Deuterodiscoelius Dalla Torre and Pachycoelius Giordani Soika are two small genera of Australian vespid wasps. While the first came from the splitting of Discoelius, the latter was created to accommodate species with a few distinct traits, some of which were not observable due to the absence of male specimens of two out of the three species designated to the genus. This already blurry delimitation of the genera was heightened when specimens of an unknown species presented a combination of characters from both genera. Hence, to properly assign this species to a genus, a phylogenetic study based on morphology was carried out before the species description. The analysis recovered both genera as unnatural, resulting in the synonymy of Pachycoelius under Deuterodiscoelius, and the new species, namely, Deuterodiscoelius delator Lopes and Carpenter, new species, was subsequently described.
INTRODUCTION
Deuterodiscoelius Dalla Torre, 1904, is a genus of zethine potter wasps comprising six extant species, all exclusive to the Australian Region. The taxonomic history of this group (fig. 1) is traced back to Discoelius Latreille, 1809, when de Saussure (1852) partitioned the genus into three divisions following both morphological and geographical distribution of its representatives. Each division was assigned a roman numeral: Division I held neotropical species, Division II, Australian, and Division III, palearctic species.
Dalla Torre (1904) kept the divisions proposed by de Saussure, agreeing with both criteria for their separation, but provided names for these divisions. Thus, the neotropical species division was established as Protodiscoelius, the Australian, Deuterodiscoelius, and the Palearctic, Tritodiscoelius. These names were ignored when Perkins (1914) created the genus Pseudozethus for a single species, P. australensis. Giordani Soika (1969) revised the genus, but only later (Giordani Soika, 1977), after learning from van der Vecht that the group had a senior name, did he synonymize Pseudozethus under Deuterodiscoelius. Two species were then properly assigned to the new name combination while the remainder were assigned to it only recently (Carpenter and Brown, 2021a).
A similar path was followed when Stange (1979) described Neodiscoelius, only to be later synonymized under Protodiscoelius by Carpenter (1986). Without these divisions, Discoelius is now comprised of only Palearctic and Oriental species and the division Tritodiscoelius is unnecessary.
Pachycoelius Giordani Soika, 1969, was created encompassing only three Australian species. Its diagnostic characters included a short propodeal valvula (fig. 2A), long maxillary palpi (fig. 2B) and males with the last flagellomere reduced (fig. 2C). These traits contrast with those of Deuterodiscoelius: an elongated propodeal valvula (fig. 2D) and shorter palpi (fig. 2E), and males with a well-developed last flagellomere (fig. 2F). However, at the time, the male of Pachycoleius was known only for the type species and the placements of the other two species were uncertain, with Giordani Soika himself raising the possibility that they could actually belong to Deuterodiscoelius.
The unclear definition of these two genera became evident even before investigating the taxonomic history, when females of an undescribed species were examined. The examined specimens presented short palpi, even shorter than those typical of Deuterodiscoelius, and a short propodeal valvula, typical of Pachycoelius. This combination of characters made it impossible to assign the new species to one of these genera. Nomenclatural acts should be done parsimoniously, avoiding unnecessary accumulated changes that might derive from a species description without the knowledge of its true generic identity. Therefore, in this case, a study on genus-level taxonomy should precede the description of the species.
Thus, the aim of this study was to closely examine species of Deuterodiscoelius and Pachycoelius in order to verify through a cladistic analysis that these genera are indeed natural phylogenetic units regarding each other, and then to describe the yet unknown species.
MATERIALS AND METHODS
The following institutions collaborated with the study by providing specimens: AMNH, American Museum of Natural History, New York; AMS, Australian Museum, Sydney; ANIC, Australian National Insect Collection, Canberra; MSNVE, Museo Civico do Storia Naturale di Venezia Giancarlo Ligabue, Venice; ZMUC, Zoological Museum, University of Copenhagen, Copenhagen.
Specimens were observed with the aid of a MZ16 Leica stereoscope. Male genitalia were extracted from relaxed specimens and were then immersed in potassium hydroxide 10% solution for up to 24 hours, followed by neutralization in 10% acetic acid, and were then rinsed and finally stored in glycerin.
For the phylogenetic analysis Deuterodiscoelius and Pachycoelius species were all used as the ingroup, examining specimens in hand, except for P. mediocris, which was examined through pictures on the ZMUC website (ZMUC, 2021) alongside its original description. Since not all its structures were observable or described, P. mediocris was eliminated from the phylogenetic analysis, as its large number of missing characters would jeopardize the study. All examined material is provided in Supplementary Material 1 (deposited in the Open Science Framework, DOI 10.17605/OSF.IO/qhs8n).
For outgroups, Discoelius zonalis (Panzer, 1801) (deposited in AMNH, ZMUC), Australozethus tasmanienesis Giordani Soika, 1969 (ANIC, ANM), Elimus australis de Saussure, 1852 (ANM), Zethus (Ischnocoelia) exiguus (Borsato, 2003) (AMNH), Macrocalymma smithianum Perkins, 1908 (ANM, ANIC), Paramischocyttarus buyssoni Gribodo. 1896 (AMNH) and Protodiscoelius merula (Haliday, 1836) (AMNH, ZMUC) were used. The last was used for rooting the tree.
The character matrix was built in Winclada (Nixon, 2015). Characters were equally weighted and unordered. Tree search was carried in the TNT software (Goloboff et al., 2008) with the branch and bound algorithm through the implicit enumeration command. Symmetric resampling (Goloboff et al., 2003) was done with 1000 replications of traditional searches for calculating support, provided as support minus contradiction group (GC) value.
Trees were plotted in Winclada and edited on Adobe Illustrator. Pictures were captured in the Leica Application Suite X with a Flexacam C1 camera attached to a Leica M205C stereo-microscope and stacked in Helicon Focus. Image plates were built in Adobe Photoshop.
The external morphology terminology follows mainly Carpenter and Garcete-Barrett (2002), Bohart and Stange (1965) and Lopes and Noll (2014) for sculpture description and Bitsch (2012) for genitalia morphology. Herein, the following terms are abbreviated: F, flagellomere, S, sternum, and T, tergum.
RESULTS AND DISCUSSION
Character List
A total of 48 characters were circumscribed, seven being from the head, 21 from the mesosoma, 12 from the metasoma, and 12 from the male genitalia. The matrix resulting from assigning character states to each terminal can be found in the Supplementary Material 2 (deposited in the Open Science Framework, DOI 10.17605/OSF.IO/qhs8n).
Head
0. Longitudinal interantennal carina: 0, absent; 1, present.
1. Transversal interantennal carina: 0, absent; 1, present.
2. Clypeus, median tooth 0, absent; 1, present.
3. Clypeus, apex: 0, not delimited; 1, delimited by a thin, usually smooth margin that contrasts with the sculpture on the remaining surface of the clypeus.
4. Maxillary palpus: 0, short, third palpomere clearly shorter than the combined three apical articles; 1, long, third palpomere subequal or longer than the combined three apical articles.
5. First labial palpomere: 0, straight; 1, recurved.
6. Occipital carina, dorsal portion: 0, continuous; 1, medially interrupted.
7. Occipital carina, ventral portion: 0, absent; 1, present as small ramification; 2, complete.
Mesosoma
8. Pronotal carina, aspect: 0, short; 1, lamellar.
9. Pronotal carina, median notch: 0, absent; 1, weak; 2, deep.
10. Humeral angles: 0, regular; 1, projected.
11. Pretegular carina: 0, absent; 1, incomplete.
12. Mesoscutum, medial carina: 0, absent; 1, present.
13. Tegula, widest at: 0, middle; 1, posterior half.
14. Mesepimeron: 0, beveled; 1, tumescent.
15. Metanotum, projections: 0, absent; 1, present.
16. Metanotum, projections, aspect 0, rounded; 1, sharp.
17. Fourth foretarsomere, apical spines: 0, absent; 1, present.
18. Hind coxa, carina: 0, complete; 1, incomplete.
19. Propodeum, lateral carina, height: 0, regular throughout; 1, raised on propodeal angle.
20. Propodeal angle, dorsal view: 0, regularly convex; 1, angled.
21. Propodeal canal, pit: 0, absent; 1, present.
22. Propodeum, apical rim: 0, absent; 1, present.
23. Propodeal orifice: 0, short; 1, elongated.
24. Propodeum, apical lamella: 0, absent; 1, present.
25. Propodeal valvula: 0, short; 1, projected posteriorly.
Metasoma
26. T1, stem: 0, short, shorter than one third of total sclerite length; 1, long, longer than one third of total sclerite length.
27. T1, longitudinal dorsal carina: 0, absent; 1, present.
28. T1, dorsal convexity: 0, regular; 1, angular.
29. T1, expansion width: 0, wide, over twice as wide as stem; 1, narrow, less than twice as wide as stem.
30. T1, expansion length: 0, short, at most as long as wide; 1, medium (at most three times longer than wide); 2, long (over three times longer than wide).
31. T1, vertical constriction: 0, gradual; 1, abrupt.
32. T1, apical horizontal constriction: 0, absent; 1, weak, apex nearly as narrow as maximum width; 2, strong, apex clearly narrower.
33. T2, apical lamella: 0, absent; 1, present.
34. T2, apical lamella length: 0, normal; 1, reduced.
35. S2, triangular smooth at base: 0, absent; 1, present.
36. T3, lamella: 0, absent; 1, present.
Male genitalia
37. Gonocoxite: 0, narrow; 1, wide.
38. Gonocoxite, apex: 0, rounded; 1, truncate.
39. Aedeagus, basal plate: 0, normal; 1, reduced.
40. Aedeagus, basal plate, length: 0, longer than wide; 1, at most as long as wide.
41. Aedeagus, basal plate, hood—when vey tip is strongly curved ventrally: 0, absent; 1, present.
42. Aedeagus, apodemes: 0, rounded; 1, flat.
43. Aedeagus, ventral lobe, size: 0, small, not passing ventralmost reach of apodeme; 1, large, reaching or passing ventralmost reach of apodeme.
44. Aedeagus, apex alignment: 0, ventrally curved; 1, dorsally curved.
45. Aedeagus, apex of head: 0, closed; 1, open.
46. Digitus, basal projection: 0, absent; 1, present.
47. Cuspis, basiventral projection: 0, absent; 1, present.
Cladistic Analysis and Genus-level Taxonomy
The analysis obtained a single tree (fig. 3). This tree supports Giordani Soika's (1969) claim of Pachycoelius being closely related to Deuterodiscoelius, because their representatives are all included in an exclusive clade with high support values (fig. 4). This group is supported by five synapomorphies; two of them homoplastic: presence of the interantennal transverse carina (fig. 5A) and presence of a median clypeal tooth (fig. 5B); and three of them unique synapomorphies: pronotal carina interrupted by a strong median notch (fig. 5C), projected humeri (fig. 5C), and mesoscutum with an anterior median carina (fig. 5C).
However, this close relationship does not uphold the current classification. Both genera were recovered as unnatural groups, Deuterodiscoelius as paraphyletic and Pachycoelius as polyphyletic. While P. carinatus is sister group to all other species in the clade, Deuterodiscoelius species are distributed in two clades. The first clade is comprised of D. confusus and D. verreauxi, and is supported by two unique synapomorphies: the strongly angulate metanotal projections (fig. 5D) and the basal plate of the aedeagus hood shaped (fig. 5E). The second clade, where P. brevicornis is more closely related to the remaining Deuterodiscoelius, is also supported by two synapomorphies: one homoplastic (the absence of an apical lamella on T3; fig. 5G) and one unique (apex of aedeagus open; fig. 5F). The new species stands as sister group to this clade.
Two of the diagnostic traits that separated Pachycoelius and Deuterodiscoelius reveal themselves as probable convergences. The elongated palpi appears as autapomorphies for P. carinatus and P. brevicornis. Meanwhile, the propodeal valvula presents ambiguous optimization for the entire clade. The reduced last flagellomere of the male should be considered a specific trait of P. brevicornis, as the male of P. carinatus was examined and presented a well-developed last flagellomere (the male of P. mediocris is still unknown).
Due to this configuration, wherein the genera do not represent natural groups, Pachycoelius is here proposed as a new junior synonym of Deuterodiscoelius.
Deuterodiscoelius Dalla Torre, 1904
Discoelius (p.p., Division II) de Saussure, 1852: 24 (division of Discoelius).
Deuterodiscoelius Dalla Torre, 1904: 18 (name for Division II of Discoelius). Giordani Soika, 1977: 109 (senior syn. of Pseudozethus). Carpenter, 1986: 19 (cat.). Vecht and Carpenter, 1990: 19 (cat.). Naumann et al., 1991: 979. (characters, distribution, species number), 981 (figure). Carpenter and Brown, 2021a: 10 (cat.); 2021b: 88 (key).
TYPE SPECIES: Discoelius verreauxii de Saussure, 1852.
Pseudozethus Perkins, 1914: 622 (genus). Bohart and Stange, 1965: 11 (key to genus). Giordani Soika, 1969: 26 (key for genus), 39 (revision, key for species); 1977: 109 (junior syn. of Deuterodiscoelius). Stange, 1979: 735 (key to genus), 738 (species number, distribution). Vecht and Carpenter, 1990: 51 (cat.).
TYPE SPECIES: Pseudozethus australensis Perkins, 1914 [= Discoelius verreauxii de Saussure, 1852].
Pachycoelius Giordani Soika, 1969: 54 (genus). Stange, 1979: 735 (key), 738 (species number, distribution). Vecht and Carpenter, 1990: 39 (cat.). Cardale, 1985: 181–182 (cat.). Naumann et al., 1991: 979 (characters, distribution, species number). Carpenter and Brown, 2021a: 15 (cat.); 2021b: 88 (key). New synonymy.
TYPE SPECIES: Pachycoelius brevicornis Giordani Soika, 1969.
Included species: D. brevicornis (Giordani Soika), D. carinatus (Meade-Waldo), D. confusus (Giordani Soika), D. delator Lopes and Carpenter, D. ephippium (de Saussure), D. insignis (de Saussure), D. mediocris (Giordani Soika), D. pseudospinosus (Giordani Soika), D. verreauxii (de Saussure).
Diagnosis: Clypeus usually with a median tooth. Maxillary palpi of variable length. Pronotum with dorsal carina lamellar and projected humeri, medially interrupted by a deep notch. Propodeal valvula variable.
Observations: Giordadni Soika (1969) affirmed there is close relationship between Deuterodiscoelius and Zethus, the main difference being the absence of the apical lamella on T2. However, a thorough examination reveals existence of the structure in most species, but in a reduced state, as the structure is only truly absent in D. insignis and D. pseudospinosus. Still, we find Macrocalymma as sister group to Deuterodiscoelius, confirming Giordani Soika's (1969) claim of unarguable affinity between the genera.
The structure of the propodeal valvula is a character usually constant among eumenine genera and commonly used as diagnostic to separate genera in identification keys (Buck et al., 2008; Carpenter and Garcete-Barrett, 2002; Giordani Soika, 1969; Tan et al., 2020). The projected trait of the structure appears in Zethus, Paramischocyttarus, Macrocalymma and Deuterodiscoelius, but the great variation among lineages of the latter renders the optimization ambiguous. The analysis allows three scenarios (fig. 6): the projected state could be plesiomorphic with independent origins to the not-projected trait in former Pachycoelius species and D. delator; the non-projected state originated once for D. carinatus and once in the closest common ancestor of D. delator and D. brevicornis with a reversion in the clade of D. insignis, D. ephippium and D. pseudospinosus; or the not-projected state could be a synapomorphy for the clade, with reversals in the two clades of Deuterodiscoelius s. str. Nevertheless, the fact that there is such variability within a genus with only a few species is an exception in the subfamily.
Species-level Taxonomy
Due to the proposed generic synonymy, species previously assigned to Pachycoelius must be formally transferred to Deuterodiscoelius through new combinations and the new species (fig. 7) is described.
Deuterodiscoelius brevicornis (Giordani Soika, 1969), new combination.
Pachycoelius brevicornis Giordani Soika, 1969: 55 (key, male). Carpenter and Brown, 2021a: 15 (cat.).
TYPE: Holotype, male – Australia, Dudiniu (MSNVE).
Deuterodiscoelius carinatus (Meade-Waldo, 1910), new combination.
Discoelius carinatus Meade-Waldo, 1910: 38 (female). Bequaert, 1928: 158 (cat.). Yasumatsu, 1934: 14 (cat.). Carpenter and Brown, 2021a: 15 (cat.).
TYPE: Holotype, female – Australia, Victoria (NHM).
Pachycoelius carinatus Giordani Soika, 1969: 55 (key), 58 (female, n. comb. for Discoelius carinatus).
Deuterodiscoelius mediocris (Giordani Soika, 1969), new combination.
Pachycoelius mediocris Giordani Soika, 1969: 55 (key), 60 (female). Carpenter and Brown, 2021a: 15 (cat.).
TYPE: Holotype, female – Australia (ZMUC).
Deuterodiscoelius delator Lopes and Carpenter, new species
Diagnosis: The maxillary palpi shorter than the labial palpi (fig. 7C) is unique in the genus, as is the combination of a short palpi with a short propodeal valvula (fig. 7D).
Description: Female.
Coloration: Black, with the following yellow markings: subdorsal transversal band on clypeus; anterior band on pronotum behind the base of dorsal carina and interrupted medially; small scrobal spot; subapical bands on T1-2 and S2. Testaceous: femora apically; tibiae and tarsi, except for mid and hind fifth tarsomere; outer two-thirds of tegula.
Structure: Clypeal apex biconcave. Maxillary palpi very short, shorter than labial palpi. Humeral angles projected. Lowermost region of mesepimeron slightly tumescent. Metanotum strongly bent, with posterior surface flat. Stem of T1 one-third the length of petiole. Expansion of T1 more than 2× as wide as stem and only 1.2× as wide as long. Parategula slender and digitiform. Apical lamellae on T2 and 3 very short, the latter not differentiated form remainder of sclerite. S2 regularly convex. S2–5 with well-developed lamellae. Forewing length: 13.4 mm.
Sculpture: Dense macropuntation without micropunctation on clypeus. Very dense micropunctation on frons, eventually coalescent. Vertex with moderate macropunctation and dense micropunctation, except for smooth ocellar region. Pronotal carina lamellar, medially interrupted. Pronotum, mesoscutum, and scutellum densely macro- and micropuntate, presenting only micropuntation surrounding discoid puncture. Notauli present on posterior third. Metanotum very densely macropuntate on dorsal surface, shimmering on posterior surface with microstriae. Propodeum also shimmering, with very dense macropunctures with cariniform interspace, less so medially, with oblique microstriae there. Lateral propodeal carina obsolete. T1 densely macro and micropunctate, with a dorsal longitudinal carina. T2–3 with sparse to moderate macropunctures intercalated by dense micropunctures. T4–6 with dense micropunctures obscuring any macropunctation. S1 smooth with very sparse macropunctures. S2 with sparse to moderate macropunctation, intercalated by smooth areas, except for a completely smooth, impunctate, basal triangle. S3–6 macropunctate with microstriae.
Pilosity: Overall pilosity long and golden. Shorter and sparser on gastral terga, where a dense whitish tomentum prevails.
Male: unknown.
Variations: Forewing length: 11.2–13.4 mm.
Etymology: The specific epithet, from Latin, means “denouncer,” or “informer,” referring to the fact of bearing traits from two distinct genera, which led to the suspicion of unnatural grouping in these taxa and execution of this study.
Examined material: Holotype: AUSTRALIA: ACT, Blundells/31.i.1970/E.F. Riek [f#, ANIC]. Paratypes: AUSTRALIA: NSW, Kiandra/27.i.1963/E.F. Riek [f#, ANIC]. CF, Genbrook [f#, ANIC].
CLOSING REMARKS
Our study illustrates how an initially simple idea can expand into greater projects, as we originally aimed to describe an unknown species of zethine Australian wasp. But to do so, we first had to establish the relationship between species of two genera, leading to the synonymy of Pachycoelius under Deuterodiscoelius. Finally, this study showcases that significant results in taxonomy and systematics can still be achieved with morphology-exclusive approaches.
ACKNOWLEDGMENTS
We thank all institutions and collections that provided material for the study. R.B.L. thanks Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for providing a scholarship for his doctoral work. R.B.L. and F.B.N. thank São Paulo Research Foundation (FAPESP) (grants #2021/00766-0, #2019/09215-6 and #2020/07895-7) and Prope/UNESP (grant # 03/2023 PROPE) for financial support. Lastly, we thank Bolívar Rafael Garcete-Barrett and an anonymous reviewer for suggestions on the manuscript.
Copyright © American Museum of Natural History 2024