal oblique view. C. Tergum VII, right lateral view. D. Tergum VII, apical view. E. Terga VII and VIII, ventral view. F. Sternum VI, ventral view. G. Sternum VII. H. Sternum VIII.

    img-z21-1_01.jpg

    Aspidosmiella Engel, new genus

  • Type species: Ctenoplectrella phaeton Gonzalez and Engel, 2011.

  • Diagnosis: This genus is generally as described for Ctenoplectrella: tridentate female mandible, pronotal collar virtually absent, and 2Rs orthogonal to 2M but is more robust and with a noticeably sparse pubescence, especially on the mesoscutum and mesoscutellum where the surfaces are nearly asetose and the metasomal scopa composed of sparse bands of setae (refer to Gonzalez and Engel, 2011).

  • Etymology: The new generic name is a combination of Aspidosmia and the Latin suffix -ellus, indicating a diminutive. The gender of the name is feminine.

  • Glyptosmiini Engel, new tribe

  • Type genus: Glyptosmia Engel, new genus.

  • Diagnosis: This tribe possesses a unique intermingling of features of Glyptapini, Dioxyini, and Ctenoplectrellini as well as its own distinctive apomorphies. Like Glyptapini and Ctenoplectrellini the metatibia bears a scopa composed of abundant, long setae (fig. 13), although in this tribe many of the setae on the posterior surface are characteristically hooked apically, and some of those on the anterior surface have short apical branches. Likewise, the pretarsal claws have a short, subapical ramus, arolia on all legs, the basal vein is arched, and 2m-cu is slightly basal to 2rs-m, and thus both 1m-cu and 2m-cu enter the second submarginal cell. Like Glyptapini the basal area and lateral surface of the propodeum are areolate and the mandible is long, slender, and has a large edentate margin above the apical tooth. Unlike glyptapines the compound eyes are bare, the omaular ridge is not carinate, and the mesosoma lacks large craterlike punctures and is instead covered with dense, strong, small punctures. Similar to Dioxyini the metanotum is medially beset with tubercles, albeit paired in the fossil versus the single tubercle or spine in the former tribe. Unlike Dioxyini, the sting and its associated structures are well developed. Quite distinctively, the mesoscutellum is flattened and projected posteriorly over the metanotum and the basal edge of the propodeum, and the apical margin is sinuate, and the metatibial spurs are noticeably thickened.

  • Female. Integument without maculation and shiny black (fig. 13). Mandible with single apical tooth and edentate margin apical to expansion of trimmal extension; interdental laminae lacking. Malar space linear. Maxillary palpus tetramerous. Labial palpus tetramerous; palpomeres I and II flattened, elongate, palpomere I subequal to length of palpomere II; palpomere III projecting obliquely from axis of palpomere II. Clypeus somewhat flattened, not extending below lower tangent of compound eyes; clypeus apical margin not covering labral base, thus labroclypeal articulation not obscured. Single subantennal sulcus directed to outer margin of antennal torulus. Juxtantennal carina absent. Compound eyes bare (fig. 14). Posterior margin of vertex gently concave; preoccipital area sharply angled above but not carinate, otherwise rounded.

  • Pronotal collar virtually absent, posterior border blending uninterrupted onto dorsal surface without transverse ridge medially to demarcate collar from dorsal surface of neck, with dorsal ridge rounded laterally without dorsolateral angle (carinate laterally in Glyptapini), without dorsolateral ridge extending vertically across lateral surface. Mesoscutum broadly rounded anteriorly (fig. 13A, 15), raised above pronotal posterior margin, with dense, small punctures (without the deep craterlike punctures of Glyptapini); parapsidal lines linear; mesoscutalmesoscutellar sulcus deeply impressed and wide (fig. 15), without V-shaped notch on mesoscutellum; mesoscutellum flat (figs. 15, 16A), posterior margin extending over metanotum and basal edge of propodeum (fig. 16A), apical margin with broad medial concavity (fig. 15); axillae simple (fig. 15); metanotum vertical, with pair of medial tubercles on either side of midline (fig. 16A); basal area of propodeum horizontal, asetose, areolate (as in Glyptapini); omaular ridge angled but not carinate (carinate in Glyptapini); scrobal sulcus absent; preepisternal sulcus absent; mesepisternum with dense, small punctures (areolate in Glyptapini); dorsal lamella of metepisternum absent.

  • Forewing (fig. 17) with pterostigma proximal to r-rs longer than pterostigmal width; pterostigmal length more than 2× its basal width; prestigma more than 2× as long as broad; basal vein (1M) arched, thus orthogonal to 1Cu (as in Glyptapini and Ctenoplectrellini); marginal cell apex rounded, slightly offset from anterior wing margin; two submarginal cells; 1m-cu and 2m-cu entering second submarginal cell, i.e., 2rs-m proximal to 2rs-m; hind wing with six distal hamuli.

  • Protibial calcar with anterior ridge on rachis bordering velum (as in Anthidiini and Glyptapini), malus simple, short, less than 0.5× length of velum, apical margin of velum simple; tibiae not spiculate; pro- and mesotibiae with apical outer spine, spine fainter on protibia; mesotibial spur long, slender, minutely serrate; metabasitibial plate absent; metatibial scopa present (figs. 13B, 18A) (as in Glyptapini and Ctenoplectrellini; absent in other megachiline tribes); metatibial spurs long, distinctly thickened (fig. 18B), minutely serrate (not crescentic, ciliate, nor comb-like); pretarsal claws with short inner ramus (fig. 18C, D); arolium present (fig. 18C).

  • Metasomal scopa present; sting and associated structures well developed (fig. 19).

  • Male. Unknown.

  • FIGURE 12.

    Synchrotron-radiation CT-scan reconstruction of male genitalia of Ctenoplectrella viridiceps Cockerell. A. Genital capsule, ventral view. B. Genital capsule, dorsal view. C. Genital capsule, apical view. D. Genital capsule, right lateral view.

    img-z23-1_01.jpg

    Glyptosmia Engel, new genus

  • Type species: Glyptosmia hemiaspis Engel, new species.

  • Diagnosis: As for the tribe (above).

  • Etymology: The new genus-group name is a combination of ancient Greek adjective γλυπτός (glyptos, meaning “engraved”) and Osmia Panzer (derived from ỏσµή/osme, meaning “odor”). The gender of the name is feminine.

  • Glyptosmia hemiaspis Engel, new species
    Figures 1319

  • Diagnosis: As for the genus (above).

  • Description: FEMALE. Total body length (as preserved) 4.36 mm; forewing length (based on the left forewing as the right wing is taphonomically stretched) 2.27 mm. Integument black, shining (fig. 13); wing veins dark brown, membranes hyaline and clear (fig. 17). Head slightly longer than wide, length 1.03 mm, width 0.81 mm; labrum about as long as wide, with truncate apical margin; mandible elongate, with distinct apical tooth and larger preapical edentate margin (apex similar to Glyptapis except mandible more elongate); malar space linear; maxillary palpus tetramerous; labial palpomeres I and II long and flattened, albeit each shorter than prementum; clypeus flat, not extending below lower tangent of compound eyes; single subantennal sulcus straight to lower outer angle of torulus (visible faintly from right side; base of sulcus faint on left side), sulcus slightly longer than torular diameter; compound eye bare, slightly broader than gena in profile, inner margins straight and slightly converging below; upper interorbital distance greater than lower interorbital distance; scape short, length 0.18 mm, shorter than torulocellar distance; flagellomere I shorter than combined lengths of flagellomeres II and III (fig. 14B), flagellomeres II and III of equal lengths; preoccipital ridge sharply angled dorsally, but not carinate, rounded laterally. Mesoscutum length 0.88 mm, mesoscutellum length 0.30 mm; intertegular distance 0.85 mm; parapsidal lines short, linear; transverse mesoscutal-mesoscutellar sulcus deeply impressed and wide, seemingly weakly foveolate within depression (challenging to observe as it is difficult to illuminate properly); tegula oval, brown and semitranslucent, with weak punctures. Forewing (fig. 17) with basal vein (1M) noticeably arched, orthogonal to 1Cu; hind wing with six distal hamuli arranged in a single series; jugal lobe more than 0.5× but less than 0.65× length of anal lobe (challenging to see precise measurement).

  • Labrum, clypeus, and supraclypeal area with dense, small punctures, punctures separated by much less than a puncture width, nearly contiguous in some places, integument between punctures apparently smooth and shining. Lower face punctured as on clypeus, frons with similar punctures but more spaced (fig. 14), separated by 0.5–1.0× a puncture width, although those on central frons denser, integument between punctures smooth and shining; ocellocular area (fig. 14A) and vertex with punctures separated by 0.5–1.2× a puncture width; gena apparently with punctures as on vertex. Mesosoma with dense punctures (figs. 13A, 15), punctures coarser and slightly larger than those of head, punctures nearly contiguous on mesoscutum, mesoscutellum, and pleura, those of pleura seemingly slightly larger than those of mesoscutum (challenging to see clearly given poor state of preservation), integument between punctures smooth and shining. Metanotum irregularly roughened with two rounded tubercles on either side of midline. Basal area of propodeum areolate, with broad seemingly impunctate triangular area on upper posterior surface. Metasomal terga II–V with pregradular surfaces depressed and polished relative to disc and apical marginal zones, pregradular areas smooth to finely imbricate with some scattered, shallow, small punctures apically; tergal discs imbricate with small punctures separated by a puncture width (fig. 16B), although laterally punctures denser, anterior-facing surface of tergum I with punctures much sparser and integument between smooth (fig. 16B).

  • Head (fig. 14) with numerous, short, fine, white, erect to suberect, minutely branched setae, such setae not obscuring integument, setae denser in antennal basin and supraantennal area; such setae present on labrum, labrum without distinct patches of setae; clypeus and supraclypeal area with sparsely scattered short, simple setae; frons with scattered, short, simple to minutely branched setae, such setae denser in antennal basins and becoming sparser in ocellocular area and vertex. Mesoscutum with scattered, short, simple, erect, pale setae, such setae slightly longer on mesoscutellar surface except setae along margin of axilla and mesoscutellum denser, long to elongate, lightly fuscous, and minutely branched; pleura with scattered, short, erect, largely simple setae, such setae slightly longer posteriorly and on lateral surface of propodeum; basal area of propodeum without setae. Legs with short, simple or minutely branched, pale, suberect setae except mesotibia with such setae more numerous, slightly longer, and intermixed with some thicker setae, setae on inner surface distinctly longer; mesobasitarsus with similar setae except noticeably longer than those of mesotibial outer surface; metatibia with scopa composed of abundant elongate erect setae, those of outer surface minutely branched apically or some simple, those of inner surface as on outer surface except intermixed with numerous elongate, simple, characteristically hooked setae; metabasitarsus with abundant erect to suberect, elongate, minutely branched, pale setae; meso- and metafemora with sparse, short, largely simple to minutely branched setae, such setae especially minute and simple on ventral and posterior apical surfaces; meso- and metabasitarsi with setae similar to those of corresponding tibiae. Metasomal terga with scattered, short, fine, decumbent, pale setae, such setae sparse on anterior-facing surface of tergum I, intermixed with slightly longer, erect to suberect setae, some with minute branches, laterally; tergum II onward with short discal setae intermixed with somewhat more erect setae and slightly denser apically (but not forming setal bands); narrow apical marginal zones without setae, imbricate; tergum VI with setae denser medioapically; metasomal scopa on sterna II–V composed of bands of long simple erect setae; sternum VI without scopal setae, with dense brush of minute plumose fuscous setae apically, apical margin narrowly truncate; sting simple, sting sheaths thick, short, with some apical setae.

  • Male. Unknown. HOLOTYPE: Female, CNU-HYM-LF-2023-002 (fig. 13), in a single piece of amber without syninclusions, Fushun coalfield, Liaoning Province, northeastern China, Guchengzi Formation, Ypresian (Eocene); deposited in the fossil insect collection of the Key Laboratory of Insect Evolution and Environmental Changes, College of Life Sciences, Capital Normal University, Beijing, China.

  • The holotype of Glyptosmia hemiaspis is challenging to interpret given its state of preservation (fig. 13). The bee is preserved with the head slightly pulled forward and slightly detached from mesosoma, and twisted to the right. The metasoma is partially curved ventrally and the legs are more or less pulled inward alongside and partially under the body, thereby obscuring various views of their surfaces depending on the relation of a given podite to the mesosoma, metasoma, and sclerites from other legs. The wings are extended posteriorly and upright, thereby not overlapping the body but making a posterior view of the mesosoma challenging. Most challenging, however, are the presence of several internal fractures across the body and particularly around the head (fig. 14), the head is also obscured by a bubble to its right, which extends posteriorly to envelope some of the podites of the right fore- and midlegs as well as the entire right side of the mesosoma (fig. 13B). The metasoma also has some smaller bubbles ventrally and captured amid the scopal setae and the dorsal surface is partially damaged at the surface (fig. 13B). There is also a frothy spread of microscopic bubbles across part of the face to its right, which is also slightly damaged by a partial compression of the face in the supraantennal basins and perhaps also by the torsion of the head (fig. 14). The mouthparts are partially extended beneath the head. Collectively, these imperfections make this a frustrating specimen. Nonetheless, sufficient characters can be observed to support our proposed classification. Although the bee is obviously a megachilid it is nonetheless worth noting that the presence of a V-shaped postmentum (evident in the CT-scan data) and a fully exposed mesocoxa support its placement among the long-tongued bees. Despite the challenges with the head, the course of the epistomal sulcus can be followed and the left anterior tentorial pit located, with the attachment of the subantennal sulcus seen and the small portion of its course would place it at the outer lower corner of the antennal torulus, a decidedly megachiline trait. Also consistent with a megachilid is the presence of elongate, scopallike setae on the metasomal sterna, and although these are entirely simple and looser than that of most megachilids, the metasomal scopa is much like that of some other Eocene megachiline tribes (e.g., Glyptapini and some Ctenoplectrellini).

  • Etymology: The specific epithet is the ancient Greek prefix ἡµι- (hemi-, meaning “half”) and the noun fi01_01.gif (aspis, meaning “shield”), and refers to the extended mesoscutellum with a sinuate margin superficially calling to mind half of a Boeotian shield.

  • FIGURE 13.

    Photographs of holotype (CNU-HYM-LF-2023-002) female of Glyptosmia hemiaspis, n. gen. and sp., in Fushun amber. A. Left dorsal oblique view. B. Right lateral oblique view.

    img-z25-1_01.jpg

    FIGURE 14.

    Photographs of head of Glyptosmia hemiaspis. A. Oblique facial view from upper right. B. Facial view.

    img-z26-1_01.jpg

    FIGURE 15.

    Oblique dorsal view of mesoscutum and mesoscutellum of Glyptosmia hemiaspis.

    img-z27-1_01.jpg

    FIGURE 16.

    Profiles of Glyptosmia hemiaspis. A. Mesoscutellum, metanotum, and propodeum. B. Metasomal terga I and II and base of tergum III.

    img-z28-1_01.jpg

    TABLE 2

    Species Diversity and Geographic and Temporal Occurrence of the Early Diverging Megachiline Bee Tribes Ctenoplectrellini, Glyptapini, and Glyptosmiini Note that there is also an undescribed ctenoplectrelline species from the Green River Formation (M.S.E., unpubl. data).

    img-ATdt_01.gif

    FIGURE 17.

    Forewing of Glyptosmia hemiaspis.

    img-z30-1_01.jpg

    FIGURE 18.

    Photographs of leg podites of Glyptosmia hemiaspis. A. Dorsal view of metatibia, metatarsus, and metapretarsus. B. Apical view of metatibial spurs (note also hooked scopal setae in background). C. Detail of metapretarsus, dorsal view. D. Detail of metapretarsal claws, lateral view, arrow points to inner (subapical) ramus of left claw.

    img-z31-1_01.jpg

    Family Apidae Latreille
    Subfamily Apinae Latreille
    Clade Corbiculata Engel

  • There are four extant tribes of corbiculate bees, along with an additional three tribes known only from fossils. These are some of the most familiar of all bees (Michener, 2007; Engel and Rasmussen, 2021): orchid bees (Euglossini), bumble bees (Bombini), stingless bees (Meliponini), and honey bees (Apini). Currently available evidence indicates that the three extinct tribes— Electrobombini, Electrapini, and Melikertini— disappeared during the Eocene-Oligocene transition (Engel, 2001).

  • Some assert that the formal name Corbiculata is attributable to Shuckard (1866). In fact, Shuckard (1866) never established such a name and instead classified corbiculate bees as his Section Cenobites Shuckard, 1866 (160, 302); he also renamed all bees as Mellicolligerae Shuckard [= Anthophila Latreille]. In discussing the systematic arrangement of Apidae, Shuckard (1866: 165) wrote about the character uniting the Cenobites being, “the glabrous surface of the posterior tibiae, with their lateral edges fringed with bristles slightly curved inwards... a sort of natural basket for the conveyance of pollen or other stores to the nest.” He continues, “This, however, has not been made use of as a main feature for scientific distribution, although they might follow the Dasygasters [Megachilidae], as corbiculated bees, or little basket bearers.” Thus, Shuckard introduces the term corbicula for this structure and uses the term as an adjectival description of the bees he actually classifies as Cenobites. Accordingly, Shuckard never proposed a “Corbiculata” and the formal name, rather than an adjective or new morphological term, cannot be said to derive from his work.

  • FIGURE 19.

    Photographs of metasomal apex of Glyptosmia hemiaspis. A. Ventral view. B. Left oblique lateral view.

    img-z32-1_01.jpg

    Tribe Melikertini Engel

  • Melikertines are the most commonly encountered bees in Eocene amber and have been recovered from the Baltic (inclusive of Bitterfeld), Cambay, and Fushun amber deposits, thus spanning a rather impressive paleogeographical distribution from at least the Ypresian through Bartonian (Engel and Davis, 2021). A species of melikertine has been discovered recently in Eocene Lublin amber (Celary et al., 2023; J. Szwedo, personal commun.). No melikertines are currently known from after the Eocene-Oligocene transition and none have been discovered in older amber deposits. Likewise, they have yet to be found as compressions or impressions in contemporaneous sedimentary settings, such as the maar lakes of Messel or Eckfeld in Germany where electrapine and ctenoplectrelline bees have been recovered in addition to their more widely known amber inclusions (Wappler and Engel, 2003; Wedmann et al., 2009; Wappler et al., 2015), and alongside early-diverging genera/subgenera of surviving tribes, such as Xylocopini and Bombini (Geier et al., in press; M.S.E., unpubl. data). The restriction of melikertines to amber may be a bias driven by their resin-collecting habits (Engel and Davis, 2021). The tribe currently comprises 15 species in nine genera, inclusive of the species described herein (table 3). Currently available phylogenetic evidence supports melikertines as related to Meliponini, the stingless bees, and that these together are sister to the honeybees, Apini, with the bumble bees more further removed (Engel, 2001; Schultz et al., 2001). Unfortunately, the fossil record of bumble bees is scant and in need of further revision (Wappler et al., 2012; Prokop et al., 2017; Dehon et al., 2019; M.S.E., unpubl. data), while that of Apini is richer but in even greater need of reconsideration (Engel, 1998a, 1999b; Kotthoff et al., 2013; M.S.E., unpubl. data).

  • Melikertines were eusocial, likely living in colonies similar to those of modern stingless bees (Engel, 1998b; Barden and Engel, 2021; Engel and Davis, 2021). For all but one of the species we know only the worker caste and therefore lack data for variations between workers and the other castes. Accordingly, it remains unknown whether some of the exaggerated morphologies observed in workers may also be present in queens or males. One would initially hypothesize that the queens lack these structures (e.g., facial protuberances), since they would seem to be associated with resin collection (Engel and Davis, 2021), although it cannot be ruled out that they may also be deployed in nest construction or less likely in defense (given the morphologies known, none seem suitable as a defensive weapon). An extensive account of the tribe and its diversity and putative paleobiology was recently provided by Engel and Davis (2021) and is therefore not repeated here except to update their key to genera.

  • Key to Genera of Melikertini
    (updated and revised from Engel and Davis, 2021)

    1. Disc of clypeus comparatively flat, without distinct lateral carinae and without belllike concavity 2

    —. Clypeus with lateral carinae rising from apex to form margins of belllike concavity arising from base of clypeus and overhanging disc Aethemelikertes Engel

    2(1). Clypeal protrusion present, i.e., base of clypeus produced into variously modified facial prominences, prominence bending upward over fronto-clypeal portion of epistomal sulcus and obscuring supraclypeal area or even lowermost frons in facial view 3

    —. Clypeal protrusion absent 4

    3(2). Apex of clypeal protrusion narrow, narrower than intertorular distance Succinapis Engel

    —. Apex of clypeal protrusion broad, as wide as or slightly wider than intertorular distance Haidomelikertes Engel

    4(2). Mesoscutellum compressed (comparatively flattened) and extended 5

    —. Mesoscutellum rounded, not produced as extension 6

    5(4). Mesoscutellum formed as flattened, extended trapezoid, slightly longer than wide, with medioapical margin truncate [only workers known] Thyreomelikertes Engel, n. gen.

    —. Mesoscutellum with prominent, elongate, tonguelike medioapical extension projecting over metanotum, propodeum, and portions of metasoma [only male known] Mochlomelikertes Engel, Breitkreuz, and Ohl

    6(4). Mesoscutellum bulging, overhanging metanotum and propodeum; apical margins of metasomal terga distinctly lighter than remainder of metasoma, thus metasoma appears banded; anterior and posterior margins of metabasitarus distinctly converging toward apex 7

    —. Mesoscutellum not bulging, not overhanging metanotum or propodeum; metasomal terga uniformly colored; anterior and posterior margins of metabasitarsus approximately parallel 8

    7(6). Forewing with anterior margin of first submarginal cell approximately equal to length of anterior margin of second submarginal (i.e., r-rs as long as 3Rs); compound eyes converging below; flagellomere II distinctly shorter than flagellomere III, flagellomere I distinctly shorter than combined lengths of flagellomeres II and III Melissites Engel

    —. Forewing with anterior margin of first submarginal cell many times longer than length of anterior margin of second submarginal cell (i.e., r-rs many times longer than 3Rs); compound eyes approximately parallel; flagellomeres II and III approximately equal in length, flagellomere I approximately equal to combined lengths of flagellomeres II and III Roussyana Manning

    8(6). Apical margin of clypeus straight, flat, not flared anteriorly and not projecting over plane of labrum (genus Melikertes Engel, s.l.) 9

    —. Apical margin of clypeus flared and projecting anteriorly over plane with labrum, with shallow U-shaped emargination medioapically Amelikertotes Engel

    9(8). Forewing with two submarginal cells (1rs-m absent) Paramelikertes Engel and Ortega-Blanco

    —. Forewing with three submarginal cells (1rs-m present) Melikertes Engel, s.s.

    TABLE 3

    Current Classification of Tribe Melikertini (Apinae: Corbiculata)

    img-AL9i_01.gif

    Thyreomelikertes Engel, new genus

  • Type species: Thyreomelikertes electrosinicus, new species.

  • Diagnosis: This is a distinctive genus (figs. 2031), noteworthy for the flattened, trapezoidal mesoscutellum that projects posteriorly well over the metanotum and wholly declivitous propodeum (figs. 23A, 27). In addition, the mesosoma is densely setose, giving the bees a shaggy appearance (figs. 20, 21, 23A, 26, 27). The clypeus lacks the protrusions and other modifications that characterize several genera of Melikertini. Among other genera, the new genus is similar to that of Mochlomelikertes Engel, Breitkreuz, and Ohl in Baltic amber as both have elongate and projecting mesoscutella, although that of the latter is vastly more extreme and covers portions of the metasoma (refer to Engel et al., 2014). Mochlomelikertes is noticeably larger at 8 mm in total length (forewing length 6.75 mm), while species of Thyreomelikertes are less than 6 mm in length.

  • A full diagnosis for the genus is as follows: Small bees, less than 6 mm in length; integument seemingly dark brown to black, without maculation (figs. 2031). Mandible with outer mandibular grooves largely reduced, outer ridge prominent, faint indication of auxiliary ridge, clear acetabular groove and pollicar basin, latter divided into auxiliary basin, rutellar basin evidence, condylar ridge forming dorsal ramus apically (refer to Discussion), basins vanishing preapically and proximally, apical margin oblique, with shallow, broad incision demarcating short first preapical tooth, above which a broad concave margin between first preapical tooth and second preapical tooth at upper edge of mandible (upper distal angle), condylar groove in apical third with line of 3–4 long, ventrally directed, suberect (obliquely angled toward apex of mandible), thickened, flat setae (fig. 22); malar space linear (fig. 22); labrum flat, slightly broader than long, lateral margins parallel (not converging apically), without lateral fringe setae, apical margin with broad, shallow concavity medially, weakly and broadly rounded lateral to concavity (resulting in a weakly and broadly bilobed appearance to apical margin), surface with sparsely scattered long, fine, erect, simple setae, apical margin with numerous long, simple, fuscous setae, such setae thicker than those of disc; clypeus weakly convex, unmodified (i.e., without clypeal specializations or protrusions); epistomal sulcus laterally forming obtuse angle; upper torular tangent slightly below head midlength; scape elongate, longer than torulocellar distance; flagellomere I longer than wide, longer than flagellomere II; ocelli high on vertex, situated above upper tangent of compound eyes; vertex rounded in facial view, unmodified (no depressions or ridges); preoccipital area rounded; gena narrower than compound eye in profile.

  • \Mesotibia elongate; mesotibial spur present, simple, elongate; mesotibia and mesobasitarsus densely setose (fig. 24), setae intermixed with bristles, some setae minutely branched and sometimes capitate (fig. 24) (in T . electrosinicus); metatibia slender, elongate (fig. 25A), posterior margin gently convex and slightly widening medially; surface of corbicula not depressed; posterior margin with fringe of long plumose setae (branches minute and along length of setal rachis) (fig. 25B), anterior margin with simple or minutely branched setae and bristles, corbicular surface with scattered, erect bristles and long setae; inner surface with keirotrichiate zone field covering most of surface except posterior, narrow, slightly depressed (i.e., a weak clivulus present) subglabrate zone and a broad, squarish apical subglabrate zone (sensu Rasmussen et al., 2017), latter with some scattered short, suberect bristles, apical subglabrate zone slightly longer than apical width of metatibia; rastellum composed of stiff bristles along entire inner apical width of metatibia; single metatibial spur present, spur minutely ciliate along inner margin; metabasitarsus with auricle present on proximal surface facing apex of metatibia; metabasitarsus rectangular (fig. 25A) to squarish, longer or slightly longer than wide, margins roughly parallel, apical margin comparatively straight to concave, inner surface with abundant, elongate, suberect, simple bristles arranged in loose comb rows; pretarsal claws with minute inner subapical ramus; arolium present (figs. 20, 21, 25A).

  • Metasoma broad, ovoid (figs. 20, 21), with scattered, short, erect to suberect setae, such setae more abundant laterally; metasomal sterna II–V with abundant, long, erect, fine, simple setae, those setae of sterna IV and V shorter and more suberect to decumbent; sting present (difficult to observe in type species, although with care it can be found, but easily seen in the second species where the sting is exserted), simple (i.e., without proximally directed barbs).

  • Etymology: The new genus-group name is combination of the ancient Greek noun θυρεός (thyreos, meaning, “long shield”) and Melikertes, type genus of the tribe. The gender of the name is masculine.

  • Key to Species of Thyreomelikertes

  • 1. Fundal, corbicular, and retromarginal fringe setae of metatibia about as long as maximum width of metatibia (fig. 25), such setae with numerous short branches, those of corbicular surface somewhat capitate, and fringes intermixed with numerous, thick, simple bristles (fig. 25B); setae of mesotibia and mesobasitarsus about as long as maximum width of corresponding podite (fig. 24), setae on prolateral margin intermixed with simple bristles and long, capitate setae (fig. 24); metabasitarsus with apical margin comparatively straight (fig. 25A); larger species, forewing length greater than 5 mm T. electrosinicus, n. sp.

  • —. Fundal, corbicular, and retromarginal fringe setae of metatibia greatly elongate (figs. 26, 27), up to 2× maximum width of metatibia (figs. 29B, 30, 31C), such setae with numerous short branches (figs. 30B, 31C), those of corbicular surface not capitate, and fringes without thick, simple bristles (figs. 30B, 31C); mesotibial setae much longer than maximum width of mesotibia, setae never capitate; metabasitarsus apical margin concave (fig. 30A); smaller species, forewing length less than 4 mm T. kongi, n. sp.

  • Thyreomelikertes electrosinicus, new species
    Figures 2025

  • Diagnosis: This species differs from its only congener (see below for alternative character states) in the that the metatibial fringe, fundal, and corbicular setae are as long as than the maximum width of the metatibia (figs. 20, 21, 25), and similarly, the setae of mesotibia are about as long as the maximum mesotibial width and are intermixed with capitate setae (fig. 24). The species is also larger, with a forewing length over 5 mm. It also seems that the facial setae are sparser than those of the other species (fig. 22).

  • Description: WORKER. Total body length (as preserved) 5.44 mm; forewing length (as preserved) 5.30 mm. Head slightly longer than wide, length (summit of vertex to clypeal apical margin) 1.47 mm, width (maximum width across compound eyes, eyes partially collapsed inward thereby shortening the value) 1.29 mm; upper interorbital distance 1.26 mm, lower interorbital distance 1.22 mm. Scape elongate, length 0.58 mm, longer than torulocellar distance; flagellomere I longer than flagellomere II, about as long as combined lengths of flagellomeres II and III, flagellomere II shorter than flagellomere III. Gena distinctly narrower than compound eye in profile. Pronotum short, declivitous; mesoscutum anterior border broadly rounded, anterior lip gently curving to meet posterior pronotal margin; mesoscutum medial length 1.04 mm; intertegular distance 1.07 mm; mesoscutellum medial length 0.39 mm. Metatibia slender, length 2.09 mm, maximum width 0.41 mm; metabasitarsus longer than wide, length 0.81 mm, maximum width 0.34 mm, apical margin comparatively straight. Forewing with basal vein (1M) straight, confluent with 1cu-a, pterostigma longer than wide, maximum width slightly proximal to midlength, margin inside marginal cell sloping to costal margin; marginal cell broadly rounded apically, well separated from anterior wing margin; 2Rs arched posteriorly; 3Rs shorter than r-rs and slightly shorter than 4Rs; 1rs-m straight, 2rs-m strongly arched apically in posterior half; 2m-cu proximal 2rs-m by about vein width; wing membranes hyaline clear; veins dark brown.

  • Labrum faintly imbricate and impunctate; clypeus seemingly smooth with some weak, shallow punctures laterally; integument of much of face difficult to discern owing to collapse of face, but vertex smooth and shining, seemingly impunctate. Mesoscutum and mesoscutellum apparently smooth with sparse faint punctures (integument exceptionally challenging to observe); metasomal terga smooth to faintly imbricate and shining, with minute punctures at bases of setae.

  • Pubescence generally fuscous and rather abundant; labrum with sparse, erect, simple setae on surface, apical margin with fringe of elongate setae, such setae about 0.75× labral length; clypeus with scattered, simple, erect, short setae, such setae more numerous on remainder of head and becoming quite elongate on vertex. Scape with scattered short, erect to decumbent, simple, fine setae, without bristles. Mesosoma rather densely pubescent but not obscuring integument, setae elongate, erect, and simple or with minute branches along rachis, many setae noticeably thick on all surfaces except metanotum and basal area of propodeum; plumose setae more numerous on posterior of mesoscutellum, with branches slightly longer and more numerous, thus appearing brushy. Wing membranes with abundant microtrichia. Profemur and protibia with abundant, short setae, those of dorsal surfaces longer than those of prolateral, retrolateral, and ventral surfaces; probasitarsus with long, erect bristles on dorsal surface, similar bristles on other surfaces except slightly shorter; mesofemur with abundant short, erect setae, intermixed with short bristles on outer surface, ventrally setae shorter and bristles lacking; mesotibia with dense, erect setae and bristles, on dorsal surface, such setae and bristles progressively longer from base to near apex, setae with some minute branches, retrolateral surface with bristles sparse and setae longer, about as long as mesotibial width, such setae with abundant minute branches and some with thickened apices, thus appearing somewhat clavate apically, ventral surface with setae shorter; mesobasitarsus similar to mesotibia except bristles longer and more numerous on dorsal surface and more evident on retrolateral surface, setae of retrolateral surface not appear clavate; metafemur with short, erect setae on dorsal and prolateral surfaces, retrolateral and ventral surfaces with setae exceedingly sparse, with a patch of keirotrichia on retrolateral surface apically; metatibia with abundant, long, erect bristles along anterior edge of proventral surface and inferiorly on prolateral surface, such setae intermixed with long, minutely branched setae, proventral surface largely asetose, prolateral surface with sparsely scattered long, erect, simple setae and bristles, such setae similar on profundal and corbicular surfaces, retromarginal fringe composed of dense, long, minutely branched setae, such setae about as long as maximum metatibial width, retrolateral surface with keirotrichiate zone covering majority of surface except narrow, slightly depressed, superior, subglabrate zone and squarish, apical subglabrate zone; rastellum composed of dense series of thick bristles tapering in length; metabasitarsus with abundant erect bristles on proventral and prolateral surfaces, bristles intermixed with some short setae, retromarginal edge with bristles less prominent, with dense minutely branched setae, such setae slightly shorter than bristles and rounding onto superior apical angle, retrolateral surface with loose comb rows composed of long, suberect bristles; auricle bordered by fringe of short, branched setae. Metasomal setation as described for genus (above).

  • QUEEN AND MALE. Unknown.

  • Holotype: Female worker, CNU-HYM-LF-2023-001 (figs. 20, 21), in a single piece of amber without syninclusions, Fushun coalfield, Liaoning Province, northeastern China, Guchengzi Formation, Ypresian (Eocene); deposited in the fossil insect collection of the Key Laboratory of Insect Evolution and Environmental Changes, College of Life Sciences, Capital Normal University, Beijing, China.

  • The holotype is the best preserved of the available bees from Fushun amber and is virtually complete, and observation is not hampered by prominent syninclusions, bubbles, or other imperfections in the amber. The bee is rather close to some curved surfaces that prevent preparation of the amber surface from some angles— the only hindrance to a perfect view of certain features. The wings are extended posteriorly over the body, although the left forewing is slightly more askew. The forelegs are raised near the head, although the left foreleg is more extended into the amber beneath the bee. The mid and hind legs are extended below and, with the exception of the left midleg, project out from the body. The worst defects concern the head, which is somewhat obscured by an oblique internal fracture. More critically, the integument of the face is partially collapsed, rendering interpretation a challenge. The top of the head and the mesosomal nota are partially obscured by encrusted material entangled in the setae, somewhat obscuring the integument. The mesosoma and metasoma are partially compressed but inconsequentially. The wings are excellent although the left forewing was slightly stretched during preservation, resulting in slight fractures along some of the fenestrae, bullae, and across the marginal cell. The right forewing is unmolested.

  • Etymology: The specific epithet is a combination of the Latin adjective electricus (from electrum), meaning, “of amber” and the Medieval Latin adjective sinicus, meaning “of or related to China.”

  • FIGURE 20.

    Photograph, dorsal view, of holotype (CNU-HYM-LF-2023-001) worker of Thyreomelikertes electrosinicus, n. gen. and sp.

    img-z37-1_01.jpg

    FIGURE 21.

    Photograph, ventral view, of holotype (CNU-HYM-LF-2023-001) worker of Thyreomelikertes electrosinicus.

    img-z38-1_01.jpg

    FIGURE 22.

    Photograph of head of Thyreomelikertes electrosinicus, right oblique facial view.

    img-z39-1_01.jpg

    Thyreomelikertes kongi, new species
    Figures 2631

  • Diagnosis: This species, which is known from three individuals (figs. 26, 27, 31A), can be distinguished by the distinctly more elongate fringe, fundal, and corbicular setae on the metatibia (figs. 27, 29, 30, 31A, B), the setae much longer than the maximum metatibal width. Likewise, the setae of the mesotibia are much longer than the maximum mesotibial width (up to about 1.5× mesotibial width) and none of the setae are capitate. The species is overall smaller than that of the type species (refer to metrics) and the apical margin of the metabasitarsus is concave (fig. 30A) (rather than straight). The facial setae, particularly on the clypeus, are seemingly denser than in T . electrosinicus (fig. 31B).

  • Description: As described for Thyreomelikertes electrosinicus (above), with the following exceptions: WORKER. Total body length (as preserved) 3.51 mm; forewing length (as preserved) 3.78 mm. Head slightly wider than long (fig. 31B), length (summit of vertex to clypeal apical margin) 1.16 mm (paratype 1.21 mm), width (maximum width across compound eyes) 1.66 mm (paratypes 1.32–1.46 mm). Scape elongate, length 0.47 mm (artificially stretched) (paratypes 0.23–0.38 mm, both damaged), longer than torulocellar distance; flagellomere I longer than flagellomere II, shorter than combined lengths of flagellomeres II and III, flagellomere II shorter than flagellomere III. Mesoscutum medial length 0.57 mm; intertegular distance 0.67 mm (paratype 0.55 mm); mesoscutellum medial length 0.36 mm. Metatibia slender, length 1.50 mm (paratypes 1.11–1.61 mm), maximum width 0.45 mm (paratypes 0.22–0.37 mm); metabasitarsus longer than wide, length 0.50 mm (paratype 0.34 mm), maximum width 0.37 mm (paratype 0.22 mm), apical margin concave. Forewing (fig. 28) with basal vein (1M) slightly proximad 1cu-a, pterostigma maximum width at midlength; marginal cell apex well separated from anterior wing margin, with nebulous appendiculate vein; 3Rs much shorter than both r-rs and 4Rs; 1rs-m arched; 2m-cu proximal 2rs-m by about 2–3× vein width.

  • Mesotibia with dense, erect setae and bristles (fig. 29A), on dorsal surface, such setae and bristles progressively longer from base to apex, setae with some minute branches, retrolateral surface with bristles sparse and setae longer, longer than mesotibial width, such setae with abundant minute branches, no capitate setae; metatibia with dense elongate (distinctly longer than metatibial width), minutely branched setae forming retromarginal fringe (figs. 29B, 30, 31A), similar but shorter setae along proventral margin, retromarginal fringe sparsely intermixed with long, erect bristles and such bristles sparse along anterior edge of proventral surface and inferiorly on prolateral surface, proventral surface largely asetose, prolateral surface with sparsely scattered long, erect, simple setae and bristles, such setae similar on profundal and corbicular surfaces.

  • QUEEN AND MALE. Unknown.

  • Holotype: Female worker, CNU-HYM-LF-2023-003 (figs. 26, 27), in a single piece of amber with two further bees as syninclusions, Fushun coalfield, Liaoning Province, northeastern China, Guchengzi Formation, Ypresian (Eocene); deposited in the fossil insect collection of the Key Laboratory of Insect Evolution and Environmental Changes, College of Life Sciences, Capital Normal University, Beijing, China.

  • The holotype is preserved in a relatively clear piece of amber and has two syninclusions of the same species as well as a tiny chalcidoid wasp near the wing apex. The holotype is located at one end of the piece, while the two other bees are clustered together and overlapping each other. The holotype is well preserved and can be observed in all orientations except ventral. The wings are obliquely folded over the body and the legs tucked up alongside the body.

  • Paratypes: Female workers, CNU-HYM-LF-2023-004 and CNU-HYM-LF-2023-005, in the same piece of amber with the holotype (fig. 26), Fushun coalfield, Liaoning Province, northeastern China, Guchengzi Formation, Ypresian (Eocene); deposited in the fossil insect collection of the Key Laboratory of Insect Evolution and Environmental Changes, College of Life Sciences, Capital Normal University, Beijing, China.

  • The paratypes are distinctly less well preserved relative to the holotype and, given that they are partially entwined as well as damaged through compression or missing sclerites, it was not possible to see each structure as well as in the holotype. It was also not possible to get all of the same measurements from the paratypes as was possible from the holotype. From observable characters, the paratypes agree perfectly with the holotype and so we are confident in their con-specificity with the holotype.

  • Etymology: The specific epithet honors Chuijin Kong, who graciously donated the specimens for this study.

  • FIGURE 23.

    Photographs of Thyreomelikertes electrosinicus. A. Mesosomal dorsum. B. Forewing.

    img-z41-1_01.jpg

    FIGURE 24.

    Photographs of Thyreomelikertes electrosinicus. A. Mesotibia and mesotarsus. B. Detail of mesotibia, note capitate setae.

    img-z42-1_01.jpg

    FIGURE 25.

    Photographs of Thyreomelikertes electrosinicus. A. Metatibia, metatarsus, and metapretarsus. B. Detail of metatibial setae and bristles on proventral margin.

    img-z43-1_01.jpg

    FIGURE 26.

    Photographs of entire amber piece with holotype (CNU-HYM-LF-2023-003) and paratypes (CNU-HYM-LF-2023-004 and CNU-HYM-LF-2023-005) of Thyreomelikertes kongi, n. sp. A. Upper view, with holotype at right. B. Lower view, with holotype at left.

    img-z44-1_01.jpg

    FIGURE 27.

    Photographs of holotype (CNU-HYM-LF-2023-003) worker of Thyreomelikertes kongi, with microhymenopteran syninclusion. A. Left lateral-posterior oblique view. B. Dorsal view.

    img-z45-1_01.jpg

    FIGURE 28.

    Photograph of forewing venation of holotype (CNU-HYM-LF-2023-003) worker of Thyreomelikertes kongi.

    img-z46-1_01.jpg

    DISCUSSION

    Despite the condition of the holotype, Glyptosmia hemiaspis is a remarkable species that uniquely intermingles a number of interesting traits, while at the same time encapsulating those character states typical of early-diverging Megachilinae. The presence of a metatibial scopa, claws with a short subapical inner tooth, arolia, and an arched basal vein are familiar features of Ctenoplectrellini and Glyptapini. Likewise, the areolate propodeum and elongate mandible with an edentate upper apical margin are akin to Glyptapini, while the tuberculate metanotum recalls that of Dioxyini. Glyptosmiini will be an important taxon for future cladistic treatments of Megachilidae.

    The discovery of melikertines in Fushun amber reflects that these are the most frequent bees in Eocene amber. The species from Fushun corroborate the conclusion that these social bees were widespread during the epoch, extending from at least western Europe to northeastern Asia and south to northwestern India. It is assuredly the case that the tribe was ubiquitous across Eurasia and there is the potential that they occurred in North America and Africa during the time, necessitating an exploration of Eocene amber from Arkansas for inclusions. The tribe Electrapini had a similar distribution, with species not only frequent in Europe but also occurring in Cambay amber and as compressions from the Eocene Green River Formation (Engel and Davis, 2021).

    Mandibular Structure across Corbiculate Bees

    As imaging techniques and technologies have improved and new fossils with orientations more suitable for observation have accumulated, it has been possible to refine the description of key characters for many fossil bees (as is done herein for Glyptapini and Ctenoplectrellini). For melikertines this has been particularly important, such as a more accurate accounting of the prominent clypeal modifications of many genera (Engel and Davis, 2021). Here we can provide some clarification to the structure of the worker mandible as the holotype of Thyreomelikertes electrosinicus has the left mandible extended such that the surfaces can be quite nicely observed. These observations are complemented by specimens of other melikertines in Baltic amber (Engel and Davis, 2021).

    The mandibles of bees are complex structures, with a great variety of modifications in relation to the different histories and biologies of the numerous lineages of Anthophila. Michener and Fraser (1978) provided a general overview of the various structures of the mandible, identifying consistent patterns of ridges and grooves (several of which broaden into distinct basins near the apical margin), many of these clearly homologous across lineages, which are useful in phylogenetic estimation and classification. In corbiculate bees, the mandibles are quite divergent from other Apidae, and especially from the Centridini, the closest relatives of the Corbiculata (e.g., Bossert et al., 2019; Orr et al., 2022). Much of the interpretation of Michener and Fraser (1978) holds nicely across corbiculate bees, although we believe an alternative interpretation of the location of the “acetabular carina” and concomitantly the position and extent of the “acetabular groove” in Euglossini is necessary and that such a reinterpretation forces a cascade of minor but important changes for homology assignments in the other corbiculate tribes. Ultimately, this permits an understanding of the mandible of Melikertini, Meliponini, and Apini. Accordingly, a rather lengthy digression is necessary to establish proper homologies before carefully considering melikertines. In the account that follows we have used the terminology and forms of annotation presented by Michener and Fraser (1978) except where we believe reinterpretations are necessary (as explained below: table 4). Some of the key structural elements of the bee mandible are presented in figure 32, exemplified by Centris (Centris) adani Cockerell (Centridini) and Eulaema (Eulaema) meriana (Olivier) (Euglossini).

    FIGURE 29.

    Photographs of holotype (CNU-HYM-LF-2023-003) worker of Thyreomelikertes kongi. A. Mesotibia, dorsal view. B. Metatibia, prolateral view.

    img-z48-1_01.jpg

    FIGURE 30.

    Photographs of holotype (CNU-HYM-LF-2023-003) worker of Thyreomelikertes kongi. A. Metatibia, metatarsus, and metapretarsus, retrolateral view. B. Detail of retromarginal fringe of metatibia, retrolateral view.

    img-z49-1_01.jpg

    FIGURE 31.

    Photographs of paratype workers of Thyreomelikertes kongi. A. Two workers together, left worker (CNU-HYM-LF-2023-005) in posterior view, right worker (CNU-HYM-LF-2023-004) in facial view. B. Facial view of CNU-HYM-LF-2023-004. C. Retromarginal fringe of CNU-HYM-LF-2023-005.

    img-z50-1_01.jpg

    In large part we agree with the interpretations of Michener and Fraser (1978), although we believe those authors misinterpreted the acetabular carina in Euglossini and thereby the associated acetabular groove and associated structures (e.g., extent of mandibular pollex). In Bombini Michener and Fraser (1978: fig. 41) correctly noted the position of the acetabular carina apically, but then in Euglossini (1978: fig. 39), rather than associating it with the same carina in the same position and with the same connection proximally toward the acetabulum, they identified a separate ridge much further down on the outer surface. From our interpretation this seems due to their selection for their comparative study of derived species of Eulaema Lepeletier, Eufriesea Cockerell, and Euglossa Latreille among free-living Euglossini. In more derived euglossines the surface above this strong ridge, here called the auxiliary ridge (fig. 32B, 35), is typically modified such that any indication of a basin and difference between this surface and the upper margin is slight, although often a faint carina can be seen running parallel to the upper margin, and which we believe (as argued below) is the actual acetabular carina. Indeed, if one expands the comparison to putatively more plesiomorphic euglossines, such as the subgenus Euglossella Moure, then the comparison with Bombini becomes more evident and an acetabular carina in the same position as in the bombine mandible is clear, and this carina extends proximally to the acetabulum (fig. 35). The ridge identified in Euglossini as the “acetabular carina” by Michener and Fraser (1978), however, is never connected to any ridge or carina that extends proximally to the position of the acetabulum and is instead always separate from both an acetabular ridge/ carina as well as, in Euglossini, the outer ridge of the mandible (figs. 32B, 35, 36).

    To be precise, in Euglossa (Euglossella) decorata Smith, there is a weak carina on the outer surface bordering the trimmal extension at the upper margin of the mandible (fig. 35). This carina extends proximally to the acetabular outer margin of the trimma in the area of the acetabulum, i.e., unlike the ridge identified in Euglossini by Michener and Fraser (1978), it is in a direct path with the actual acetabulum and its course perfectly matches that of the acetabular carina and acetabular groove in Centridini (and even more broadly across bee mandibles). This uppermost carina is similarly adjoined to a piliferous groove beset with intermingled, longer, erect setae similar to the acetabular fimbrial setae of the acetabular groove in other bees (figs. 3336). Collectively, these features serve to identify this uppermost carina as the actual acetabular carina and the associated groove as the acetabular groove, which expands apically into a pollicar basin (fig. 36). This pollicar basin extends apically until a thin, weak premarginal carina delimits the glabrous apical margin of the pollex (figs. 35, 36). Michener and Fraser (1978) considered this same area as merely an augmented part of the trimmal extension and the corresponding groove as a pubescent area of the trimma, all above their prominent “acetabular carina” (= auxiliary ridge herein). Accordingly, the evidence supports the interpretation that the euglossine acetabular carina is a thin, carinate ridge, which is sometimes so faint as to be effectively effaced in derived species, although even in these groups it remains thinly and faintly evident near midlength (at least in unworn or lesser-worn mandibles). This means that the acetabular carina is identical in position in both Euglossini and Bombini (figs. 35, 37).

    Importantly, the acetabular groove in corbiculate bees, at least plesiomorphically, is typically a piliferous surface with a more or less developed acetabular fimbria composed of longer, erect, stiffer setae (figs. 3542, 44B, C) (centridines have the narrow groove and fimbriae but lack the broader piliferous basins apically: figs. 33, 34), quite clear in Euglossella and Bombini (figs. 3538). The acetabular groove runs to the preapical margin and, in pairing with the outer ridge just beneath, separates at the margin the pollex from the rutellum (figs. 33, 34), and in corbiculate bees the acetabular groove expands apically into a basin, the piliferous pollicar basin (based on its comprising the apex of the pollex). In Euglossini, the auxiliary ridge is discontinuous with the outer ridge at its base, completely separated by the proximal merging of the piliferous grooves above and below the auxiliary ridge, isolating the auxiliary ridge from the outer ridge proximally (fig. 35). This supports the conclusion that the grooves both above and beneath the auxiliary ridge are collectively the entire acetabular groove, merely divided by the auxiliary ridge in Euglossini and Bombini. The piliferous groove beneath the auxiliary ridge, as it borders the apical margin of the pollex, is an inferior portion of the acetabular groove and basin (beneath the auxiliary ridge), and when separated completely by the auxiliary ridge can be called an auxiliary groove of the acetabular groove, which if widened apically, extends into an auxiliary basin (= lower apical part of the pollicar basin when an auxiliary ridge is present). This interpretation indicates that the pollex occupies a far greater portion of the apical margin of the euglossine mandible (fig. 37). Thus, the auxiliary ridge, which is nowhere present in the outgroups (figs. 33, 34), is a secondary structure that effectively divides what is a far broader acetabular groove and its apical basin. In most Eulaema the auxiliary groove and basin are the only portions of the acetabular groove evident; the main course of the latter and its apical basin above the auxiliary ridge are only faintly depressed and largely glabrous. The groove and basin above the auxiliary ridge become more continuous with the trimmal extension, likely a result of orchid bee biology. The same is the case for most Eufriesea, although there the main course and basin often have some scattered minute setae hinting at their former presence.

    Michener and Fraser (1978) interpreted the auxiliary ridge in bumble bees to be a bifurcation of the outer ridge, what they referred to as the upper distal branch of the outer ridge. Given the earlier-diverging position of Euglossini among corbiculates (e.g., Plant and Paulus, 1987, 2016; Prentice, 1991; Engel, 2001; Schultz et al., 1999, 2001; Bossert et al., 2019; Engel and Rasmussen, 2021) conjoined with the aforementioned interpretation of this ridge in orchid bees, we believe that the outer ridge does not bifurcate. If we follow our interpretation, then the pattern in Euglossini is identical to that observable in Bombini. Here again, the acetabular carina and acetabular groove are continuous and in line with the proximal acetabulum, while the auxiliary ridge is not and is instead connected via a narrow proximal extension to the outer ridge but plesiomorphically separated by a narrow piliferous channel from the apical margin of the pollex (as evident in many species of the early diverging bumble bee subgenera Mendacibombus Skorikov and Bombias Robertson). In fact, an apomorphic feature of the bombine mandible, and likely a new synapomorphy for the tribe, is this proximal attachment of the auxiliary ridge to the outer ridge of the mandible (fig. 37). In early diverging bumble bees, the pollicar basin is continuous apically, with the main groove extending above the auxiliary ridge and the auxiliary groove just below it (fig. 38). In more derived lineages of Bombus Latreille the auxiliary ridge connects to the apical margin of the pollex and separates the auxiliary groove and auxiliary basin from the remainder of the acetabular groove and basin (much like it is in Euglossini). Michener and Fraser's (1978) lower distal branch of the outer ridge is simply the outer ridge as interpreted by us and a separate structure from the auxiliary ridge (= upper distal branch of outer ridge of bombines sensu Michener and Fraser). The reason for this interpretation is that this lower “branch,” when present (e.g., Euglossini and Bombini), with the acetabular groove bordering it above, always makes contact with the apical margin at the boundary between the pollex and rutellum, distinctions more apparent in other bee lineages (e.g., Centridini, Anthophorini, Xylocopinae, Lithurginae, Fideliinae, Melittidae, Halictidae, Andrenidae, Colletidae). In some derived subgenera of bumble bees, such as some Bombus, s.s., the medial portion of the outer ridge is somewhat effaced and a continuous ridge extends strongly from the outer ridge root through what is here considered the auxiliary ridge. We interpret this as merely a secondary derivation in some “higher” bumble bees. As already noted, the structure of these ridges in the early diverging subgenera of bumble bees (Mendacibombus, Bombias) have the outer ridge continuous across the lower of Michener and Fraser's two branches and a weak connection to the auxiliary ridge, which in the most basal extant Bombus is interrupted by a piliferous channel before the apical margin (figs. 37, 38). Accordingly, in Bombini the attachment of the auxiliary ridge to the outer ridge is a synapomorphy, plesiomorphically the auxiliary ridge does not reach the apical margin, and the contact of the auxiliary ridge with the apical margin and/ or the medial effacement of the outer ridge (sometimes with a secondary carina running across the outer ridge root and auxiliary ridge) are secondary apomorphies within subordinate clades of Bombus. Another secondary derivation among derived Bombus is the complete loss of the auxiliary ridge, as in some Psithyrus Lepeletier. The presence in many bumble bees of an interstitial ridge is either a secondary structure entirely or perhaps a modification of the dorsal ramus of the condylar ridge (outer condylar ridge of Michener and Fraser). The dorsal ramus broadens near midlength in many bees (e.g., Centridini: figs. 32A, 33) but then tapers again toward the apex. In Bombus the broadened dorsal ramus has a shallow channel running parallel to its dorsal margin before rising again to form the lower boundary of the outer groove. The result is the appearance of a second ridge dorsal to the dorsal ramus of the condylar ridge and below the outer ridge. Secondary ridges in Euglossini are derived as proximal extensions of the rutellar cap and are therefore of independent derivation (figs. 32B, 35).

    TABLE 4

    Comparison of Terminologies for Structures and Areas on the Outer Surface of Corbiculate Bee Mandibles

    img-z53-4_01.gif

    FIGURE 32.

    Photographs of the outer surfaces of noncorbiculate and corbiculate apine mandibles, with particular structures labeled. A. Centris (Centris) adani Cockerell (Centridini). B. Eulaema (Eulaema) meriana (Olivier) (Euglossini).

    img-z54-1_01.jpg

    The interpretation of these carinae, ridges, and grooves have implications for the interpretation of the composition of the apical margin of the mandible and especially the teeth. The dentition of bee mandibles derives from various mandibular components. The apex or lower apical angle is the apical tooth (= At, R1, or Mt1, first mandibular tooth) (fig. 32). It is equivalent to the apex of the cap of the rutellum (Eickwort, 1969; Michener and Fraser, 1978). When present, additional teeth may derive from the rutellum, the pollex, or even the apicalmost angle of the trimmal extension. For example, in Centridini the second tooth is derived from the rutellum and is therefore the second rutellar tooth (R2 = Mt2, second mandibular tooth), whereas the apical tooth is the first rutellar tooth (fig. 32A, 33). Alternatively, only the apical tooth may be from the rutellum and the second tooth is at the lower edge of the pollex and is thus the first tooth of the pollex (P1 = Mt2). Accordingly, “mandibular tooth” is a descriptive term, while refinements such as R2 versus P1 are indications of morphological identity. In those bees where there is a sharp upper distal angle or “tooth” situated above the level of the acetabular carina, this is typically considered a tooth of the pollex (and often numbered as such: P2) but is more properly a continuation of the trimmal extension (and for purposes of homology should likely be called Tt1 = first trimmal tooth).

    FIGURE 33.

    Carinae, ridges, and regions of outer mandibular surface of Centris (Centris) adani Cockerell (Centridini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing carinae, ridges, and regions with color overlay: red = acetabular carina; yellow = outer ridge; cyan = condylar ridge; dark blue = margin of pollex; pink = rutellar cap.

    img-z56-1_01.jpg

    FIGURE 34.

    Basins and grooves of outer mandibular surface of Centris (Centris) adani Cockerell (Centridini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing basins and grooves with color overlay: red = condylar groove; yellow = rutellar basin; green = outer groove; cyan = acetabular groove.

    img-z57-1_01.jpg

    From these fundamentals it is possible to interpret the pattern of structures on the mandibles of extinct corbiculate bee lineages. Not all specimens or species are preserved in such a manner as to allow for optimal observation of the mandibles, but we record here structures for Protobombus indecisus Cockerell (fig. 43) (Electrapini) and three species of Melikertini: Aethemelikertes emunctorii Engel (fig. 45A), Succinapis micheneri Engel (fig. 45B, C), and Thyreomelikertes electrosinicus (fig. 44A). A fourth melikertine, Amelikertotes clypeatus (Engel), is figured in Engel and Davis (2021). The mandible of Protobombus has, like Euglossini and Bombini, the outer surface of the mandible beset by strongly developed ridges and grooves (fig. 43), whose pattern shows similarities to both of these tribes. The acetabular carina is in the same position as identified for Euglossini and Bombini, and extends proximally where the abutting acetabular groove meets with the proximal end of the auxiliary groove, the two grooves separated by a strongly raised and arching auxiliary ridge, which is itself clearly separated from the outer ridge and even arches at its base toward the acetabular carina, but does not join it owing to the acetabular groove, and away from the outer ridge. It is interesting to note that in Euglossini the auxiliary ridge has a broad arch medially, which projects the ridge in a dorsal direction before tapering obliquely downward and apically to the apical margin (figs. 32B, 35). This is not the case in Bombini where the auxiliary ridge, attached to the outer ridge proximally, extends in a more or less straight manner toward the apical margin (fig. 37). Protobombus has the same broad medial arch observed in Euglossini (fig. 43A). As is typical for corbiculate bees, the piliferous pollicar basin is broad along the apical margin, with the portion of the apical margin attributed to the pollex bearing a notch in its lower third and setting off a broad second preapical tooth (P2). The outer ridge is strong throughout its course and extends obliquely from near the acetabulum to the lower apical margin of the rutellar cap. The condylar ridge is simple and does not divide apically into dorsal and ventral rami (fig. 43B). The rutellar basin is small and poorly developed, without any intercalary rutellar ridge and the outer groove is comparatively weak, there is no interstitial ridge, and there does not appear to be an inferior groove (fig. 43B). Although not part of the mandible, it is worth noting that the labrum is transverse and its surface flat, lacking the tubercles and other elements that characterize bombine labra (fig. 43A).

    Melikertine mandibles are best considered in conjunction with those of Meliponini and Apini. The mandibles of Meliponini and Apini are noteworthy for the reduction of the grooves and ridges that are clearly evident on the outer surface of the mandibles of other corbiculate bees (figs. 3538) (Euglossini, Bombini, Electrobombini, and, where evident, Electrapini) (Michener and Fraser, 1978; Michener, 1990, 2007; Engel and Rasmussen, 2021). In addition, the apical margins are more oblique and, given the apparently loss of the acetabular groove, a clear rutellar cap was considered absent or at least impossible to differentiate. The mandible of Apini certainly presents an extreme reduction; given the similarities of Apini and Meliponini, the mandible of Apis best discussed in light of meliponine mandibles as the faint remnants of outer structures in stingless bees aids the identification of elements even in such a dramatic state as that of honey bees.

    Although reduced, the Meliponini retain some vestiges of the acetabular carina, identified not only by a faint ridge but also by the shallow or downright faint acetabular groove. In many stingless bees the acetabular carina parallels the upper margin of the mandible, such as in Cephalotrigona zexmeniae (Cockerell) (fig. 44B) and Melipona (Melipona) quadrifasciata Lepeletier (fig. 44C). In others, such as Geniotrigona thoracica (Smith) (fig. 41), the ridge is faint but discernible and extends straight along the trimma. The acetabular groove is shallow but easily discerned not only by a depression but also by the acetabular fimbria that runs along its upper and apical border (figs. 41, 42, 44B, C). In most the pollicar basin that forms apically is not only evident from the fimbria and depressed surface but also by a change in integumental sculpture and may at times be weakly piliferous (figs. 41, 42, 44B, C). Thus, we observe that the corbiculate trait of a pollicar basin that expands along the apical margin of the mandible is also present, albeit fainter, in Meliponini. Importantly, the ventral termination of this base, along with the acetabular fimbria, informs us of the limit of the pollex along the apical margin and, therefore, the apical margin ventral to this is the rutellar cap (fig. 41). Remarkably, there is nearly always a thin piliferous groove medially on this rutellar cap, which opens proximally into the outer groove, and is therefore putatively a reduced rutellar basin (figs. 41, 42, 44B, C). Another interesting feature is what appears to be a faint auxiliary ridge, at least in G . thoracica (fig. 41). The change in integumental sculpture that seems to cut into the base of the pollicar basin in M . quadrifasciata could also be a vestigial auxiliary ridge (fig. 44C). Given how faint these structures are, we draw them to the attention of the reader but refrain from conclusively stating that Meliponini retain a vestigial auxiliary ridge. Regardless, the features are exceedingly reduced as is typical of other outer mandibular features of Meliponini and Apini. The outer ridge is nearly indiscernible were it not for the presence in most Meliponini of a distinct outer groove that arches apically to the end of the rutellar basin and into a shallow notch on the ventral margin, which appears to be a reduced condylar groove (figs. 41, 42, 44B, C). The upper margin of this groove would presumably be the outer ridge (fig. 41). With all this in mind, if we turn to the more extremely modified mandible of Apini (figs. 39, 40) we find a nearly identical pattern to that in Meliponini. An acetabular carina is weak but certainly present and is even more easily observed as the acetabular groove (along with its fimbria) deepens preapically. The pollicar basin extends, like other corbiculates, along the oblique apical margin, but, unlike other corbiculates, remains a narrow depression rather than broadening proximally (fig. 40). The termination of this groove differentiates the lower extreme of the pollex from the otherwise unrecognizable rutellar cap (fig. 39). There are two nearly indistinguishable irregular depressions, appearing almost like the fusion of a short series of irregular punctures (one possible interpretation for these “grooves”). Given their location and that the sculpturing is different in each, these could be exceptionally vestigial outer grooves and rutellar basins, respectively (fig. 40). A more noticeable short, subapical groove on the ventral surface is assuredly the remnant of a condylar groove (fig. 40). There is a thin and weak ridge proximally that is likely the outer ridge and if one attempts to follow it there is a faint indication of a course that takes it across the medial surface (where it is almost entirely effaced) and then up along the upper margin of the two irregularly pitted areas that may represent the outer groove and rutellar basin (fig. 39). While the acetabular carina, acetabular groove, acetabular fimbria, pollex, rutellar cap, and condylar groove are clear, the remainder of elements on the honey bee mandible might be best considered working hypotheses for the moment. Regardless, what is often called the “oblique groove” in Meliponini is certainly the outer groove of other corbiculates and is so labeled herein.

    The outer structures of the mandibles of Melikertini are certainly reduced relative to Euglossini, Bombini, Electrobombini, and Electrapini, but are more apparent in some genera than those of Meliponini and Apini. As in all of the tribes, the acetabular carina is evident and borders an acetabular groove that takes on the same course as in Meliponini (figs. 44A, 45), the latter broadening apically along the apical margin. In Thyreomelikertes this basin is much broader than in other melikertines and there is a faint longitudinal ridge dividing the basin, which we interpret as a vestigial auxiliary ridge (fig. 44A). This feature is quite variable among melikertines, with it particularly prominent in Succinapis (fig. 45C) but apparently absent in Aethemelikertes Engel (fig. 45A). Interestingly, in Succinapis the auxiliary ridge has a weak medial arch, putatively a plesiomorphic form based on the shape of Euglossini and Protobombus.

    FIGURE 35.

    Carinae, ridges, and regions of outer mandibular surface of Euglossa (Euglossella) decorata Smith (Euglossini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing carinae, ridges, and regions with color overlay: red = acetabular carina; orange = auxiliary ridge; yellow = outer ridge; cyan = condylar ridge; dark blue = margin of pollex; pink = rutellar cap.

    img-z60-1_01.jpg

    FIGURE 36.

    Basins and grooves of outer mandibular surface of Euglossa (Euglossella) decorata Smith (Euglossini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing basins and grooves with color overlay: red = condylar groove; orange = inferior groove; yellow = rutellar basin; green = outer groove; cyan = acetabular groove and pollicar basin (and auxiliary groove and basin).

    img-z61-1_01.jpg

    FIGURE 37.

    Carinae, ridges, and regions of outer mandibular surface of Bombus (Mendacibombus) defector Skorikov (Bombini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing carinae, ridges, and regions with color overlay: red = acetabular carina; orange = auxiliary ridge; yellow = outer ridge; green = interstitial ridge; cyan = condylar ridge; dark blue = margin of pollex; pink = rutellar cap.

    img-z62-1_01.jpg

    FIGURE 38.

    Basins and grooves of outer mandibular surface of Bombus (Mendacibombus) defector Skorikov (Bombini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing basins and grooves with color overlay: red = condylar groove; orange = inferior groove; yellow = rutellar basin; green = outer groove; cyan = acetabular groove and pollicar basin (and auxiliary groove and basin).

    img-z63-1_01.jpg

    FIGURE 39.

    Carinae, ridges, and regions of outer mandibular surface of Apis (Megapis) dorsata Fabricius (Apini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing carinae, ridges, and regions with color overlay: red = acetabular carina; yellow = outer ridge; cyan = condylar ridge; dark blue = margin of pollex; pink = rutellar cap.

    img-z64-1_01.jpg

    FIGURE 40.

    Basins and grooves of outer mandibular surface of Apis (Megapis) dorsata Fabricius (Apini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing basins and grooves with color overlay: red = condylar groove; yellow = rutellar basin; green = outer groove; cyan = acetabular groove and pollicar basin.

    img-z65-1_01.jpg

    FIGURE 41.

    Carinae, ridges, and regions of outer mandibular surface of Geniotrigona thoracica (Smith) (Meliponini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing carinae, ridges, and regions with color overlay: red = acetabular carina; yellow = outer ridge; cyan = condylar ridge; dark blue = margin of pollex; pink = rutellar cap.

    img-z66-1_01.jpg

    FIGURE 42.

    Basins and grooves of outer mandibular surface of Geniotrigona thoracica (Smith) (Meliponini); see table 4 for terminology. A. Original photograph. B. Grayscale version showing basins and grooves with color overlay: red = condylar groove; yellow = rutellar basin; green = outer groove; cyan = acetabular groove and pollicar basin (and auxiliary groove and basin).

    img-z67-1_01.jpg

    Like Meliponini, there is an outer groove that follows a nearly identical course with the exception that it extends to the apical margin in most genera, including Thyreomelikertes and Succinapis (figs. 44A, 45B, C), but may arch to the ventral margin near the apex (rutellar cap) in Aethemelikertes, Melikertes Engel, Amelikertotes Engel. In Melikertini, Meliponini, and Apini the condylar groove is scarcely evident and is effectively the same as the ventral margin of the outer surface. However, in Thyreomelikertes, unlike other melikertines observed, the condylar carina is evident apically and extends dorsally over the condylar groove (fig. 44A), a noteworthy plesiomorphy for this genus as such a form is otherwise known only in Euglossini, Bombini, Electrobombini, and most Electrapini (it may be lacking in Thaumastobombus Engel and this should be more fully explored).

    In Thyreomelikertes the outer groove is faint for most of its course until it opens into the rutellar basin, when it is clearly present. There is potentially a thin longitudinal ridge dividing the rutellar basin, but this is so difficult to observe it may be an illusion. If present it is either a vestigial remnant of a dorsal ramus to the condylar ridge or, less likely, an intercalary ridge of the rutellar cap. What makes such a faint ridge evident is the proximal lower part of the rutellar basin, which appears separate from the upper part, thereby seemingly forming a wide inferior groove. This requires exploration in more perfectly preserved material.

    In Thyreomelikertes, the first and second preapical teeth are separated by a rather broad, shallow concavity as is sometimes the case in Meliponini. The loss of teeth in the mandibles of workers of Apini is a derived feature of that tribe, as is the secondary loss of teeth in some Meliponini. Euglossini and Bombini retain teeth, although those of Bombini are modified in that they are largely formed by notches in an otherwise continuous arch of the apical margin. Interestingly, the teeth of Succinapis, Aethemelikertes, Haidomelikertes Engel, and Melikertotes are similarly formed to those of bombines (fig. 45; see also Engel and Davis, 2021), albeit by somewhat deeper notches and with the first preapical tooth a broad continuation of the apical margin, while the second, if present, is either similar to the first or is a bit more toothlike in appearance. Aetheomelikertes, Haidomelikertes, Amelikertotes, Melikertes, and Succinapis have a single, blunt preapical tooth narrowly separated by a short, thin incision from the otherwise edentate apical margin and without a sweeping expansion to a second preapical tooth (Engel and Davis, 2021), except Haidomelikertes and Melissites appear to have a small second preapical tooth similar to the first and narrowly separated (at least in Haidomelikertes uraeus Engel and Melissites trigona Engel). This is distinctly not the case in Thyreomelikertes, which, as just noted, has the teeth formed like those of most stingless bees with broader concavities and sharper teeth (figs. 41, 42, 44). The sweeping expansion between the preapical tooth and a second preapical tooth of the mandible is a unique feature of Thyreomelikertes among Melikertini. The inner (mesal) surface of the melikertine mandible is weakly concave in its apical half, much like that of Meliponini and Apini, and reflects the manipulation of resins in melikertes (Engel and Davis, 2021), much like in Meliponini, and of wax in Apini.

    Features of the corbiculate worker mandible relative to Centridini are: (1) the great expansion of the pollex, concomitant with the apical expansion of the acetabular groove into a pollicar basin apically along the upper apical margin of the mandible; (2) the division of the acetabular groove (at least plesiomorphically) by an auxiliary ridge previously termed either as the “acetabular carina” in Euglossini or the “outer ridge upper distal branch” in Bombini (Michener and Fraser, 1978); and (3) narrowing and effacement of the acetabular groove proximally rather than extending from near the mandibular base to at least midlength, as exemplified in Centris Fabricius (Centridini) and many other apid lineages (e.g., Xylocopini, Manueliini, several Anthophorini, many Nomadinae). The presence of an auxiliary ridge is likely a further synapomorphy for Corbiculata, retained plesiomorphically in Euglossini, Bombini, Electrobombini, Electrapini, and at least some Melikertini, while it is largely effaced in a least some Melikertini, Meliponini, and entirely indiscernible in Apini. While it is easy to think that the reduction of ridges and grooves in Meliponini and Apini are a shared feature, at least some elements of this reduction are seemingly independent. For example, the presence of the auxiliary ridge in some Melikertini suggests that the further reduction and loss of this structure in Meliponini is perhaps independent of such a loss in Apini. Such a scenario seems unlikely based on existing phylogenies (Engel, 2001), with the condition in Apini seemingly an autapomorphic extreme. What is noteworthy is the variation observed among genera of Melikertini and the manner by which Thyreomelikertes is an outlier and seemingly more like meliponines than others in this group. Does this mean melikertines are paraphyletic to Meliponini? That remains to be tested more critically. Although there is abundant morphological and behavioral evidence from diverse sources to support Meliponini + Apini among extant corbiculate tribes (e.g., Plant and Paulus, 1987, 2016; Prentice, 1991; Schultz et al., 1999, 2001; Engel, 2001; Noll, 2002; Cardinal and Packer, 2007; Porto et al., 2016, 2017), molecular data consistently place Bombini and Meliponini together (e.g., Sheppard and McPheron, 1991; Cameron and Mardulyn, 2001; Kawakita et al., 2008; Rodriguez-Serrano et al., 2012; Bossert et al., 2019). Might the more bombinelike dentition of some melikertines (e.g., Aethemelikertes, Succinapis) be a holdover from a common ancestor with Bombini and Electrobomini, which then became further modified in Meliponini, with a form reminiscent of Apini merely owing to a reduction/loss as a result of specialization for manipulating resin/wax? Alternatively, the bombinelike dentition could be plesiomorphic for the entire corbiculate clade, as evidence of the same form in Protobombus. Might Protobombus (perhaps with Electrapis Cockerell) be an even earlier diverging branch between the divergence of Euglossini and the eusocial corbiculate lineages? These are all worthy hypotheses to be tested, but for the moment the existing data and pattern from the mandibles, albeit not explored cladistically, are instead perfectly consistent with the Darwinian null hypothesis of (Euglossini + (Bombini + (Meliponini + Apini))). In fact, the oblique apical margins of Thyreomelikertes, some other Melikertini, Meliponini, and Apini could be added as a potential synapomorphy for these groups relative to the broader and more transverse apical margins of Euglossini, Bombini, Electrobombini, and Electrapini. This would seem to further strengthen the traditional conception of relationships.

    FIGURE 43.

    Mandibles of Protobombus indecisus Cockerell (AMNH B-JH98) (Electrapini) in Eocene Baltic amber. A. Frontal view of clypeal apex, labrum, and mandibles. B. Outer-ventral oblique view.

    img-z69-1_01.jpg

    FIGURE 44.

    Mandibles of Melikertini and Meliponini. A. Drawing of mandible for Thyreomelikertes electrosinicus in Fushun amber. B. Cephalotrigona zexmeniae (Cockerell). C. Melipona (Melipona) quadrifasciata Lepeletier.

    img-z70-1_01.jpg

    Before any further conclusions are drawn a new study is necessary on the comparative morphology of corbiculate mandibles and particularly more refined accounts of the character states present in Electrobombini and Electrapini, the latter potentially paraphyletic with the inclusion of Thaumastobombus (subtribe Thaumastobombina), which could be a sister group to Apini. Once such a comprehensive consideration of living and fossil corbiculate mandibles is completed and as data on more fossil corbiculate bees continue to accumulate, a revised cladistic analysis will be needed to provide new clarity to their relationships. Such work must wait for the time being, as there are already a large number of additional corbiculate fossils awaiting description from other Eocene deposits (M.S.E., unpubl. data) and these should be included in any such comprehensive phylogenetic framework.

    Distribution and Paleoecology

    The bees of the Fushun coalfield deposit are both uniquely distinctive and simultaneously consistent with our expanding understanding of this paleofauna from Eocene paratropical to tropical forests. Both of the lineages recovered here are present in the Eocene forests of Europe (Engel, 2001; Engel and Davis, 2021) and melikertines are also present in the dipterocarp forest resins of the Cambay Basin in India (Engel et al., 2013). Glyptosmiini, as another representative of the early-diverging Megachilinae, correspond to the Glyptapini and Ctenoplectrellini more extensively known from Europe. Ctenoplectrellines are also present in the Eocene of North America (Engel and Davis, 2021; M.S.E., unpubl. data). These occurrences give a certain familiarity to the bees found in Fushun amber, highlighting the wide distribution of these bees, at least across the Holarctic during the Eocene and to some degree southward into Asia. At the same time, the two genera found in the Fushun coalfield are distinctive and not easily confused with the bees of other amber deposits. This is most evident with Glyptosmia, here considered to constitute a tribe separate from Glyptapini and Ctenoplectrellini, given its unique integumental sculpturing, thickened metatibial spurs, and mesoscutellar shield. Similarly, Thyreomelikertes stands apart from the melikertine diversity of European ambers. Of all of the potential forest bees living at the time, these bees are perhaps the most likely to be expected as they both seemingly belong to groups that collect resin. The mandible of Thyreomelikertes certainly corresponds to a bee that manipulated wax, propolis, and resins, as is hypothesized for other Melikertini (Engel and Davis, 2021) and is well known for modern Meliponini. Likewise, the mandible of Glyptapis has been hypothesized for use in resin collection (Gonzalez et al., 2019) and the identical form present in Glyptosmia presumably served an identical function. In essence, if any bees were to be captured in Fushun amber, then such groups are precisely those for which the greatest probability existed for entrapment. Other lineages of forest bees would have been far less likely to become ensnared, as since they would have had little or no contact with exuded resins given their biological profile. Thus, solitary ground-nesting bees, which were likely present, would have had virtually no chance of inclusion in amber. Furthermore, larger insects are often not captured as they can more easily free themselves if they ever come in contact with resin. Amber, while offering incredible insights into past biotas, is a highly skewed representation of any given fauna, preferentially sampling small, arboreal, and/or resin-associated arthropods from tropical, subtropical, and paratropical environments. Sometimes the resin, if it accumulated at the base of the trees, can encapsulate fauna in the litter in the vicinity or even at the edges of aquatic to semiaquatic habitats. There remain things amber cannot and will never resolve, particularly when it comes to the past diversity of bees. Given that the richest diversity of bees today is concentrated in xeric environments, which includes most of the early-diverging lineages of each family (Michener, 1979), it is likely the case that the greatest past diversity was concentrated in desert habitats, places notoriously poor for fossilization (Engel, 2004). Accordingly, it is no wonder that our insights into past bee diversity are significantly skewed, both in terms of habitat (i.e., forests) and body size as well as paleobiology. As noted, many of our species are documented from fossiliferous resins, which preferentially samples forest bees of smaller sizes and particularly those who collect resins as some part of their biology. It may be that paleomelittology can never give us the kind of fuller perspective that paleomyrmecology or other fields enjoy (e.g., Barden, 2017; Barden and Engel, 2021; Sosiak et al., 2024).

    FIGURE 45.

    Representative Baltic amber melikertines. A. Clypeal margin, labrum, and mandibles of Aethemelikertes emunctorii Engel. B. Slightly oblique facial view of Succinapis micheneri Engel. C. Detail of left mandible and right mandibular apex in S. micheneri

    img-z72-1_01.jpg

    The “Punctuated Equilibrium” of Paleomelittological Research

    Paleontology is about patience with incomplete data and imperfect material. A burst of new fossils, methods of study, or new technologies may scrape at the grime obscuring one's vision or one may confront years or decades before any glimmer appears. Paleomelittology is an excellent example of this. Although Fushun amber was explored in depth for 30 years, especially the Hymenoptera (e.g., Hong, 2002a), no bees were forthcoming, and then in a flash specimens were found that exposed this unique fauna at the far end of Asia during the Paleogene. The material from Fushun is a microcosm of the history of paleomelittological research, which is a story of fits and starts.

    Although the first fossil bees were “described” in the mid-1800s (Hope, 1836; Germar, 1849; Heer, 1849, 1865, 1867; Giebel, 1856; Menge, 1856; Motschulsky, 1856; Unger, 1867; Oustalet, 1870; Novák, 1877), aside from an isolated paper by Tosi (1896), the subject did not experience a true birth until Theodore D.A. Cockerell (1866–1948) made the first true systematic attempt to piece together a picture of fossil Apoidea, placing disparate elements into a framework at the time consistent with his broader understanding of bee diversity (e.g., Cockerell, 1906, 1907, 1909a, 1909b, 1925, 1931). Cockerell was one of a handful of leading melittologists of his era, and so his contributions to fossil bees benefitted from his considerable insight into the modern diversity. This is abundantly evident from his prescient remarks regarding Glyptapis and Ctenoplectrella (above) as well as other insights on bees he published over the course of more than a half century of study. Little changed subsequent to his efforts. Sundry smaller works appeared, but all suffered from the lack of understanding that Cockerell had so masterfully delivered to his interpretation of fossil bees (e.g., Meunier, 1915; Salt, 1931; Statz, 1931, 1934, 1936; Zeuner, 1931, 1938; Roussy, 1937, 1960; Théobald, 1937; Armbruster, 1938a, 1938b, 1938c, 1939; Piton, 1940). Although Friedrich E. Zeuner (1905–1963) and Francis J. Manning (1912–1966) had great hopes for a deeper understanding of fossil bees and planned a grand monograph of all specimens then known, neither lived to see their work finished and the published version of their notes (Zeuner and Manning, 1976) was sadly a step back from the perspective advanced by Cockerell many decades earlier. Nonetheless, flashes of further clarity would come during these paleomelittological dark ages, such as Simone Kelner-Pillault's (1925–1985) initial forays into the Baltic amber fauna and Alvaro Wille's (1928–2006) discovery of fossil stingless bees in Miocene Mexican and Dominican ambers (e.g., Wille, 1959, 1977; Wille and Chandler, 1964; Kelner-Pillault, 1969, 1970a, 1970b, 1974), refined by Michener (1982), and then significantly built upon by Camargo et al. (2000).

    One might say that our current epoch began with the publication of a meliponine from the Maastrichtian of New Jersey (Michener and Grimaldi, 1988a, 1988b; Engel, 2000), as within a few years of this work, especially between the years 1993 and 1996, there began a more or less steady stream of publications on fossil bees, with the report of fossil social bees from Eckfeld, Germany, and the expansion of the New World amber record beyond Meliponini into other families of the Anthophila (Lutz, 1993; Engel, 1995, 1996; Michener and Poinar, 1996; Rozen, 1996). From that time onward, Cockerell's valiant start has been pushed in directions he could only have dreamt and would certainly have excited him tremendously (as assumed from stories told of Cockerell and his passions and manner from the personal remembrances of the late Charles D. Michener to M.S.E.). And yet, despite the explosive growth of material and the many ways in which these have been studied—from higher resolution photography to CT scans, from geometric morphometrics to inclusion in cladistic analyses—the fossil record of bees remains woefully meager compared with other groups of ecologically dominant Hymenoptera (e.g., Engel, 2004; Ohl and Engel, 2007; Michez et al., 2012; Barden and Engel, 2021). Although already stated above it's worth repeating that many bees thrive in dryer places and these habitats are notoriously underrepresented in the fossil record as good preservation requires what they lack—water. Our sampling of the past for insects is decidedly biased toward forests, and often hot and humid forests, and this must be kept in mind when attempting to connect our few data points and extrapolate a bigger picture. We may never get past this hurdle when it comes to a fuller accounting of past bee diversity. Nonetheless, even if the total picture remains obscured, fortuitous discoveries have greatly expanded our understanding of paleomelittological diversity, including their intimate associations with fossil flowers (e.g., Wappler et al., 2015; Geier et al., in press), dramatically even within a wisp of time as brief as three decades. Paleomelittology is now being advanced by a fresh generation whose vision shall vastly exceed our own by scrutinizing new material to fill the vast geographic and temporal gaps (e.g., Lepeco and Melo, 2022; Celary et al., 2023) in our current understanding. We all work in the shadow of Cockerell's ghost, the king of paleomelittology and to whom we are merely pretenders, building on his excellent foundation to bring a fuller accounting of the history of Earth's preeminent pollinators.

    ACKNOWLEDGMENTS

    The authors are grateful to Chuijin Kong for generously donating Fushun amber to Capital Normal University thereby making this work possible. M.S.E. is immensely grateful to Dong Ren and Taiping Gao for generously providing the support for his visit to Beijing and Xi'an and their considerable hospitality during his stay, and to Zhen Wang for her friendly assistance during that visit. M.S.E. is further thankful to the late You-Chong Hong for stimulating correspondence in the late 1990s and early 2000s and sharing many interesting reprints of his and other Chinese works. M.S.E. is appreciative of Hans Jahnke for hosting his 1999 visit to the Institut und Museum für Geologie und Paläontologie, Göttingen, when Cockerell's holotype of G . mirabilis was rediscovered, and to the current curator, Alexander Gehler, and Uwe Kaulfuss for their kindness in providing more modern photographs of the specimen (presented here as fig. 2). Likewise, Agnieska Pierwola and David A. Grimaldi, AMNH, supplied new images for representative Baltic amber bees and bee mandibles, particularly the photographs presented here in figures 8A, 32–44, C; and Olivier Béthoux provided the photograph of the holotype of C . eocenica (fig. 8B). Production of synchrotron scans was supported by Argonne National Laboratory–Advanced Photo Source grant GUP-39093 (to M.S.E.), and we are thankful to Carmen Soriano and Ryan C. McKellar for assistance with segmentations. Lastly, we are grateful to David A. Grimaldi and an anonymous reviewer for their deeply helpful feedback on the manuscript, and Mary Knight for her exceptional assistance in arranging and editing the final version.

    Copyright © American Museum of Natural History 2024

    REFERENCES

    1.

    Armbruster, L. 1938a. Versteinerte Honigbienen aus dem obermiocänen Randecker Maar. Archiv für Bienenkunde 19 (1): 1–48. Google Scholar

    2.

    Armbruster, L. 1938b. Versteinerte Honigbienen aus dem obermiocänen Randecker Maar. Archiv für Bienenkunde 19 (2): 73–93. Google Scholar

    3.

    Armbruster, L. 1938c. Versteinerte Honigbienen aus dem obermiocänen Randecker Maar. Archiv für Bienenkunde 19 (3–4): 97–133. Google Scholar

    4.

    Armbruster, L. 1939. Eine miocäne Insectenfauna (mit meinem Präparierverfahren). Verhandlungen der VII Internationaler Kongress für Entomologie, Berlin [Band I–IV] 1938 (2): 1365–1371. Google Scholar

    5.

    Azar, D., S. Maksoud, C. Nammour, A. Nel, and B. Wang. 2018. A new trogiid genus from Lower Eocene Fushun amber (Insecta: Psocodea: Trogiomorpha). Geobios 51 (2): 101–106. Google Scholar

    6.

    Barden, P. 2017. Fossil ants (Hymenoptera: Formicidae): ancient diversity and the rise of modern lineages. Myrmecological News 24: 1–30. Google Scholar

    7.

    Barden, P., and M.S. Engel. 2021. Fossil social insects. In C.K. Starr (editor), Encyclopedia of social insects: 384–403. Cham: Springer, xxvi + 1049 pp. Google Scholar

    8.

    Bossert, S., et al. 2019. Combining transcriptomes and ultraconserved elements to illuminate the phylogeny of Apidae. Molecular Phylogenetics and Evolution 130: 121–131. Google Scholar

    9.

    Camargo, J.M.F., D.A. Grimaldi, and S.R.M. Pedro. 2000. The extinct fauna of stingless bees (Hymenoptera: Apidae: Meliponini) in Dominican amber: two new species and redescription of the male of Proplebeia dominicana (Wille and Chandler). American Museum Novitates 3293: 1–24. Google Scholar

    10.

    Cameron, S.A., and P. Mardulyn. 2001. Multiple molecular data sets suggest independent origins of highly eusocial behaviour in bees (Hymenoptera: Apinae). Systematic Biology 50 (2): 194–214. Google Scholar

    11.

    Cardinal, S., and L. Packer. 2007. Phylogenetic analysis of the corbiculate Apinae based on morphology of the sting apparatus (Hymenoptera: Apidae). Cladistics 23 (2): 99–118. Google Scholar

    12.

    Celary, W., B. Bojarski, and J. Szwedo. 2023. Buzzers from the past—the first Melikertini bee from Eocene Lublin amber. In J. Szwedo, B. Bojarski, K. Cierocka, and E. Sontag (editors), Fossil record in resins and sediments: books of abstracts: 31–32. Gdańsk: University of Gdańsk, vi + 122 pp. Google Scholar

    13.

    Cockerell, T.D.A. 1904. The bee genus Apista, etc. Canadian Entomologist 36 (12): 357. Google Scholar

    14.

    Cockerell, T.D.A. 1906. Fossil Hymenoptera from Florissant, Colorado. Bulletin of the Museum of Comparative Zoology 50 (2): 33–58. Google Scholar

    15.

    Cockerell, T.D.A. 1907. A fossil honey-bee. Entomologist 40 (533): 227–229. Google Scholar

    16.

    Cockerell, T.D.A. 1909a. Descriptions of Hymenoptera from Baltic amber. Schriften der physikalisch-ökonomischen Gesellschaft zu Königsberg 50 (1): 1–20. Google Scholar

    17.

    Cockerell, T.D.A. 1909b. Some European fossil bees. Entomologist 42 (559): 313–317. Google Scholar

    18.

    Cockerell, T.D.A. 1920. On South African bees, chiefly collected in Natal. Annals of the Durban Museum 2 (5): 247–262. Google Scholar

    19.

    Cockerell, T.D.A. 1925. Tertiary insects from Kudia River, Maritime Province, Siberia. Proceedings of the United States National Museum 68 (2606): 1–16, pls. 1–2. Google Scholar

    20.

    Cockerell, T.D.A. 1931. Insects from the Miocene (Latah) of Washington, II. Hymenoptera and Hemiptera. Annals of the Entomological Society of America 24 (2): 309–312, + pl. 1. Google Scholar

    21.

    Dehon, M., et al. 2019. Morphometric analysis of fossil bumble bees (Hymenoptera, Apidae, Bombini) reveals their taxonomic affinities. ZooKeys 891: 71–118. Google Scholar

    22.

    Eickwort, G.C. 1969. A comparative morphological study and generic revision of the augochlorine bees (Hymenoptera: Halictidae). University of Kansas Science Bulletin 48 (13): 325–524. Google Scholar

    23.

    Engel, M.S. 1995 [1996]. Neocorynura electra, a new fossil bee species from Dominican amber (Hymenoptera: Halictidae). Journal of the New York Entomological Society 103 (3): 317–323. Google Scholar

    24.

    Engel, M.S. 1996 [1997]. New augochlorine bees (Hymenoptera: Halictidae) in Dominican amber, with a brief review of fossil Halictidae. Journal of the Kansas Entomological Society (supp. 2) 69 (4): 334–345. Google Scholar

    25.

    Engel, M.S. 1998a. Fossil honey bees and evolution in the genus Apis (Hymenoptera: Apidae). Apidologie 29 (3): 265–281. Google Scholar

    26.

    Engel, M.S. 1998b. A new species of the Baltic amber bee genus Electrapis (Hymenoptera: Apidae). Journal of Hymenoptera Research 7 (1): 94–101. Google Scholar

    27.

    Engel, M.S. 1999a. Megachile glaesaria, the first megachilid bee fossil from amber (Hymenoptera: Megachilidae). American Museum Novitates 3276: 1–13. Google Scholar

    28.

    Engel, M.S. 1999b. The taxonomy of Recent and fossil honey bees (Hymenoptera: Apidae; Apis). Journal of Hymenoptera Research 8 (2): 165–196. Google Scholar

    29.

    Engel, M.S. 2000. A new interpretation of the oldest fossil bee (Hymenoptera: Apidae). American Museum Novitates 3296: 1–11. Google Scholar

    30.

    Engel, M.S. 2001. A monograph of the Baltic amber bees and evolution of the Apoidea (Hymenoptera). Bulletin of the American Museum of Natural History 259: 1–192. Google Scholar

    31.

    Engel, M.S. 2004. Geological history of the bees (Hymenoptera: Apoidea). Revista de Tecnologia e Ambiente 10 (2): 9–33. Google Scholar

    32.

    Engel, M.S. 2005. Family-group names for bees (Hymenoptera: Apoidea). American Museum Novitates 3476: 1–33. Google Scholar

    33.

    Engel, M.S. 2008. A new species of Ctenoplectrella in Baltic amber (Hymenoptera: Megachilidae). Acta Zoologica Academiae Scientiarum Hungaricae 54 (4): 319–324. Google Scholar

    34.

    Engel, M.S., and S.R. Davis. 2021. New genera of melikertine bees with facial modifications in Baltic amber (Hymenoptera: Apidae). Journal of Melittology 103: 1–52. Google Scholar

    35.

    Engel, M.S., and C.D. Michener. 2013. A minute stingless bee in Eocene Fushan [sic] amber from northeastern China (Hymenoptera: Apidae). Journal of Melittology 14: 1–10. Google Scholar

    36.

    Engel, M.S., and E.E. Perkovsky. 2006. An Eocene bee in Rovno amber, Ukraine (Hymenoptera: Megachilidae). American Museum Novitates 3506: 1–11. Google Scholar

    37.

    Engel, M.S., and C. Rasmussen. 2021. Corbiculate bees. In C.K. Starr (editor), Encyclopedia of social insects: 302–310. Cham: Springer, xxvi + 1049 pp. Google Scholar

    38.

    Engel, M.S., J. Ortega-Blanco, P.C. Nascimbene, and H. Singh. 2013. The bees of Early Eocene Cambay amber (Hymenoptera: Apidae). Journal of Melittology 25: 1–12. Google Scholar

    39.

    Engel, M.S., L.C.V. Breitkreuz, and M. Ohl. 2014. The first male of the extinct bee tribe Melikertini (Hymenoptera: Apidae). Journal of Melittology 30: 1–18. Google Scholar

    40.

    Engel, M.S., A.S. Alqarni, and M.A. Shebl. 2017. Discovery of the bee tribe Tarsaliini in Arabia (Hymenoptera: Apidae), with the description of a new species. American Museum Novitates 3877: 1–28. Google Scholar

    41.

    Engel, M.S., H.W. Herhold, S.R. Davis, B. Wang, and J.C. Thomas. 2021a. Stingless bees in Miocene amber of southeastern China (Hymenoptera: Apidae). Journal of Melittology 105: 1–83. Google Scholar

    42.

    Engel, M.S., C. Rasmussen, and V.H. Gonzalez. 2021b. Bees: phylogeny and classification. In C.K. Starr (editor), Encyclopedia of social insects: 93–109. Cham, Switzerland: Springer, xxvi + 1049 pp. Google Scholar

    43.

    Fedotova, Z.A., E.E. Perkovsky, A.J. Ross, and Q.-Q. Zhang. 2022. A new genus and species of gall midges [sic] the tribe Winnertziini (Diptera, Cecidomyiidae, Porricondylinae) from Lower Eocene Fushun amber from China. Palaeoentomology 5 (1): 90–98. Google Scholar

    44.

    Gadek, P.A., D.L. Alpers, M.M. Heslewood, and C.J. Quinn. 2000. Relationships within Cupressaceae sensu lato: a combined morphological and molecular approach. American Journal of Botany 87 (7): 1044–1057. Google Scholar

    45.

    Geier, C., et al. In press. The earliest large carpenter bee (Xylocopa) and its adhering pollen (Araliaceae, Theaceae). Palaeobiodiversity and Palaeoenvironments. Google Scholar

    46.

    Germar, E.F. 1849. Ueber einige Insekten aus Tertiärbildungen. Zeitschrift der Deutschen geologischen Gesellschaft 1: 52–66, + pl. II. Google Scholar

    47.

    Giebel, C.G. 1856. Fauna der Vorwelt mit steter Berücksichtigung der lebenden Thiere: Zweiter Band: Gliederthiere: Erste Abtheilung: Insecten und Spinnen. Leipzig: F.A. Brockhaus, xviii + 511 pp. Google Scholar

    48.

    Giłka, W., M. Zakrzewska, V. Baranov, B. Wang, and F. Stebner. 2016. The first fossil record of Nandeva Wiedenbrug, Reiss and Fittkau (Diptera: Chironomidae) in Early Eocene Fushun amber from China. Alcheringa 40 (3): 390–397. Google Scholar

    49.

    Gonzalez, V.H., and M.S. Engel. 2011. A new species of the bee genus Ctenoplectrella in middle Eocene Baltic amber (Hymenoptera, Megachilidae). ZooKeys 111: 41–49. Google Scholar

    50.

    Gonzalez, V.H, T. Griswold, C.J. Praz, and B.N. Danforth. 2012. Phylogeny of the bee family Megachilidae (Hymenoptera: Apoidea) based on adult morphology. Systematic Entomology 37 (2): 261–286. Google Scholar

    51.

    Gonzalez, V.H., G.T. Gustafson, and M.S. Engel. 2019. Morphological phylogeny of Megachilini and the evolution of leaf-cutter behavior in bees (Hymenoptera: Megachilidae). Journal of Melittology 85: 1–123. Google Scholar

    52.

    Gürsoy, D., F. De Carlo, X. Xiao, and C. Jacobsen. 2014. Tomopy: a framework for the analysis of synchrotron tomographic data. Journal of Synchrotron Radiation 21 (5):1188–1193. Google Scholar

    53.

    Heer, O. 1849. Die Insektenfauna der Tertiärgebilde von Œningen und von Radoboj in Croatien. Zweiter Abtheilung: Heuschrecken, Florfliegen, Aderflügeler, Schmetterlinge und Fliegen. Zürich: Druck von Zürcher und Furrer, iv + 264 pp., + pls. I–XVII. Google Scholar

    54.

    Heer, O. 1865. Die Urwelt der Schweiz. Zürich: F. Schulthess, xxix + 622 pp., + pls. I–XI. Google Scholar

    55.

    Heer, O. 1867. Fossile Hymenopteren aus Oeningen und Radoboj. Neue Denkschriften der allgemeinen Schweizerischen Gesellschaft für die gesammten Naturwissenschaften 22: 3–42, + 3 pls. Google Scholar

    56.

    Hong, Y.-C. 1981. Eocene fossil Diptera Insecta in amber of Fushun coalfield. Beijing: Geological Publishing House, vi+ 166 pp., + 27 pls. [In Chinese] Google Scholar

    57.

    Hong, Y.-C. 1982. Discovery of new fossil spiders in amber of Fushun coalfield. Scientia Sinica, Series B 25 (4): 431–436, + 1 pl. Google Scholar

    58.

    Hong, Y.-C. 2002a. Amber insects of China. Beijing: Beijing Science and Technology Press, 653 pp., + 48 pls. [in Chinese] Google Scholar

    59.

    Hong, Y.-C. 2002b. Atlas of amber insects of China. Zhengzhou: Henan Scientific and Technological Publishing House, 394 pp. [in Chinese] Google Scholar

    60.

    Hong, Y.-C., et al. 1974. Stratigraphy and palaeontology of Fushun coalfield, Liaoning Province. Acta Geologica Sinica 48 (2): 113–149. [in Chinese] Google Scholar

    61.

    Hope, F.W. 1836. Observations on succinic insects. Transactions of the Entomological Society of London 1: 133–147. Google Scholar

    62.

    ICZN [International Commission on Zoological Nomenclature]. 1999. International code of zoological nomenclature [4th ed.]. London: International Trust for Zoological Nomenclature, xxix+ 306 pp. Google Scholar

    63.

    Kawakita, A., J.S. Ascher, T. Sota, M. Kato, and D.W. Roubik. 2008. Phylogenetic analysis of the corbiculate bee tribes based on 12 nuclear protein-coding genes (Hymenoptera: Apoidea: Apidae). Apidologie 39 (1): 163–175. Google Scholar

    64.

    Kelner-Pillault, S. 1969. Les abeilles fossiles. Memorie della Società Entomologica Italiana 48: 519–534. Google Scholar

    65.

    Kelner-Pillault, S. 1970a. L'ambre Balte et sa faune entomologique avec description de deux apoides nouveaux. Annales de la Société Entomologique de France 6 (1): 3–24. Google Scholar

    66.

    Kelner-Pillault, S. 1970b. Une mélipone (s.l.) de l'ambre Balte (Hym. Apidae). Annales de la Société Entomologique de France 6 (2): 623–634. Google Scholar

    67.

    Kelner-Pillault, S. 1974. État d'évolution des apides de l'ambre Balte. Annales de la Société Entomologique de France 10 (3): 623–634. Google Scholar

    68.

    Kotthoff, U., T. Wappler, and M.S. Engel. 2013. Greater past disparity and diversity hints at ancient migrations of European honey bee lineages into Africa and Asia. Journal of Biogeography 40 (10): 1832–1838. Google Scholar

    69.

    Krüger, E. 1920. Beiträge zur Systematik und Morphologie der mittel-europäischen Hummeln. Zoologische Jahrbücher, Abteilung für Systematik, Geographie und Biologie der Tiere 42 (5–6): 289–464, + pls. 3–7. Google Scholar

    70.

    Lepeco, A., and G.A.R. Melo. 2022. Another piece in the puzzle: a fossil stingless bee from Ethiopian amber (Apidae, Meliponini). Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 304 (2): 151–157. Google Scholar

    71.

    Litman, J.R., T. Griswold, and B.N. Danforth. 2016. Phylogenetic systematics and a revised generic classification of anthidiine bees (Hymenoptera: Megachilidae). Molecular Phylogenetics and Evolution 100: 183–198. Google Scholar

    72.

    Lutz, H. 1993. Eckfeldapis electrapoides nov. gen. n. sp., eine “Honigbiene” aus dem Mittel-Eozän des “Eckfelder Maares” bei Manderscheid/Eifel, Deutschland (Hymenoptera: Apidae, Apinae). Mainzer Naturwissenschaftliches Archiv 31: 177–199. Google Scholar

    73.

    Menge, A. 1856. Lebenszeichen vorweltlicher, im Bernstein eingeschlossener Thiere. Danzig: A.W. Kafeman [Programm der öffentlichen Prüfung der Schüler der Petrischule], [i] + 32 pp. Google Scholar

    74.

    Meunier, F. 1915. Über einige fossile Insekten aus den Braunkohlenschichten (Aquitanien) von Rott (Siebengebirge). Zeitschrift der Deutschen Geologischen Gesellschaft. A. Abhandlungen 67 (3): 205–217, + pls. xxi–xxv. Google Scholar

    75.

    Michener, C.D. 1944. Comparative external morphology, phylogeny, and a classification of the bees (Hymenoptera). Bulletin of the American Museum of Natural History 82 (6): 151–326. Google Scholar

    76.

    Michener, C.D. 1979. Biogeography of the bees. Annals of the Missouri Botanical Garden 66 (3): 277–347. Google Scholar

    77.

    Michener, C.D. 1982. A new interpretation of fossil social bees from the Dominican Republic. Sociobiology 7 (1): 37–45. Google Scholar

    78.

    Michener, C.D. 1990. Classification of the Apidae (Hymenoptera). University of Kansas Science Bulletin 54 (4): 75–163. Google Scholar

    79.

    Michener, C.D. 2007. The bees of the world [2nd ed.]. Baltimore: Johns Hopkins University Press, xvi + [i] + 953 pp., + 20 pls. Google Scholar

    80.

    Michener, C.D., and A. Fraser. 1978. A comparative anatomical study of mandibular structure in bees (Hymenoptera: Apoidea). University of Kansas Science Bulletin 51 (14): 463–482. Google Scholar

    81.

    Michener, C.D., and L. Greenberg. 1980. Ctenoplectridae and the origin of long-tongued bees. Zoological Journal of the Linnean Society 69 (3): 183–203. Google Scholar

    82.

    Michener, C.D., and D.A. Grimaldi. 1988a. A Trigona from Late Cretaceous amber of New Jersey (Hymenoptera: Apidae: Meliponinae). American Museum Novitates 2917: 1–10. Google Scholar

    83.

    Michener, C.D., and D.A. Grimaldi. 1988b. The oldest fossil bee: apoid history, evolutionary stasis, and antiquity of social behavior. Proceedings of the National Academy of Sciences of the United States of America 85 (17): 6424–6426. Google Scholar

    84.

    Michener, C.D., and G.[O.] Poinar, Jr. 1996 [1997]. The known bee fauna of the Dominican amber. Journal of the Kansas Entomological Society (suppl. 2) 69 (4): 353–361. Google Scholar

    85.

    Michez, D., M. Vanderplanck, and M.S. Engel. 2012. Fossil bees and their plant associates. In S. Patiny (editor), Evolution of plant-pollinator relationships: 103–164. Cambridge: Cambridge University Press, xv + 477 +[6] pp. Google Scholar

    86.

    Milliron, H.E. 1971. A monograph of the Western Hemisphere bumblebees (Hymenoptera: Apidae; Bombinae). I. The genera Bombus and Megabombus subgenus Bombias. Memoirs of the Entomological Society of Canada 82: 1–80. Google Scholar

    87.

    Motschulsky, V., de. 1856. Voyages. Lettres de M. de Motschulsky à M. Ménétriés. Études Entomologiques 5: 3–38. Google Scholar

    88.

    Nel, A., and J.F. Petrulevičius. 2003. New Palaeogene bees from Europe and Asia. Alcheringa 27 (4): 277–293. Google Scholar

    89.

    Noll, F.B. 2002. Behavioral phylogeny of corbiculate Apidae (Hymenoptera; Apinae), with special reference to social behavior. Cladistics 18 (2): 137–153. Google Scholar

    90.

    Novák, O. 1877. Fauna der Cyprisschiefer des Egerer Tertiärbeckens. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften: Mathematisch-Naturwissenschaftliche Classe 76 (Erste Abtheilungen: 7): 71–96, + pls. I–III. Google Scholar

    91.

    Ohl, M., and M.S. Engel. 2007. Die Fossilgeschichte der Bienen und ihrer nächsten Verwandten (Hymenoptera: Apoidea). Denisia 20: 687–700. Google Scholar

    92.

    Orr, M.C., et al. 2022. Phylogenomic interrogation revives an overlooked hypothesis for the early evolution of the bee family Apidae (Hymenoptera: Apoidea), with a focus on the subfamily Anthophorinae. Insect Systematics and Diversity 6 (4): 1–15. Google Scholar

    93.

    Oustalet, M.E. 1870. Recherches sur les insectes fossiles des terrains tertiaires de la France. Première partie. Insectes fossiles de l'Auvergne. Annales des Sciences Géologiques 2 (3): 1–178, + pls. 1–6. Google Scholar

    94.

    Parizotto, D.R., D. Urban, and G.A.R. Melo. 2022. Phylogeny and generic classification of the Anthidiini bees from the Neotropical region (Hymenoptera: Apidae). Zoological Journal of the Linnean Society 194 (1): 80–101. Google Scholar

    95.

    Peters, D.S. 1972. Über die Stellung von Aspidosmia Brauns 1926 nebst allgemeinen Erörterungen der phylogenetischen Systematik der Megachilidæ (Insecta, Hymenoptera, Apoidea). Apidologie 3 (2): 167–186. Google Scholar

    96.

    Ping, C. 1931. On a blattoid insect in the Fushun amber. Bulletin of the Geological Society of China 11 (2): 205–207. Google Scholar

    97.

    Piton, L.-E. 1940. Paléontologie du gisement Éocène de Menat (Puy-de-Dôme) (flore et fauna). Mémoires de la Société d'Histoire Naturelle d'Auvergne 1: 1–303, + 26 pls. Google Scholar

    98.

    Plant, J.D., and H.F. Paulus. 1987. Comparative morphology of the postmentum of bees (Hymenoptera: Apoidea) with special remarks on the evolution of the lorum. Zeitschrift für Zoologische Systematik und Evolutionsforschung 25: 81–103. Google Scholar

    99.

    Plant, J.D., and H.F. Paulus. 2016. Evolution and phylogeny of bees: a review and a cladistic analysis in light of morphological evidence (Hymenoptera, Apoidea). Zoologica 161: 1–368. Google Scholar

    100.

    Porto, D.S., L. Vilhelmsen, and E.A.B. Almeida. 2016. Comparative morphology of the mandibles and head structures of corbiculate bees (Hymenoptera: Apidae: Apini). Systematic Entomology 41 (2): 339–368. Google Scholar

    101.

    Porto, D.S., E.A.B. Almeida, and L. Vilhelmsen. 2017. Comparative morphology of internal structures of the mesosoma of bees with an emphasis on the corbiculate clade (Apidae: Apini). Zoological Journal of the Linnean Society 179 (1): 303–337. Google Scholar

    102.

    Praz, C.J., et al. 2008. Phylogeny and biogeography of bees of the tribe Osmiini (Hymenoptera: Megachilidae). Molecular Phylogenetics and Evolution 49 (1): 185–197. Google Scholar

    103.

    Prentice, M. 1991. Morphological analysis of the tribes of Apidae. In D.R. Smith (editor), Diversity in the genus Apis: 51–69. Boulder: Westview Press, xiv +265 pp. Google Scholar

    104.

    Prokop, J., M. Dehon, D. Michez, and M.S. Engel. 2017. An Early Miocene bumble bee from northern Bohemia (Hymenoptera, Apidae). ZooKeys 710: 43–63. Google Scholar

    105.

    Rasmussen, C., J.C. Thomas, and M.S. Engel. 2017. A new genus of Eastern Hemisphere stingless bees (Hymenoptera: Apidae), with a key to the supraspecific groups of Indomalayan and Australasian Meliponini. American Museum Novitates 3888: 1–33. Google Scholar

    106.

    Richards, O.W. 1927. The specific characters of the British humblebees (Hymenoptera). Transactions of the Entomological Society of London 75 (2): 233–268, + pls. xxii–xxv. Google Scholar

    107.

    Rodriguez-Serrano, E., O. Inostroza-Michael, J. Avaria-Llautureo, and C.E. Hernandez. 2012. Colony size evolution and the origin of eusociality in corbiculate bees (Hymenoptera: Apinae). PLoS ONE 7 (7): e40838 [1–8]. Google Scholar

    108.

    Roig-Alsina, A., and C.D. Michener. 1993. Studies of the phylogeny and classification of long-tongued bees (Hymenoptera: Apoidea). University of Kansas Science Bulletin 55 (4): 123–162. Google Scholar

    109.

    Roussy, L. 1937. Contributions a l'étude de l'abeille Tertiare, de ses parasites et de ses ennemis. Gazette Apicole (Montfavet) 388: 49–72. Google Scholar

    110.

    Roussy, L. 1960. Insectes et abeilles fossils de l'ambre de Sicile: Migrations, localisations, peuplement du Nouveau Monde, de l'Australie, de la Nouvelle Zélande. La Gazette Apicole (Montfavet) 635: 5–8. Google Scholar

    111.

    Rozen, J.G., Jr. 1996 [1997]. A new species of the bee Heterosarus from Dominican amber (Hymenoptera: Andrenidae; Panurginae). Journal of the Kansas Entomological Society (suppl. 2) 69 (4): 346–352. Google Scholar

    112.

    Rust, J., et al. 2010. Biogeographic and evolutionary implications of a diverse paleobiota in amber from the Early Eocene of India. Proceedings of the National Academy of Sciences of the United States of America 107 (43): 18360–18365. Google Scholar

    113.

    Salt, G. 1931. Three bees from Baltic amber. Bernstein-Forschungen 2: 136–147. Google Scholar

    114.

    Schultz, T.R., M.S. Engel, and M. Prentice. 1999. Resolving conflict between morphological and molecular evidence for the origin of eusociality in the corbiculate bees (Hymenoptera: Apidae): a hypothesis-testing approach. University of Kansas Natural History Museum Special Publication 24: 125–138. Google Scholar

    115.

    Schultz, T.R., M.S. Engel, and J.S. Ascher. 2001. Evidence for the origin of eusociality in the corbiculate bees (Hymenoptera: Apidae). Journal of the Kansas Entomological Society 74 (1): 10–16. Google Scholar

    116.

    Sheppard, W.S., and B.A. McPheron. 1991. Ribosomal DNA diversity in Apidae. In D.R. Smith (editor), Diversity in the genus Apis: 89–102. Boulder: Westview Press, xiv +265 pp. Google Scholar

    117.

    Shuckard, W.E. 1866. British bees: an introduction to the study of the natural history and economy of the bees indigenous to the British Isles. London: Lovell, Reeve, & Co., xiii + ii] + 371 pp., + pls. i–xvi. Google Scholar

    118.

    Silveira, F.A. 1993. Phylogenetic relationships of the Exomalopsini and Ancylini (Hymenoptera: Apidae). University of Kansas Science Bulletin 55 (5): 163–173. Google Scholar

    119.

    Soriano, C., et al. 2010. Synchrotron x-ray imaging of inclusions in amber. Comptes Rendus Palevol 9 (6–7): 361–368. Google Scholar

    120.

    Sosiak, C., P. Cockx, P.A. Suarez, R. McKellar, and P. Barden. 2024. Prolonged faunal turnover in earliest ants revealed by North American Cretaceous amber. Current Biology 34 (8): 1755–1761. Google Scholar

    121.

    Statz, G. 1931. Eine neue Bienenart aus Rott am Siebengebirge. Ein Beitrag zur Kenntnis der fossilen Honigbienen. Wissenschaftlichen Mitteilungen des Vereins für Natur- und Heimatkunde in Köln am Rhein 1 (2): 39–60. Google Scholar

    122.

    Statz, G. 1934. Neue Beobachtungen über fossile Bienen aus dem Tertiär von Rott am Siebengebirge. Archiv für Bienenkunde 15 (1): 1–10, +2 pls. Google Scholar

    123.

    Statz, G. 1936. Ueber alte und neue fossile Hymenopterenfunde aus den tertiären Ablagerungen von Rott am Siebengebirge. Decheniana 93: 256–312, +13 pls. Google Scholar

    124.

    Stebner, F., R. Szadziewski, and B. Wang. 2016. Biting midges (Diptera: Ceratopogonidae) in Fushun amber reveal further biotic links between Asia and Europe during the Eocene. Palaeontologica Electronica 19 (3): 31A [1–9]. Google Scholar

    125.

    Théobald, N. 1937. Les insectes fossiles de terrains Oligocènes de France. Nancy: Mémoires de la Société des Sciences de Nancy, 473 pp., +39 pls., [i: errata]. Google Scholar

    126.

    Tosi, A. 1896. Di un nuovo genre di Apiaria fossile nell'amba di Sicilia (Meliponorytes succiniM . sicula). Rivista Italiana di Paleontologia 2 (6): 352–356, + pl. VI. Google Scholar

    127.

    Tran, N.T., M.S. Engel, M.T. Nguyen, C.Q. Nguyen, and L.T.P. Nguyen. 2024. The cuckoo bee genus Pasites in central Vietnam (Hymenoptera: Apidae). American Museum Novitates 4022: 1–12. Google Scholar

    128.

    Trunz, V., L. Packer, J. Vieu, N. Arrigo, and C.J. Praz. 2016. Comprehensive phylogeny, biogeography and new classification of the diverse bee tribe Megachilini: can we use DNA barcodes in phylogenies of large genera? Molecular Phylogenetics and Evolution 103: 245–259. Google Scholar

    129.

    Unger, F. 1867. Die fossile Flora von Kumi auf der Insel Euboea. Denkschriften der Kaiserlichen Akademie der Wissenschaften: Mathematisch-Naturwissenschaftlichen Classe 27 (1): 27–90, + pls. i–xvii. Google Scholar

    130.

    Wang, B., J. Kathirithamby, and M.S. Engel. 2016. The first twisted-wing parasitoid in Eocene amber from north-eastern China (Strepsiptera: Myrmecolacidae). Journal of Natural History 50 (21–22): 1305–1313. Google Scholar

    131.

    Wang, B., et al. 2014. A diverse paleobiota in Early Eocene Fushun amber from China. Current Biology 24 (14): 1606–1610. Google Scholar

    132.

    Wang, B., et al. 2021. The mid-Miocene Zhangpu biota reveals an outstandingly rich rainforest biome in East Asia. Science Advances 7 (18): eabg0625 [1–7]. Google Scholar

    133.

    Wappler, T., and M.S. Engel. 2003. The middle Eocene bee faunas of Eckfeld and Messel, Germany (Hymenoptera, Apoidea). Journal of Paleontology 77 (5): 908–921. Google Scholar

    134.

    Wappler, T., T. De Meulemeester, A.M. Aytekin, D. Michez, and M.S. Engel. 2012. Geometric morphometric analysis of a new Miocene bumble bee from the Randeck Maar of southwestern Germany (Hymenoptera: Apidae). Systematic Entomology 37 (4): 784–792. Google Scholar

    135.

    Wappler, T., C.C. Labandeira, M.S. Engel, R. Zetter, and F. Grímsson. 2015. Specialized and generalized pollen-collection strategies in an ancient bee lineage. Current Biology 25 (23): 3092–3098. Google Scholar

    136.

    Wedmann, S., T. Wappler, and M.S. Engel. 2009. Direct and indirect fossil records of megachilid bees from the Paleogene of Central Europe (Hymenoptera: Megachilidae). Naturwissenschaften 96 (6): 703–712. Google Scholar

    137.

    Wille, A. 1959. A new fossil stingless bee (Meliponini) from the amber of Chiapas, Mexico. Journal of Paleontology 33 (5): 849–852, +1 pl. Google Scholar

    138.

    Wille, A. 1977. A general review of the fossil stingless bees. Revista de Biología Tropical 25 (1): 43–46. Google Scholar

    139.

    Wille, A., and L.C. Chandler. 1964. A new stingless bee from the Tertiary amber of the Dominican Republic (Hymenoptera; Meliponini). Revista de Biología Tropical 12 (2): 187–195. Google Scholar

    140.

    Yarrow, I.H.H. 1954. Some observations on the genus Bombus, with special reference to Bombus cullumanus (Kirby) (Hym. Apidae). Journal of the Society for British Entomology 5 (1): 34–39. Google Scholar

    141.

    Zeuner, F.E. 1931. Die Insektenfauna des Böttinger Marmors: Eine systematische und paläobiologische Studie. Fortschritte der Geologie und Palaeontologie 9 (28): 247–406, + 19 pls. Google Scholar

    142.

    Zeuner, F.E. 1938. Die Insektenfauna des Mainzer Hydrobienkalks. Palaeontologische Zeitschrift 20: 104–159, + pls. 13–17. Google Scholar

    143.

    Zeuner, F.E., and F.J. Manning. 1976. A monograph on fossil bees (Hymenoptera: Apoidea). Bulletin of the British Museum (Natural History), Geology 27 (3): 149–268, + 4 pls. Google Scholar

    144.

    Zhang, Q.-Q., A. Nel, D. Azar, and B. Wang. 2016. New Chinese psocids from Eocene Fushun amber (Insecta: Psocodea). Alcheringa 40 (3): 366–372. Google Scholar
    Michael S. Engel and Jiaying Xie "The Bee Fauna of Eocene Fushun Amber (Hymenoptera: Apoidea)," Bulletin of the American Museum of Natural History 2024(469), 1-80, (23 October 2024). https://doi.org/10.1206/0003-0090.469.1.1
    Published: 23 October 2024
    Back to Top