Search

Richter, Adrian, Boudinot, Brendon, Yamamoto, Shûhei, Katzke, Julian (2022): The First Reconstruction of the Head Anatomy of a Cretaceous Insect, †Gerontoformica gracilis (Hymenoptera: Formicidae), and the Early Evolution of. Insect Systematics and Diversity 6 (5): 1-80, DOI: 10.1093/isd/ixac013, URL: http://dx.doi.org/10.1093/isd/ixac013

GENUS † GERONTOFORMICA NEL & PERRAULT, 2004 = † Sphecomyrmodes Engel & Grimaldi, 2005 junior syn.: Barden & Grimaldi (2016): 518. TAXONOMIC SYNOPSIS OF † GERONTOFORMICA I. Species group of cretacica (newly recognized) (Charentese amber) [Note 1]: 1. † G. cretacica Nel & Perrault, 2004 nomen dubium (new status) [Note 2] – Type species of † Gerontoformica Nel & Perrault, 2004 2. † G. occidentalis (Perrichot et al., 2008) II. Species group of gracilis (newly recognized) (Kachin amber): 3. † G. gracilis (Barden & Grimaldi, 2014) 4. † G. robusta (Barden & Grimaldi, 2014) 5. † G. spiralis (Barden & Grimaldi, 2014) [Note 3] 6. † G. subcuspis (Barden & Grimaldi, 2014) [Note 3] III. Species group of pilosa (Kachin amber): 7. † G. sternorhabda sp. nov. 8. † G. contega (Barden & Grimaldi, 2014) 9. † G. magna (Barden & Grimaldi, 2014) 10. † G. pilosa (Barden & Grimaldi, 2014) IV. Incertae sedis to species group within genus (Kachin amber): (11.) † G. orientalis (Engel & Grimaldi, 2005) nomen dubium (new status) [Note 4] – Type species of † Sphecomyrmodes Engel & Grimaldi, 2005 (12.) † G. rugosa (Barden & Grimaldi, 2014) nomen dubium (new status) [Note 5] (13.) † G. tendir (Barden & Grimaldi, 2014) nomen dubium (new status) [Notes 5, 6] Notes on classification Note 1: Under the label of ‘ orientalis group’, Boudinot et al. (2020b) previously united the cretacica and gracilis groups with the species † G. orientalis. The cretacica and gracilis groups are here divided based on the distinct forms of their petioles. The species † G. orientalis is of insufficient preservation for specieslevel identification (see Note 4 below). Note 2: The holotype of † Gerontoformica cretacica is poorly preserved, as noted by Barden & Grimaldi (2016), with obvious distortions and elongation of the scapes and flagella, and little detail of the body visible. Due to this unfortunate circumstance, we newly consider † G. cretacica to be a nomen dubium. It remains in the cretacica species group of † Gerontoformica, because it is the type species of the genus. However, it will be highly desirable to clarify the identity of this species through examination and description of more Charentese ants. At present, the only distinction between † G. cretacica and the co-eval † G. occidentalis that is not apparently affected by preservation is body size, with the former having a longer mesosoma than the latter (2.1 mm vs. 1.4 mm, respectively). Body size is, of course, highly variable among nestmates of crown ants, thus is a weak diagnostic trait when used in isolation and without the quantification of variation across conspecific individuals. Note 3: † Gerontoformica spiralis and † G. subcuspis were difficult to separate in the present study based on the available anatomical evidence and may be conspecific. Specifically, we observe that, in addition to conditions outlined in the key (see that section below), the two species are highly similar in the following conditions, which we only roughly characterize here: (1) proportions and fine details of the head, including frontal carina shape; (2) the degree of mesonotal, metanotal and propodeal convexity; (3) the width of separation between the meso- and metanota plus metanotum and propodeum; (4) the shape and proportions of the petiolar node; and (5) the form of the third abdominal (first ‘gastral’) segment. It is possible that † G. spiralis and † G. subcuspis represent the smaller and larger ranges of body size of a single species, with the former reported to have a total body length of 4.22–5.22 mm and the latter 5.35–5.76 mm (Barden & Grimaldi, 2014). Our focal uncertainties relate to the shapes and setational patterns of the tarsi of † G. subcuspis, and the form of the subpetiolar process and prora of † G. spiralis. A potential feature separating the two species is the distance of the toruli from the posterior clypeal margin, which appears to be narrower in † G. spiralis relative to † G. subcuspis, but this could be a visual artefact caused by the apparent light distortion in the holotype image of † G. spiralis. We recommend further re-evaluation of these two species, ideally using µ-CT and additional light photography to resolve the uncertainties of the tarsi, toruli and sternal processes of the metasoma. See also Note 1 on the diagnosis of † G. gracilis. Note 4: † Gerontoformica orientalis, the type species of † Sphecomyrmodes, is identifiable as † Gerontoformica relative to † Sphecomyrma Wilson & Brown, 1967 by the absence of the anteromedian clypeal process and presence of traction setae/ chaetae along the anterior clypeal margin, as recognized in the original description (Engel & Grimaldi, 2005) and illustrated in Boudinot et al. (2020b). However, in † Gerontoformica, the species † G. orientalis is unidentifiable due to poor preservation. No details of the head are clearly observable, except for the antenna, mandibles and anterior clypeal margin, while the mesosoma appears strongly distorted, the petiole is obscured, and the metasoma is mostly disarticulated. Boudinot et al. (2020b) placed † G. orientalis in the orientalis species group along with the type species of † Gerontoformica based on the absence of the cinctus of abdominal segment IV. As a more refined placement is not possible, we newly consider this species to be a nomen dubium. Note 5: † Gerontoformica rugosa and † G. tendir are newly considered as nomina dubia due to their poor preservation, being strongly desiccated and thus distorted. Both species appear to have some degree of abdominal segment III petiolation, as observed in the three confirmed members of the pilosa group, but it cannot be determined whether this is natural or exaggerated due to preservation. It is possible, but not yet determinable, that † G. rugosa is conspecific with † G. gracilis. That † G. rugosa or † G. tendir do have sculptured integument remains possible but requires substantiation via additional material of these species. We note that little surface texture variation has been explicitly documented among stem ants thus far. Note 6: † Gerontoformica tendir was defined by Barden & Grimaldi (2014) as having a medial clypeal lobe. This anteromedian lobe not only bears traction setae/chaetae, as previously observed, but is also lateromedially broader and proximodistally shorter than that of † Sphecomyrma. Given the poor preservation, it is possible that the apparent lobate form may be due to distortion of the amber matrix, as the lobe consists of the entire medial portion of the clypeus, which is distinct from the lateral clypeal lobes. Based on direct examination of the holotype, it appears that there is a transverse mesonotal carina in † G. tendir, but this also requires re-evaluation. Without additional specimens having an exaggerated and broadly, medially lobate clypeus, we remain uncertain about the identity of the species. A state of potential value for confirming the identity of † G. tendir from additional material is the absence of teeth on the pretarsal claws, as illustrated by Barden & Grimaldi (2014). Remarks The genus † Gerontoformica currently consists of 13 species, including the newly described † G. sternorhabda. The holotypes of nine of these species are sufficiently preserved for species-level identification, with eight of these being sufficiently defined given the potential synonymy of two (see Note 3 above). At the generic level, † Gerontoformica differs from † Sphecomyrma by presence of traction setae/chaetae along the anterior clypeal margin and absence of a narrow anteromedian clypeal lobe. These two conditions were used by Engel & Grimaldi (2005) to establish the genus † Sphecomyrmodes, which was synonymized under † Gerontoformica by Barden & Grimaldi (2016) after examination of the holotype of the type species of the latter taxon. Collectively, † Gerontoformica and † Sphecomyrmodes have been revised piecemeal after the former’s establishment by Nel et al. (2004). Specifically, Barden & Grimaldi (2014) added nine species of † Sphecomyrmodes, Barden & Grimaldi (2016) transferred all species of † Sphecomyrmodes to † Gerontoformica, and Boudinot et al. (2020b) moved one species to † Myanmyrma and recognized two morphologically defined groups of species in the genus. To understand the shifting boundaries of † Gerontoformica in the light of the present µ-CT-driven study, we detail the species-level history of the genus. Species attributed to the genus are distributed in Kachin amber (11 total) and Charentese amber (two total). The Charentese species, † G. cretacica and † G. occidentalis, were described four years apart by Nel et al. (2004) and Perrichot et al. (2008), respectively. Perrichot et al. described the smaller-bodied † G. occidentalis as † Sphecomyrmodes in comparison to the type species of that genus – † G. orientalis from Kachin amber – without comparison to † G. cretacica. Unfortunately, given the current status of morphological knowledge, the holotypes of both † G. orientalis and † G. cretacica are too poorly preserved to allow for confident specieslevel identification. However, both Charentese species uniquely share a distinct, nearly squamiform petiolar shape, with the node being anteroposteriorly narrow and dorsoventrally tall; this indicates that the Charentese species are closely related to one another, relative to the Kachin species. For this reason, we group † G. cretacica and † G. occidentalis together in the cretacica species group. It is possible that † G. cretacica and † G. occidentalis are conspecific, but any taxonomic action should wait for the accumulation and processing of more material from the Charentese formation. The identifiable Kachin species of † Gerontoformica, i.e. excluding † G. orientalis, † G. rugosa and † G. tendir, are evenly distributed in the G. gracilis and G. pilosa groups, with four species each. All species of the pilosa group are highly distinct in body form and setation, with † G. sternorhabda being an outlier, having the smallest body size and least pronounced constriction of the fourth abdominal segment, in addition to other defining features (see the species diagnosis below). In contrast, species of the newly recognized gracilis group are all largely similar to one another, without easily recognizable diagnostic structures. The most distinct Kachin species of the gracilis group is † G. robusta, which has a boxy mesosoma, with the meso- and metanota forming a nearly linear dorsal margin in lateral view. The other species of the gracilis group have a multi-humped profile due to the bulgelike development of the meso- and metanota and are otherwise similar to one another (see Note 1 on the diagnosis of † G. gracilis)., Published as part of Boudinot, Brendon E., Richter, Adrian, Katzke, Julian, Chaul, Júlio C. M., Keller, Roberto A., Economo, Evan P., Beutel, Rolf Georg & Yamamoto, Shûhei, 2022, Evidence for the evolution of eusociality in stem ants and a systematic revision of † Gerontoformica (Hymenoptera: Formicidae), pp. 1355-1389 in Zoological Journal of the Linnean Society 195 on pages 1361-1363, DOI: 10.1093/zoolinnean/zlab097, http://zenodo.org/record/6994456, {"references":["Nel A, Perrault G, Perrichot V, Neraudeau D. 2004. The oldest ant in the Lower Cretaceous amber of Charente- Maritime (SW France) (Insecta: Hymenoptera: Formicidae). Geologica Acta 2: 23 - 29.","Engel MS, Grimaldi DA. 2005. Primitive new ants in Cretaceous amber from Myanmar, New Jersey, and Canada (Hymenoptera: Formicidae). American Museum Novitates 3485: 1 - 23.","Barden P, Grimaldi D. 2016. Adaptive radiation in socially advanced stem-group ants from the Cretaceous. Current Biology 26: 1 - 7.","Perrichot V, Nel A, Neraudeau D, Lacau S, Guyot T. 2008. New fossil ants in French Cretaceous amber (Hymenoptera: Formicidae). Naturwissenschaften 95: 91 - 97.","Barden P, Grimaldi D. 2014. A diverse ant fauna from the mid-Cretaceous of Myanmar (Hymenoptera: Formicidae). PLoS One 9: e 93627.","Boudinot BE, Perrichot V, Chaul JCM. 2020 b. † Camelosphecia gen. nov., lost ant-wasp intermediates from the mid-Cretaceous (Hymenoptera, Formicoidea). ZooKeys 1005: 21 - 55."]}

† GERONTOFORMICA GRACILIS (BARDEN & GRIMALDI, 2014) (FIGS 1A, B, 2A, B, D, E, G, H, J, K, 8, 9, 10, 11 (1’-2), 12(7-1); SUPPORTING INFORMATION, FIG. S3) † Sphecomyrmodes gracilis Barden & Grimaldi, 2014: 4–7, figs 2, 10B, 11C, D (wingless female, Kachin amber, JZC-Bu324A, AMNH). Combination in † Gerontoformica: Barden & Grimaldi (2016): 518, suppl. info. p. 16. Diagnosis (wingless female) Similarly identifiable as † G. sternorhabda to the genus † Gerontoformica (I–III above), including absence of the anteromedian lobate clypeal process [Note 1]. For the diagnosis and redescription, see Figures 2, 8 and 10 primarily. I. Among † Gerontoformica, with the following unique character : (1) proximal labial palpomere with a distinct, apicomedially situated lobate process (Fig. 7A, B) [Note 2]; II. Within † Gerontoformica, identifiable as a member of the gracilis species group : (2) mesoscutum without distinct, raised transverse carina separating mesoscutal and mesoscutellar regions (Fig. 10A); (3) tergum of petiolar node anteroposteriorly longer than dorsoventrally tall (high); and (4) abdominal segment IV without cinctus dividing the tergum and sternum into pre- and postsclerites; III. Within the gracilis species group, identified by the following : (5) petiole bun-like, with anteroposteriorly long node; vs. subsquamiform, with anteroposteriorly narrow node († G. cretacica and † G. occidentalis); (6) meso- and metanota bulging, with their dorsal silhouette bihumped; vs. meso- and metanota not bulging, with their dorsal silhouette forming an almost straight line († G.robusta);(7)transverse dorsal sulci of mesosoma broad, separating the meso- and metanota and the metanotum and propodeum by at least one tarsomere width; vs. these sulci not as broad, with the meso- and metanota and the metanotum and propodeum separated by less than one tarsomeral width († G. spiralis, † G. subcuspis) [Note 1]; and (8) propodeal spiracle covered anteriorly by an anterior flange which is directed posterolaterally; vs. spiracle not flanged († G. spiralis, † G. subcuspis; state uncertain for † G. robusta) [Note 3]. Notes on the diagnosis Note 1: † Gerontoformica gracilis is quite similar overall to † G. spiralis and † G. subcuspis. The primary structural distinction is the width of the dorsal transverse sulci that divide the mesonotal and metanotal regions, and the metanotal region and the propodeum, as well as development of an anterior flange around the propodeal spiracles. The character used by Barden & Grimaldi (2014) to distinguish † G. spiralis and † G. gracilis was the distance between the pro- and mesocoxae (their couplet 9); this is certainly a matter of preservation, as the prothorax of the holotype of † G. gracilis is elevated relative to the mesothorax, and the coxae are promoted anteriorly. Such pronotal elevation is possible in ant species with a mobile promesonotal articulation. The defining feature of † G. subcuspis provided by Barden & Grimaldi (2014) is the form of the subpetiolar process, which has an almost vertically oriented anterior margin in profile view. The process of † G. gracilis is more evenly rounded from posterior to anterior, while that of † G. spiralis is not visible. These distinctions should be re-evaluated in future study. See also Note 3 on the synopsis of † Gerontoformica classification above. Note 2: We do not know the distribution of the process of the proximal labial palpomere across † Gerontoformica, with the exception of its absence in † G. sternorhabda. Despite this, we note the development of these processes as they are unique to our knowledge of both extant and extinct species. Because of the difficulty of evaluating proximal palpomeres for ants in general, and especially for fossil ants, we strongly recommend the application of µ-CT methods to determine the phylogenetic extent of this obvious apomorphy. Note that it is possible that the holotype lacks this condition, as we discovered this character after our chance to directly examine the type specimen. Note 3: The anterior flange or hood of the propodeal spiracle is present in most † Gerontoformica examined, with the exception of † G. spiralis and † G. subcuspis. The flange could not be evaluated for † G. magna or † G. robusta. Measurements and indices Adult, specimen C-32: HWed = 0.71; HWev = 0.78; EWl = 0.17; HD = 0.62; ML = 1.49; PnLi = 0.59; PnWa = 0.18; MnL = 0.35; AIIILm = 0.52; AIIILl = 0.65; HPI = 0.99; HSI = 0.47; AIIILI = 0.80. Pupa, specimen C-31: HWed = 0.72; HWev = 0.71; EWl = 0.18; HD = 0.62; ML = 1.62; PnLi = 0.60; PnWa = 0.25; MnL = 0.34; AIIILm = 0.62; AIIILl = 0.73; HPI = 0.86; HSI = 0.44; AIIILI = 0.85. Redescription: adult Head: The head is narrow in facial view, i.e. it is lateromedially narrower than anteroposteriorly long as measured from the anterior clypeal margin to the apparent posterior head margin (Fig. 2K); in posterior and lateral view, the head is dorsoventrally broad, with the vertexal region dome-like; standing setae are most conspicuous on the clypeus. The compound eyes are situated in the posterior third of the head; they bulge laterally, breaking the silhouette of the lateral head margins in facial view; their height mesonotal carina; mspct, mesopectus; mtbt, metabasitarsus; mtcx, metacoxa; mtfm, metafemur; mtnt, metanotum; mtpfm, metaprefemur; mtplglvf, ventral flange of the metapleural gland; mtptc, metapretarsal claw; mttbsa, anterior spur of metatibia; mttbsp, posterior spur of metatibia; mttr, metatrochanter; pd, pedicel; pl, labial palp; pm, maxillary palp; pbt, probasitarsus; pnt, pronotum; ppd, propodeum; ppdsf, anterior flange of propodeal spiracle; pptc, propretarsal claw; pt, petiole; ptn, petiolar node; pts, petiolar sternum; ptsp, subpetiolar process; sc, scape; tgl, laterotergite; to, antennal torulus. above the surrounding surfaces of the cranium is comparatively high; their ommatidia count appears similar to that of † G. sternorhabda; they are apparently glabrous, i.e. lacking interstitial setation. The three ocelli are completely developed. The frontal carinae diverge posterolaterally toward but not reaching the compound eyes; they are circular in form, i.e. curving evenly from their anterior termini to their posterior termini which are directed anterolaterally, rather than posterolaterally; their anterior termini are distant from the epistomal line; the minimum distance between the frontal carinae is about 0.26× maximum head width as measured in full face view. The antennal scrobes are anteroposteriorly short and encircled by the frontal carina. The antennal toruli are distant from the posterior clypeal margin; they are in the form of a low, more-or-less even ring. The antennae are 12-merous. The scapes are somewhat flattened and curved; they are about four times as long as their maximum width; their length is somewhat more than half the width of the head, and somewhat less than half the length of the head; they lack standing setae along their shafts. The pedicels are slightly longer than twice their width; they are about one-third the length of the scapes, and about half the length of the third antennomere; they apparently lack standing setae. The flagellae are longer than the mesosoma and are simple, i.e. not thickening distally; they bear a range of standing and appressed setae; flagellomeres I are the longest, being somewhat more than four times as long as wide and about three fourths the length of the scape; flagellomeres II–IX are all longer than the pedicel; flagellomeres X are longer than each pair of flagellomeres from II to IX. The clypeus is about three times as wide lateromedially as long anteroposteriorly, with the length measured from the midpoints of the anterior and posterior clypeal margins and the width measured between the lateralmost points of the clypeus. The lateroclypeal areas are formed as lateral lobes. The medioclypeal area is anteriorly convex; its length at the midline of the head is about 0.21× head length also at midline, as measured in full-face view; it bears an array of standing (rather than appressed) setae on its disc and is margined anteriorly by chaetae. The mandibles are simple and apically bidentate. The maxillary palps are 6-merous (Figs 2B, H, 7A–D) [Note 1]; they are elongate, being almost as long as the head and longer than the scapes and mandibles; the proximal palpomeres are short compared to the others; the second palpomeres are dorsoventrally flattened and lobate apicomedially; the third through sixth palpomeres are long, thin, and cylindrical. The labial palps are 4-merous (Figs 2B, H, 7A–D) [Note 1]; they are short, being less than half the length of the maxillary palps, and with their individual lengths shorter than each of the maxillary palpomeres with the exception of the proximal maxillary ones; the proximal labial palpomere is narrow proximally and bears a distinct lobate process medially at about its apical third, with this process being about as long as wide; the second and third palpomeres are thickened medially; the fourth palpomeres are relatively more cylindrical. Mesosoma: The pronotum bears an anteromedian neck process, and lateral and posteromedian flanges; these flanges are not flared; the muscular node or ‘disc’ of the pronotum is hemispherical as observed in lateral view and almost elliptical in dorsal view, being distinctly longer than tall in lateral (profile) view, thus appearing narrow; setation was not observed on the pronotum [Note 2]. The pronotal lobes are well developed (Fig. 6E). The mesonotum is not divided into an anterior mesoscutal area and a posterior mesoscutellar area, i.e. the transverse mesonotal carina is not developed; the notum is convex in lateral view and is almost flattened along most of its length. The mesopectus is not clearly divided into dorsal and ventral areas; its dorsoventral height is almost 1.5× that of its anteroposterior length. The transverse mesonotal sulcus is anteroposteriorly long/broad. The metanotum is developed as a distinct and almost evenly convex bulge. The transverse metanotopropodeal sulcus is anteroposteriorly broad and continues ventrally toward the base of the mesocoxa, completely separating the lateral metapectal area from the lateral mesopectal area. The metapleural gland orifice is small and not remarkably hairy (Fig. 6F); it does not have a distinct bulla; it is margined dorsally and ventrally by flanges, including the ventral metapleural gland flange. The propodeum is rounded, with the dorsal and posterior margins curving broadly into one another in lateral view; its posterior surface is convex; it does not, apparently, bear standing setae. The propodeal spiracles are situated distant from the metanotum and below the dorsal propodeal margin as seen in lateral view, but they are located in the anterodorsal fourth of the sclerite. The propodeal lobes are apparently not developed. Legs: The legs are developed as expected for the Formicidae, with some notable characters and states. With the exception of the tibial apex and tarsi, they appear almost entirely glabrous. The state of the apical protibial foramina is uncertain. The mesoprefemora and metaprefemora are welldeveloped and are broader ventrally than dorsally. The protibia bears and anterior brush of dense suberect setae in their apical third, near the calcar. Each of the mesotibia and metatibia bear a pair of apicoventral spurs; the anterior tibial spurs are barbirulate; the posterior tibial spurs are simple. The calcar is apparently bifid apically, with one point being the apex of the elongate velum and the other point being a small array of hairs; two stout setae are developed posterior to the calcar. The plantar lobes of the tarsi are well developed, the ventral setation is sparse, and the apical row of chaetae is thinner. The fourth tarsomeres are only weakly notched distally, thus appearing cylindrical in dorsal view. The pretarsal claw teeth are well developed and located just past the midlength of their respective claws. The aroliae are well developed and comparatively large (Fig. 8D); they are nearly as long as the pretarsal claws. Metasoma: The petiole is nodiform and lacks tergosternal fusion between its postsclerites. The petiolar tergum is anteroposteriorly longer than dorsoventrally tall; it is asymmetrically convex, being somewhat longer anteriorly than posteriorly. The developmental state of the laterotergites is uncertain. The petiolar sternum is anteriorly flat; its main portion is at least twice as long as tall; it is narrowly convex in cross-section at its midpoint; it is weakly produced anteroventrally as a low, lobate subpetiolar process, which is shorter dorsoventrally than wide anteroposteriorly; posteriorly, the sternum is not distinctly concave or notched. The helcium is narrow relative to the third abdominal postsclerites; the helcial tergite conceals the helcial sternite in lateral view. The abdominal posttergite III is not constricted posteriorly and is not fused with the third poststernite; it is not necked or shouldered anteriorly, as the tergum evenly curves from its anterior base to posterior margin in lateral view. The abdominal tergosternal margin III is weakly curved, without distinct ‘shouldering’ as observed in various Formicinae and Dolichoderinae. The abdominal poststernite III is not constricted posteriorly, but is weakly angled lateromedially, and bears the prora anteroventrally. The prora is lip-like in lateral view, being anteroposteriorly short and moreor-less transverse in ventral view. The abdominal segments IV, V, and VI are not divided into pre- and postsclerites by a cinctus; they are homonomous in form, i.e. highly similar in shape, size and other qualities of appearance. The seventh abdominal tergum is somewhat dome-like. The seventh abdominal sternum is lateromedially cupped and narrowed distally. The sting is long and narrow. The third valvulae are digitate in form and highly exserted as preserved. Preservation: The specimen CASENT0741232 is exceptionally well preserved internally, particularly the head and mesosoma (Fig. 9). The metasoma is, however, less well preserved, and has some apparent fungal growth or decay bubbling emanating from between the third and fourth abdominal segments. There are several fractures around the specimen, including on the dorsal surface of the head, the ventral right side of the head, across the left side of the mesosoma, and across the petiole and the left side of the third and fourth abdominal segments. The right side of the face, opposite from where the scape is held, is indented, thus appearing to have a longitudinal scrobe (Fig. 10); this is definitely an artefact, as it is not symmetrically present on the left side of the face where the scape is preserved in a position that is distant from the head capsule. Similarly, the mesosoma is indented where the legs are in close proximity (Fig. 10). The left metatrochanter appears distorted in shape. Notes on description Note 1: The original description recognized 4-merous maxillary palps and did not state a labial palpomere count. A 6, 4 palp formula (sensu Bolton, 1994, 2003) was determined here based on direct examination of the holotype, and from our microphotographs and digital renders. As noted in the diagnosis, the process of the proximal labial palpomere is unique among † Gerontoformica, given our current knowledge. Note 2: We are not certain whether † G. gracilis is largely glabrous or whether it has appressed pubescence in various places. At the least we expect pubescence on inter-sclerite contact surfaces. Description: pupa The pupa (Figs 2A, D, G, J, 7, 10) is encased in a cocoon with a black meconium at the caudal end (Fig. 1A). Most structures of species-level identificatory value are incompletely developed, including the frontal carinae, perioral sclerites, mesosomal dorsum and anterior metasoma. The preservation of the specimen is fine externally but is poor internally, with the body cavity filled with a single, solid mass, despite apparent distinctions as seen from light microscopy. In overall appearance, the specimen is relatively bubbly or puffy looking, and has the propodeum crushed and incompletely differentiated from the petiole. Those metasomal segments which are posterior to the petiole are bulging, thus they appear slightly constricted as with a cinctus, but cincti are apparently absent. The metanotal and propodeal spiracles are visible externally, but those of the metasoma are not. Features of note include the following: The head is longer anteroposteriorly than broad lateromedially; the mandibles are contacting one another apically; the maxillolabial complex is exserted, with the palps and glossa clearly visible in ventral view; the maxillary palps are 6-merous; the labial palps are 4-merous; the antennae are directed caudally and reach the posterior margin of the third abdominal segment; the pronotum is longer than tall in lateral view; and the mesonotum lacks a transverse ridge, thus is not divided into anterior mesoscutal and posterior mesoscutellar regions. The pupa differs from the synincluded adults of both species as follows: The antennae appear wideset; the clypeus is anteroposteriorly longer, apparently with an additional band of cuticle along the anterior m a r g i n; t h e m a n d i b l e s a r e o b l i q u e l y o r i e n t e d relative to head length, converging anterad (vs. perpendicular to the long axis of the head at closure); and the labrum is apically notched and with distinct paramedian lobes. Remarks on the pupa Because little comparative work on pupal morphology has been done in Formicidae, it is basically unknown what the polarity of the ‘general features of note’ is within the family. Further, because there are no developmental series of anatomy available for pupal transformation, we cannot be certain of the stage that this specimen was preserved. The conditions listed above which differ between the pupa and adults are especially interesting, as they suggest that there are intermediate stages in the structural rearrangement of the body. For example, the wideset toruli and exceptionally long clypeus are hard to explain without an atlas of developmental transformation. We note the apically bilobate labrum, in particular, as the labra of the adults are apically rounded. At present we are unable to explain the labral difference, as there are no wo, Published as part of Boudinot, Brendon E., Richter, Adrian, Katzke, Julian, Chaul, Júlio C. M., Keller, Roberto A., Economo, Evan P., Beutel, Rolf Georg & Yamamoto, Shûhei, 2022, Evidence for the evolution of eusociality in stem ants and a systematic revision of † Gerontoformica (Hymenoptera: Formicidae), pp. 1355-1389 in Zoological Journal of the Linnean Society 195 on pages 1369-1377, DOI: 10.1093/zoolinnean/zlab097, http://zenodo.org/record/6994456, {"references":["Barden P, Grimaldi D. 2014. A diverse ant fauna from the mid-Cretaceous of Myanmar (Hymenoptera: Formicidae). PLoS One 9: e 93627.","Barden P, Grimaldi D. 2016. Adaptive radiation in socially advanced stem-group ants from the Cretaceous. Current Biology 26: 1 - 7.","Bolton B. 1994. Identification guide to the ant genera of the world. Cambridge: Harvard University Press, 222.","Bolton B. 2003. Synopsis and classification of Formicidae. Memoirs of the American Entomological Institute 71: 1 - 370.","Boudinot BE, Moosdorf OTD, Beutel RG, Richter A. 2021 b. Anatomy and evolution of the head of Dorylus helvolus (Formicidae: Dorylinae): patterns of sex- and castelimited traits in the sausagefly and the driver ant. Journal of Morphology 282: 1616 - 1658. Doi: 10.1002 / jmor. 21410."]}

† GERONTOFORMICA STERNORHABDA SP. NOV. (FIGS 1C, D, 2C, F, I, L, 4– 6, 7E, F, 11 (1-2), 12(2-1, 2-3, 2-4), 13(3’-1); SUPPORTING INFORMATION, FIGS S1, S 2) Z o o b a n k r e g i s t r a t i o n: u r n: l s i d: z o o b a n k. org:act: E77C1A20-1398-465F-9607-AD11E97ADF1C Type material: Holotype. Wingless female (w). Myanmar, Kachin State: Hukawng Valley [CASENT 0741233 in an amber piece labelled AMNH SY-23 and deposited at AMNH]. Paratypes. Wingless females (w). Synincluded with holotype [CASENT0741234]; same locality as synincluded holo- and paratypes [UFV-LABECOL-009656, deposited in CELC]. Diagnosis (wingless female) Conforming to the diagnosis of the Formicidae (see: Boudinot et al. 2020a, b), including the following key conditions: (1) postgenal bridge elongated, thus head ‘prognathous’ (Fig. 2C); (2) cranial condyles of the anterior/ dorsal mandibular articulation enlarged (Fig. 5B); (3) toruli oriented dorsolaterally rather than simply dorsally (Fig. 5A); (4) procoxae elongate, about twice as long as wide (Fig. 2C); (5) prodisticoxal foramen ‘closed’ and protrochanter narrowly necked (Fig. 4F) [Note 1]; (6) meso- and metathoracicocoxal articulations ‘closed’, i.e. directed ventrally rather than laterally or ventrolaterally (Fig. 4F); (7) abdominal segment II completely petiolated (Fig. 4F); (8) subpetiolar process present (Fig. 4F); (9) prora present (Fig. 4F). I. Among Formicidae, identifiable as †Sphecomyrminae: (1) Mandibles simple and bidentate (Fig. 5A), without: (a) being strongly bowed and multidentate as in the † Camelomecia group, (b) the scythe-like blade as in the haidomyrmecine † Haidomyrmex clade, (c) projecting anteriorly with numerous teeth as in the haidomyrmecine † Aquilomyrmex clade, or (d) the strong torsion of † Zigrasimeciini (the state in † Boltonimecia is uncertain); (2) the antennal scrobes on the face do not extend all the way to the compound eye (Fig. 4B) (such scrobes observed in †Zigrasimeciinae, including † Boltonimecia) [Note 2]; (3) anterolateral corners of head not produced as robust triangles (Fig. 5A) (such corners observed in † Dilobops of the †Haidomyrmecinae); and (4) (a) scapes shorter than the width of the head (Fig. 5A) and (b) clypeus not extending posteriorly between the toruli (Fig. 2L) (such extension observed in † Brownimeciinae); II. Within †Sphecomyrminae, identifiable as † Gerontoformica : (5) anteromedian clypeal margin not produced as distinct medial lobe (Fig. 2L); vs. such a lobe present († Sphecomyrma, † G. tendir) [Note 3]; and (6) (a) mandibles short and fitting against clypeus when closed (Fig. 5A) and (b) metanotum developed (Fig. 2C); vs. mandibles elongate and metanotum not developed († Myanmyrma); III. Within † Gerontoformica, with the following unique condition : (7) the anteroventral (‘subpetiolar’) process of the petiolar sternum long, orthogonal to the longitudinal axis of the petiole, and more-or-less rod-like, i.e. with anterior and posterior margins parallel to subparallel (Fig. 2C); vs. short and triangular († G. gracilis, † G. occidentalis, † G. pilosa, † G. robusta, † G. rugosa, † G. spiralis, † G. subcuspis, † G. tendir) [Note 4]; IV. Within † Gerontoformica, identifiable as a member of the pilosa species group : (8) mesoscutum with distinct, raised transverse carina separating mesoscutal and mesoscutellar regions (Fig. 6B); vs. such a carina entirely absent or only poorly developed laterally, thus incomplete medially and not forming a distinct angle between the mesoscutal and mesoscutellar regions in profile view (cretacica, gracilis groups); (9) tergum of petiolar node anteroposteriorly longer than dorsoventrally tall (Fig. 2C); vs. petiolar node tergum taller than long (cretacica group); and (10) abdominal segment IV (metasomal III) with the cinctus distinct and impressed, i.e. divided into pre- and post-sclerites by a transverse sulcus (Fig. 5E); V. Within the pilosa species group, distinguished from all species by the following : (11) cinctus developed, but transverse sulci weakly impressed, thus pre- and postsclerites of abdominal segment IV not meeting at strongly oblique angle (Fig. 5E); vs. transverse sulci deeply impressed, thus pre- and postsclerites meeting at distinct oblique angle († G. contega, † G. magna, † G. pilosa); (12) head in full-face view broader lateromedially than long anteroposteriorly, excluding eyes (Fig. 2L); vs. head longer than broad († G. contega, † G. magna, † G. pilosa) [Note 5]; (13) pretarsal claws edentate (Fig. 5G, I, J); vs. each claw with a single tooth of variable location († G. contega, † G. magna, † G. pilosa) [Note 6]; and (14) body small, mesosoma length 1.5 († G. contega, † G. pilosa) and> 2.5 († G. magna); VI. Further distinguished by species in the pilosa group by the following : (15) anteromedian clypeal margin distinctly evenly curved to an ‘incision’ at the point of contact with the rounded lateroclypeal lobes (Fig. 2L); vs. anteromedian clypeal margin weakly convex, without distinct incision between the medioclypeus and lateroclypeal lobes, which are themselves anterolaterally angled († G. contega); (16) in profile, pronotum evenly curved, mesoscutum more-or-less aligned with mesoscutellum and metanotum, and propodeal dorsal and posterior surfaces curving into one another obliquely (Fig. 2C); vs. pronotum, mesonotum, and propodeum each shouldered in appearance, i.e. pronotum with strong anterior dorsolateral bulge, mesonotum with mesoscutal and mesoscutellar regions meeting at nearly a right angle, and propodeal dorsal and posterior surfaces also meeting at nearly a right angle († G. contega); (17) standing setation on body short, relatively sparse (with the exception of the propodeum and petiolar tergum) (Fig.5); vs. longer setation present († G. magna, † G. pilosa); and (18) anteroventral process of abdominal sternum III (metasomal II) robust but short anteroposteriorly (Fig. 2C); vs. prora anteroposteriorly long, sharkfin-like in form († G. pilosa) [Note 7]. Notes on the diagnosis Note 1: The right procoxa and protrochanter of the holotype are slightly disarticulated. However, the focal details to evaluate are the circular shape of the prodisticoxal foramen and the crank-like (curved) and thin proximal neck of the protrochanter, which can be evaluated from Figure 4F. Note 2: The only species of †Sphecomyrminae to have an elongate antennal scrobe is † G. contega. Although the scrobe of this species was observed via direct examination of the holotype, it remains a possibility that this is an artefact because other specimens of † Gerontoformica may have asymmetrical, sulcuslike distortions of the head corresponding to the position of the scrobe (see the ‘preservation’ section of the † G. gracilis description; Fig. 10). Conducting a µ-CT scan of the type specimen and/ or the accrual of additional specimens are necessary. N o t e 3: T h e a n t e r o m e d i a n c l y p e a l p r o c e s s o f † Sphecomyrma is lateromedially thin and at least as long as broad; it is distinctly bordered laterally by the medioclypeus, i.e. the clypeal ‘disc’ between the lateral clypeal lobes. In contrast, the entire medioclypeus of † G. tendir is apparently produced. See Note 5 of the ‘Notes on classification’ section above. Note 4: The form of the subpetiolar process is unknown for four species within the genus: † G. contega, † G. cretacica, † G. magna and † G. orientalis. Note 5: Although it is difficult to evaluate the head shape of † G. contega from the holotype due to cut of the amber matrix, the head does appear to be longer than broad. Reevaluation of this condition through µ-CT is recommended. Note 6: Teeth on the pretarsal claws have been recorded for most species of † Gerontoformica († G. contega, † G. gracilis, † G. magna, † G. occidentalis, † G. pilosa, † G. robusta, † G. spiralis and † G. subcuspis). The condition of having reduced teeth on the pretarsal claws, however, is shared with † G. cretacica and † G. tendir. Notably, Perrichot et al. (2008) recorded minute teeth on the pretarsal claws of † G. occidentalis, which suggests the need to re-evaluate the holotype of † G. cretacica. Note 7: The prora is clearly an important structural feature for distinguishing stem ants but is not clearly visible in most specimens, including † G. contega and † G. magna in the pilosa group, † G. cretacica and † G. orientalis in the cretacica / orientalis group, plus † G. rugosa and † G. tendir. It is extremely small and nearly absent in † G. gracilis and † G. occidentalis, and it is developed, but short and comparatively inconspicuous in † G. robusta, † G. spiralis, and † G. subcuspis. MEASUREMENTS AND INDICES Holotype, specimen C-33: HWed = 0.87; HWev = 0.95; EWl = 0.14; HD = 0.55; ML = 1.25; PnLi = 0.50; PnWa = 0.30; MnL = 0.27; AIIILm = 0.48; AIIILl = 0.53; HPI = 0.64; HSI = 0.69; AIIILI = 0.91. Paratype, specimen C-34: HWed = 0.94; HWev = 1.06; EWl = 0.20; HD = 0.66; ML = 1.39; PnLi = 0.45; PnWa = 0.27; MnL = 0.28; AIIILm = 0.50; AIIILl = 0.64; HPI = 0.70; HSI = 0.67; AIIILI = 0.78. UFV-LABECOL-009656: HWed = 0.79; HWev = 0.85; EWl = 0.11; HD = 0.46; ML = 1.05; PnLi = (~0.35–0.38); PnWa = –; MnL = 0.29; AIIILm = 0.4; AIIILl = 0.48; HPI = 0.58; HSI = 0.75; AIIILI = 0.83. Description Head: The head is broad in facial view, i. e. lateromedially wider than anteroposteriorly long from the anterior clypeal margin to the apparent posterior head margin (Fig. 2L); in posterior or lateral view, the head is dorsoventrally narrow, with the vertexal region not being particularly domed; standing setae are present on the vertexal and facial regions [Note 1]; there are short, decumbent setae distributed sparsely on the head capsule, a few longer, and suberect setae medially on the vertex near the ocelli. The compound eyes are situated in the posterior third of the head (Fig. 2L); they bulge laterally, breaking the silhouette of the lateral head margins in facial view; their height above the surrounding surfaces of the cranium is comparatively low; they comprise over 100 ommatidia, but not more than 200 (Fig. 4C) [Note 2]; they are apparently glabrous, i.e. lacking interstitial setation. The three ocelli are completely developed (Fig. 2L). The frontal carinae diverge posterolaterally, toward but not reaching the compound eyes (Figs 2L, 5A); they are sinuate in form, i.e. from their anterior margins they are directed relatively medially then bulge laterally before curving laterad; their anterior termini are close to the epistomal line, but not extending on to the clypeus; the minimum distance between the frontal carinae is about 0.29× maximum head width as measured in full face view. The antennal scrobes, or depressed contact surfaces of the face laterad the frontal carinae, are parallel in orientation relative to the frontal carinae and are apparently longer than wide (Figs 2L, 4B). The antennal toruli abut but do not indent the posterior clypeal margin [Note 3]; they are in the form of a low, more-or-less even ring. The antennae are 12-merous (Fig. 4B). The scapes are somewhat flattened and curved (Fig. 2C, F, I, L); they are about three to four times as long as wide (Fig. 2I); their length is about half the width of the head, and less than half the length of the head; they bear a vestiture of short subdecumbent to suberect setae. The pedicels are about twice as long as wide (Fig. 4F); they are about one-third the length of the scapes, and somewhat more than half the length of the third antennomere; their setation is similar to that of the scapes. The flagellae are longer than the mesosoma and are simple, i.e. not thickening distally (Fig. 4F); they bear a range of standing and appressed setae; flagellomeres I are the longest, being about four times as long as wide and more than half the length of the scapes; flagellomeres II–IX are about as long as the pedicel; flagellomeres X are longer than flagellomeres II–IX. The clypeus is about four times as wide (lateromedially) as long (anteroposteriorly), with the length measured from the midpoints of the anterior and posterior clypeal margins and the width measured between the lateralmost points of the clypeus (Fig. 2L). The lateroclypeal areas are formed as lateral lobes (Figs 2L, 5A). The medioclypeal area is anteriorly convex (Fig. 2L); its length at the midline of the head is about 0.21× head length also at midline, as measured in full-face view; it bears five or six long and flexuous setae that are situated near the anterior clypeal margin and anteriorly directed, consisting of one medial seta surrounded by a pair of setae and potentially a second, even more lateral pair [Note 4]; the anterior medioclypeal margin bears a row of chaetae. The mandibles are simple and apically bidentate (Fig. 5A). The maxillary palps are 5-merous (Figs 2I, 4A, 7E, F) [Notes 5, 6]; they are conspicuously short, with their total lengths shorter than the lengths of either the mandibles or scapes; with the exception of the apical palpomere, they are thick and bulging at about their midlengths; they are adorned with erect and appressed pilosity. The labial palps are 4-merous (Figs 2I, 4A, 7E, F) [Note 6]; they are short, being just over half the length of the maxillary palps; with the exception of the apical palpomere, they are more-orless conical and thickening toward their apices; they are adorned with erect and appressed pilosity; the proximal palpomere lacks a distinct process. Mesosoma: The pronotum bears an anteromedian neck process, and lateral and posteromedian flanges (Figs 2C, F, 5C, 6A); these flanges are flared in the paratype specimen C-34 (Fig. 4C) but not in the holotype or other paratype; the muscular node or ‘disc’ of the pronotum is almost spheroidal in shape, with the lateral margins strongly convex in dorsal view and the dorsal margin strongly convex in lateral view, and with an anteroposterior length approximating its dorsoventral height; pronotal setation is sparse, being represented by a few subdecumbent setae (Fig. 6A). The pronotal lobes are well developed (Fig. 6A) [Note 7]. The mesonotum is divided into an anterior mesoscutal area and a posterior mesoscutellar area by the transverse mesonotal carina (Figs 2C, F, 4C, F, 5C, 6B). The mesoscutal area is approximately in the form of a low saddle (i.e. is a low hyperbolic paraboloid), with a concave dorsal margin in lateral view, and with the anterior rim more upcurved than the posterior rim (Fig. 2C). The mesoscutellar area is convex, but sunken relative to the mesoscutal area (Fig. 2C). The mesopectus is divided into dorsal and ventral areas by a longitudinal sulcus (Fig. 2C); its dorsoventral height is about equal to its anteroposterior length. The dorsal mesopectal area is approximately rectangular in shape, being somewhat more than twice as long anteroposteriorly as tall dorsoventrally (Fig. 2C); its dorsoventral height is one-third the dorsoventral height of the lower mesopectal area as measured from the transverse sulcus directly ventrad to an imaginary line drawn parallel to the ventrolateral margins of the meso- and metapecta. The ventral mesopectal area is approximately triangular in shape, being broad along is dorsal margin and narrowing posteroventrally to its ventrolateral margin (Fig. 2C); its anteroposterior length along its dorsal margin is approximately equal to its dorsoventral height. The transverse mesometanotal sulcus is anteroposteriorly short/ thin (Fig. 2C). The metanotum is developed as a distinct bulge between the metanotal spiracles (Fig. 2C). The transverse metanotopropodeal sulcus is anteroposteriorly long/ broad and continues ventrally toward the base of the mesocoxa, completely separating the lateral metapectal area from the lateral mesopectal area (Fig. 2C). The metapleural gland orifice is large, hairy, and in the form of a broad subelliptical pit (Fig. 6B); it is margined dorsally by a bulge, the metapleural gland bulla, and ventrally by the ventral metapleural gland flange, which itself is a continuation of the sinuate ventrolateral carina of the metapectus and is spiniform and dorsally inclined. The propodeum is boxy (Figs 2C, 6B) [Note 8]; it bears standing setae and is the hairiest mesosomal region. The propodeal spiracles are situated distant from the metanotum, but near the dorsal propodeal margin in lateral view (Figs 2C, 5C); they are posteriorly to posterolaterally directed and protected anteriorly by the anterior flange of the propodeal spiracles. The propodeal lobes are apparently not developed. Legs: The legs are developed as expected for the Formicidae, with some notable characters and states. They are densely hairy, appearing shaggy, with the setae suberect (Fig. 4A, D, G–J). Apparently, the apical protibial foramina are open, i.e. without a bridge of sclerite dividing the calcar from the probasitarsus (Fig. 5G) [Note 9]. The mesoprefemora and metaprefemora are well developed and are broader ventrally than dorsally (Fig. 5D). The protibia bears and anterior brush of dense suberect setae in their apical third, near the calcar (Fig. 4A). Each of the mesotibia and metatibia bear a pair of apicoventral spurs (Figs 4I, H, 5H); the anterior tibial spurs are pectinate; the posterior tibial spurs are barbirulate to simple. The calcar is apparently bifid apically, with one point being the apex of the elongate velum and the other point a small array of hairs (Fig. 5G); two stout setae are developed (inserted) posterior to the calcar. The plantar lobes of the tarsi are not developed (absent), but the tarsomeres have a brush of dense ventral setae, and are apically margined by thick, coarse chaetae (Figs 4G–J, 5G–J) [Note 10]. The fourth tarsomeres of each leg are deeply notched distally, thus appearing V-shaped (probasitarsi) or arrowhead-shap, Published as part of Boudinot, Brendon E., Richter, Adrian, Katzke, Julian, Chaul, Júlio C. M., Keller, Roberto A., Economo, Evan P., Beutel, Rolf Georg & Yamamoto, Shûhei, 2022, Evidence for the evolution of eusociality in stem ants and a systematic revision of † Gerontoformica (Hymenoptera: Formicidae), pp. 1355-1389 in Zoological Journal of the Linnean Society 195 on pages 1364-1369, DOI: 10.1093/zoolinnean/zlab097, http://zenodo.org/record/6994456, {"references":["Boudinot BE, Khouri Z, Richter A, Van de Kamp T, Barden P, Perrichot V, Chaul JCM. 2020 a. Chapter II. The evolution of the ants: extinct ant sister-group illuminates eusociality origin and post-K / Pg persistence. In: Boudinot BE, Systematic and evolutionary morphology: case studies on Formicidae, Mesozoic Aculeata, and hexapodan genitalia. Ph. D. Thesis, University of California, Davis, 174 - 393.","Perrichot V, Nel A, Neraudeau D, Lacau S, Guyot T. 2008. New fossil ants in French Cretaceous amber (Hymenoptera: Formicidae). Naturwissenschaften 95: 91 - 97.","Bolton B. 2003. Synopsis and classification of Formicidae. Memoirs of the American Entomological Institute 71: 1 - 370.","Beutel RG, Gorb SN. 2001. Ultrastructure of attachment specializations of hexapods (Arthropoda): evolutionary patterns inferred from a revised ordinal phylogeny. Journal of Zoological Systematics & Evolution Research 39: 177 - 207.","Boudinot BE, Beutel RG, Gorb SN, Polilov AA. 2021 a. Functional diversity of attachment and grooming structures is retained in all but the smallest insects. Journal of Zoology 313: 99 - 113.","Wohrl T, Richter A, Guo S, Reinhardt L, Nowotny M, Blickhan R. 2021. Comparative analysis of a geometric and an adhesive righting strategy against toppling in inclined hexapodal locomotion. Journal of Experimental Biology 224: jeb 242677.","Bolton B. 1994. Identification guide to the ant genera of the world. Cambridge: Harvard University Press, 222."]}

The hypogaeic ant subfamilies Leptanillinae and Martialinae likely form the sister group to the remainder of the extant Formicidae. In order to increase the knowledge of anatomy and functional morphology of these unusual and phylogenetically crucial ants, we document and describe in detail the cranium of a leptanilline, Protanilla lini Terayama, 2009. The mandibular articulation of the species differs greatly from that of other ants studied so far, and clearly represents a derived condition. We propose a mode of movement for the specialized mandibles that involves variable rotation and sophisticated locking mechanisms. While a wide opening gape and a unique articulation are characteristics of the mandibular movement of P. lini, the observed condition differs from the trap-jaw mechanisms occurring in other groups of ants, and we cannot, at present, confirm such a functional configuration. Protanilla lini displays hardly any plesiomorphies relative to the poneroformicine ants, with the possible exception of the absence of the torular apodeme. Instead, the species is characterized by a suite of apomorphic features related to its hypogaeic and specialized predatory lifestyle. This includes the loss of eyes and optic neuropils, a pronouncedly prognathous head, and the derived mandibular articulation. The present study is an additional stepping-stone on our way to reconstructing the cephalic ground plan of ants and will contribute to our understanding of ant evolution.