Systematic Entomology (1987) 12, 413-431

Phylogenetic relationships and classification of the Vespinae (Hymenoptera: Vespidae)·

JAMES M. CARPENTER Museum of Comparative Zoology, Harvard University,'! Cambridge, Massachusetts, U.S.A.

ABSTRACT. The phylogenetic relationships of the genera, subgenera and species-groups ofthe Vespinae are analysed using cladistic techniques. The results are used as the basis for a natural classification of these wasps. The cladogram for the four genera recognized is: Vespa + .cProvespa + (Dolichovespula + Vespula)). No subgenera are recognized; all those previously described are synonymized with the appropriate genus. The . synonymies of Nyctovespa with Vespa and Rugovespula with Vespula are new.

Introduction Nomenclature. More serious is the lack of a comprehensive phylogenetic system. As a The yellowjack6ts and hornets are the· most result, paraphyletic taxa are in use, and much­ familiar of wasps. Almost ubiquitous through­ discussion of the evolutionary development of out the north temperate regions and oriental behaviour is misplaced. Without differentiation tropics, they are generally recognized - and of features into derived and primitive states, feared - by layrrmh. The history of human disputes ongroupings and evolutionary develop­ knowledge of these wasps is ancient (cf. Sprad­ ment are unresolvable (cf. Yamane, 1976; bery, 1973; Edwards, 1980). The fascination Greene, 1979; Matsuura & Yamane, 1984; Mac­ engendered by their eusocial behaviour is also Donald & Matthews, 1975, 1984). The present ancient, and so they are the subject of an ever­ work is intended to redress this situation, by burgeoning number of behavioural studies. In providing the first comprehensive cladistic treat­ addition to social organization, considerable ment of supraspecific taxa in the Vespinae. current interest is focused on aspects of venom chemistry, economic injury and biological con­ trol potential (e.g. Akre & MacDonald, 1986). Previous treatments With the attention paid to these wasps, it might be expected that their classification would be The taxonomic history of the vesplne genera has stable, and their evolutionary relationships been reviewed by Bequaert (1930, 1932), grasped at least in outline. But this is not the Guiglia (1971) and Edwards (1980). I review case. Great controversy currently exists; both here only recent studies dealing with evolution­ nomenclature and classification vary from ary relationships, independent of taxonomy. author to author, and evolutionary relationships MacDonald & Matthews (1975, 1984) ques­ are in dispute. Some current classifications do tioned the placement ofVespula squamosa in the not follow the International Code of Zoological ruta species group, or Vespula s.str. (Bequaert, 1932). Correspondence: Dr J. M. Carpenter, Museum This stance has been followed by general of Comparative Zoqlogy, Harvard University, authors, who have regarded this species and its Cambridge, Mass. 02138, U.S.A. sister-species sulphurea as of uncertain place-

413 414 lames M. Carpenter

ment (Edwards, 1980; Alae et aI., 1981). Mac­ characters (tyloides) is also incorrectly inferred, Donald & Matthews (1975: 1004) stated: and the states of the other two characters 'Comparative nest architectural, behavioral, associating Boreovespula and Dolichovespula morphological, and electrophoretic data cur­ are also true ofsylvestris (Metavespula). Finally, rently under study strongly indicate that the autapomorphic state for Boreovespula is not V.squamosa is wrongly aligned with the V.rufa true of all the included species, and this taxon is group'. No morphological or electrophoretic paraphyletic. data pertinent to this question have ever been Greene (1979) presented a discussion of published, but MacDonald (1977) and Mac­ behavioural characters which he considered to Donald & Matthews (1984) listed a number of show a 'phyletic sequence' with Dolichovespula nest architecture and behavioural characters as the most primitive yellowjacket genus and the which, they suggested, showed greater similarity vulgaris group the most recently derived group. to the V. vulgaris species group (Paravespula). This sequence, he argued, was the opposite It will be shown below that these features are of that proposed by Yamane (1976). In fact, either plesiomorphic in Vespinae, or uninforma­ Greene's discussion conflated character state tive. MacDonald (1977) stated that the male change and branching sequence. Although genitalia of squamosa are distinctive (also men­ Greene (1979: 614) stated that Dolichovespula tioned by Yamane et aI., 1980), and MacDonald 'probably has a relatively closer ancestral rela­ & Matthews (1984) cited a study by Kugler et al. tionship with Vespa', he later stated (p. 618): 'I (1976) showing seven ovarioles per ovary in have assumed the yellowjackets constitute a squamosa (versus six in most other vespines) as monophyletic group in which the genus Vespula also indicating wrong placement ofsquamosa in evolved from Dolichovespula stock, which in the rufa group. These traits, of course, are turn arose from the Vespa lineage'. This is the autapomorphies, and such unique characters are same sequence as Yamane proposed. Thus, irrelevant to the placement of the species - they whereas Greene is probably correct in the . only show that it is monophyletic. Synapomor­ polarities inferred for the behaviours discussed, phies uniting the squamosa and rufa groups are they are irrelevant to the question ofphylogene­ presented below. tic relationships among yellowjackets, as either Yamane (1976) presented a cladogram for polarity is in accord with the same scheme. vespine subgenera; however, it did not include Yamane et al. (1980) argued that the division Provespa or the parasitic subgenera (viz Vespula of Vespula s.l. into the rufa and vulgaris groups (only with austriaca), and Pseudovespula). This was inappropriate, and subdivided Vespula s.l. latter omission is unimportant, as recognition of into seven groups. Their character states were a subgenera for the parasitic species renders other mixture of primitive and derived, and they subgenera paraphyletic, as discussed below. stated (p. 34): 'These groups are, however, Yamane (1976: Table 3) listed fifteen characters rather phenetic'. They did not investigate the used in the construction of his tree, and differen­ interrelationships of these groups (nor establish tiated these into primitive and derived states. their monophyly), but did adopt Guiglia's (1972) He upheld the monophyly of Vespula + division of Vespula into the subgenera Vespula Dolichovespula, and Dolichovespula, but could and Paravespula! They considered the assign­ not determine whether Allovespula was the sis­ ment of the koreensis group to Paravespula and ter-group of Paravespula or Dolichovespula. As the squamosa group to Vespula to throw doubt developed below, the polarity of two of on this division. They stated (p. 34): 'They have Yamane's characters was incorrectly inferred, characters both peculiar to them and transitional and there are no grounds for grouping between the two subgenera'. The interrelation­ Allovespula with Dolichovespula (done on the ships of these groups are discussed below, aI).d it basis of grey, pliable nest paper). Yamane's is shown that the peculiar characters are cladogram also showed the interrelationships of autapomorphies, whereas the 'transitional' are the subgenera of Dolichovespula as: Meta­ symplesiomorphies. vespula + (Boreovespula + Dolichovespula). Archer (1982) described the subgenus The latter two taxa were grouped by three char­ Rugovespula for the koreensis group, primarily acters, and single autapomorphies were noted because-the male genitalia ofthis group 'are very for each of them. The polarity of one of these different from those of other species of Para- Phylogenetic relationships and classification ofVespinae 415 vespula'. It is shown below that the differences are vespine autapomorphies not mentioned by discussed all show the pnmitive state in Carpenter (1981). The autapomorphies are Rugovespula, and this taxon is the sister-group included in the matrix to establish the mono­ of Paravespula. phyly of each terminal taxon in the 'Cladistic Varvio-Aho et al. (1984) presented an Diagnoses' section. allozyme data set for the eight species of Euro­ pean yellowjackets, and analyses which they Character matrix (Table 1) suggested showed that Dolichovespula was not a 1 Prestigma length. Prestigma shorter than monophyletic group, and that Paravespula + pterostigma, O. Prestigma ~ pterostigma, 1. Pre­ Vespula was not monophyletic. Carpenter stigma 3x pterostigma, 2. (1987) showed that these authors incorrectly 2 lRI cell length. Length of first submarginal (lRI) analysed their data, and that their data are in fact cell ::; distance from ap<;x of cell to apex of wing, largely uninformative· on the relationships of O. Length of cell> this distance, 1. these yellowjacket species. 3 Base of lRS cell. M vein obliquely oriented with Finally, Matsuura & Yamane (1984) respect to m-euI at base of second submarginal (IRS) cell, O. M vein vertically oriented (apex of published two matrices totalling forty-two char­ discal cell truncate), 1. acters for the Vespinae (including Vespula, 4 lRS cell M: RS. Second submarginal cell with Paravespula, Rugovespula and the squamosa section ofM vein distal to m-eu2 crossvein shorter group as terminal taxa). The matrix is than section of RS vein basal to r2 crossvein, O. M reproduced here as Table 2, with correction of section equal to orlonger than RS section, 1. This some of the inferred polarities as explained is a way of characterizing the autapomorphic dis­ below. Matsuura and Yamaue could not resolve tal elongation of the second submarginal cell found in most vespines. It is not found in the the placement ofProvespa, although they recog­ Vespulajlaviceps group (Yamaneetal., 1980: Fig. nized Provespa + Vespula + Dolichovespula as 21), which is here considered an apomorphic a monophyletic group, with Vespa as the sister­ reversal. group of this. Reanalysis of their data is pre­ 5 Hamuli placement. Beginning basad of fork of RI sented below, which resolves the relationships and RS, O. Beginning at this fork, 1. as: Provespa + (Vespula + Dolichovespula). 6 Occipital carina. Running to base of mandible, O. Effaced near base of mandible, 1. The derived state is also found in the jlaviceps group of Para­ vespula, which is another apomorphy of the Materials and Methods group. 7 Occipital carina. Present dorsolateraIIy, O. Themonophyly ofthe subfamily is established in Absent dorsolateraIIy, 1. Carpenter (1981); its sister-group is the 8 Malar space. Length less than thatoflast antennal Polistinae. Characters have been polarized with segment, O. Length greater than or equal to that reference to Polistinae. In the tables which of last antennal segment, 1. Vespa has a range of variation in this character which includes the follow the state of '0' indicates the primitive derived state (e.g. inandarinia), but this is clearly state. Where the taxon shows both the primitive convergent, as most species Show the primitive and derived states, the parsimoniou·s inference state. of the groundplan is coded. Usually this is the 9 Vertex length. Ocelloccipital distance shorter primitive state, except as noted. The vespine than or equal to distance between posterior species examined are listed in the appendix. ocellus and eye, O. Ocelloccipital distance greater than distance between posterior ocellus and eye, Cladistic analyses (Hennig, 1966) were per­ 1. formed using the PHYSYS system by J. S. Farris 10 Ocelli. Ocellar diameter less than distance and M. F. Mickevich, State University of New between posterior ocellus and eye, O. Ocellar York at Stony Brook and Maryland Center for diameter greater than this distance, 1. An adapta­ Systematic Entomology, as implemented on the tion to nocturnal habits, this state is also found in VAX 11/780 computer running VMS at Vespa binghami, which is apparently also noctur­ nal (van der Vecht, 1959). However, the ocelli in Harvard. this species are not as large as those in Provespa. The tree lengths and consistency indices reported do not include autapomorphies or 11 Labial palpus. Third segment with a strong seta. O. Third segment without this seta, but with invariant characters. The invariant characters hairs, 1. Duncan (1939) stated that there was 'no are listed in the character matrix because they striking difference' between the two states, 416 lames M. Carpenter

TABLE 1. Character matrix for Vespinae. Character Vespa Provespa Dolichovespula Vespula V. sq. P. R. 1. Prestigma length 2 1 1 1 1 1 1 2. R1 cell length 0 1 0 0 0 0 0 3. Base of lRS cell 0 1 1 1 1 1 1 4. 1RS cell M: RS 1 1 1 1 1 1 1 5. Hamuli placement 0 ], 1 1 1 1 1 6. Occipital carina 0 0 1 1 1 0 0 7. Occipital carina 0 1 0 0 0 0 0 8. Malar space 0 0 1 0 0 0 0 9. Vertex length 1 0 0 0 0 0 0 10. Ocelli 0 1 0 0 0 0 0 11. Labial palpus 0 0 1 1 1 1 1 12. Tyloides 0 0 0 1 1 1 1 13. Pronotal carina 0 1 1 2 2 2 2 14. Pretegular carina 0 1 1 1 1 1 1 15. Mesepisternum 1 1 1 1 1 1 1 16. Scutallamella 0 0 1 1 1 1 1 17. Propodeal striae 0 0 0 0 0 0 1 18. Coxal carina 0 1 1 0 0 0 0 19. ~ SVI process 0 0 0 0 0 1 1 20. d'TVII 0 0 0 0 - 0 1 1 21. d'SVII 0 0 0 0 0 1 1 22. Aedeagus apex 1 0 0 0 0 0 0 23. Aedeagus apex 0 0 0 0 0 1 0 24. Aedeagus apex 0 0 0 1 1 1 1 25. Aedeagus width 0 0 1 0 0 0 0 26. Paramere process 0 1 0 1 0 0 0 27. Paramere process 0 0 0 0 0 1 0 28. Volsella 0 0 0 1 1 0 0 29. Digitus 0 0 0 1 1 0 0 30. Larval mandible 0 0 1 0 0 0 ? .~~.\ 31. Larval c1ypeus 0 0 1 0 0 0 ? 32. Spiracle 1 0 1 0 0 0 ? 33. Nest aerial 0 0 0 1 1 1 ? 34. Paper type 1 0 0 0 0 0 ? 35. Scallopping 1 0 0 0 0 1 ? 36. Queen cells 1 1 1 1 1 1 ?' 37. Suspensoria 0 0 2 1 0 0 ? 38. Worker-cell comb 0 0 0 1 0 0 ? 39. Mixed cell combs 0 0 0 1 0 0 ? 40. Colony size 0 0 0 0 1 1 ? 41. First pedicel 0 ? 0 1 1 1 ? 42. First attachment 0 0 1 1 1 1 ? 43. Swarming 0 1 0 0 0 0 ? 44. Royal court 1 1 0 0 0 0 ? 45. Prey 0 0 0 0 0 1 ? 46. Nocturnal 0 1 0 0 0 0 0

attributing the size of the seta in Vespa to the are absent in a few species in Provespa and relatively large size of the species in this genus. Dolichovespula (sylvestr.is , social parasites) . Duncan did not examine any Provespa or mem- Yamane et al. (1980) stated that they were present bers of other vespid subfamilies with the seta pre- in Vespula vulgaris, but whereas there are shiny sent. The presence of the seta in these other areas ventrally on the antennal segments, there groups shows that it has nothing to do with size of are no raised ridges. the individual, and the distinction of these two 13 Pronotal carina. Present, O. Dorsally reduced, 1. states is thus well supported. Laterally effaced, 2. The state in Dolichovespula 12 Tyloides. Present, O. Absent, 1. Tyloides are pre- is problematic. Although some species, e.g. sent in the groundplan of Polistinae, and the maculata, have the carina developed dor- Vespinae as well, contra Yamane (1976). They solaterally close to the scutal margin, as in Vespa, Phylogenetic relationships and classification ofVespinae 417

it is not as strong dorsolaterally, is weaker in throughout vespines, in groups that 'otherwise workers, and most other species have it effaced have little in common' (Kluge & Farris, 1969). well laterad ofthe scutum. Coding this genus with The parsimonious inference is therefore that it is the plesiomorphic state does not affect the result­ part ofthe groundplan ofVespinae (Farris, 1982). ing tree shape, and when the data are analysed 27 Paramere process. Parameral process not project­ using this scoring, the state in Dolichovespula is ing fingerlike, O. Parameral process strongly pro­ inferred to be a reversal on Fig. 1 (due to median jecting, fingerlike, 1 (see Bequaert, 1932: Figs. 2, state optimization; Farris, 1970). Therefore I am 3). treating the state in Dolichovespula as the first reduction state, although it is not as reduced as 28 Volsella. Volsella long, extending to apex ofpara­ the state in Provespa. meral spine and aedeagus or further, O. Volsella short, not extending as far as apex of parameral 14 Pretegular carina. Present, O. Absent, 1. spine or aedeagus, 1 (see Bequaert, 1932: Figs. 4, 15 Mesepisternum. Epicnemial carina and dorsal 5). groove present, O. Epicnemial carina and dorsal 29 Digitus. Digitus a large, pincer-shaped lobe, O. groove absent, 1. Digitus a very slender, fingerlike lobe, 1 (see 16 Scutallamella. Rim ofscutum raised into lamella Bequaert, 1932: Figs. 4, 5). beside tegula, O. Scutal lamella effaced beside 30 Larval mandible. Larval mandible tridentate, O. tegula, 1. Bindentate, 1. Yamane (1976) listed D.sylvestris 17 Propodeal striae. Propodeum not striate, O. Pro­ (Metavespula) as tridentate. He did not see this podeum striate in queens, 1. species, but followed Short (1952). Wagner (1978) examined the larva ofsylvestris, and stated 18 Coxal carina. Carina present on pos­ that it does not differ in its mandible from other terior surface of hind coxa, O. Carina absent Dolichovespula. I have examined the larva of this ventrally on posterior surface of hind coxa, 1. species and confirm this. The mandible is more or 19 ~ SVI process. Female without dorsolateral pro­ less bidentate, with serration between the teeth, cess on metasomal sternum VI (SVI), O. Female as figured for other species by Yamane (1976). with dorsolateral process on Sternum VI, 1 (see 31 Larval clypeus. Larval frontoclypeal suture Archer, 1982: Fig. 1). developed, O. Larval frontoclypeal suture effaced 20 cf TVII. Male with disc ofmetasomal tergum VII dorsally, 1. (TVII) evenly convex, O. Male with disc of TVII 32 Spiracle. Larval spiracular collar processes sim­ depressed, 1 (see Bequaert, 1932: Fig. 2). ple, O. Collar processes branched, 1. 21 cf SVII. Male with metasomal sternum VII 33 Nest aerial. Nesting site aerial, O. Nesting site in' (SVII) triangular, O. Male with SVII transverse, 1 cavity or underground, 1. This character ()cca-, (see Archer, 1982: fig. 3). sionally varies in Vespa and Dolichovespula, but 22 Aedeagus apex. Apex of aedeagus not projecting the polarity is clear. laterally, O. Apex of aedeagus projecting later­ 34 Paper type. Grey, pliable, O. Brittle brown, 1.' ally, apex transverse, 1 (see Buysson, 1905: The difference in colour and strength is attributed Plate 7). A state approaching this is found in to collection of sound wood fibres versus rotten Provespa anomala (Buysson, 1905: Plate 5, Fig. wood fibres (Spradbery, 1973). Yamane (1976) 1). The genitalia in this genus are very diverse in inferred the opposite polarity, but this is not sup­ form, but a state similar to Vespa does not seem to ported by outgroup comparison. This character be the groundplan state. apparently varies in Vespa (van derVecht, 1957), 23 Aedeagus apex. Apex ofaedeagus not subcircular Dolichovespula (Greene, 1979) and Paravespula on spoon-shaped, O. Apex of aedeagus subcircu­ (MacDonald, 1977). The species of Paravespula lar, spoon-shaped, 1 (see Bequaert, 1932: Figs. 2, which have the plesiomorphic state (germanica 3). andpensylvanica; group 3 ofYamane et al., 1980) 24 Aedeagus apex. Rods forrning aedeagus separ­ are apparently relatively basal in the group, ated apically, O. Rods forming aedeagus fused by therefore tbe apomorphic state cannot even be a membrane apically, 1. Although in general the groundplan characteristic ofthe group. This char­ Vespidae have the rods forming the aedeagus acter is not very informative. fused apically (Carpenter, 1981), the par­ 35 Scallopping. Envelope laminar, O. Envelope simonious interpretation is that they are separ­ scallopped, 1. The situation is complicated by ated apically in the vespine groundplan. variation within Vespa (e.g. Matsuura, 1984) and 2S Aedeagus width. Aedeagus as wide or wider Dolichovespula (Greene, 1979), as well as a apically as medially, O. Aedeagus tapering mixed type of construction in Provespa (Mat­ towards apex, narrower there than medially, 1 suura, 1985). This character cannot be regarded (see Bequaert, 1932: Figs. 1,6). as very informative. 26 Paramere process. Paramere with dorsal process 36 Queen cells. No special cells constructed for rear­ near spine, O. Paramere without process, margin ing queens, O. Special queen cells constructed, 1. smooth, 1. Although not found in the outgroup, a 37 Suspensoria. Comb suspensoria pillarlike, O. Rib­ process of some sort is rather generally found bonlike supporting first comb, 1. Ribbonlike 418 lames M. Carpenter

throughout nest, 2. MacDonald (1977), citing second of the remaining eighteen. These latter Spradbery (1973), suggested that pilIarlike were characterized as 'difficult to determine the suspensoria were secondary, but these are like the suspensoria built in polistines. ancestral or derived condition'. I have deter­ mined the polarities for these features using the 38 Worker-cell comb. Multiple combs of worker cells, O. One comb of worker cells, 1. Although Polistinae as an outgroup. there is some variation (e.g. Greene, 1979, men­ 1 Vertex length. OOD 2:: 2POD (exceptionally tions the derived state as occurring in OOD=POD), O. Distance between ocellus and D. maculata), it is confined as a groundplan trait to posterior of head (POD) > distance between the ruta group (Akre et al., 1981). Thus, it is ocellus and eye (OOD), 1. Matsuura & Yamane apomorphic within Vespinae, contrary to Greene originally coded these states in reverse. Thus, (1984). they considered the elongate vertex ofVespa to be 39 'Mixed cell combs' (MacDonald & Matthews, primitive, which is clearly erroneous; it is not 1984) present, O. Worker and queen cells on approached in any primitive polistine. different combs, 1. This is probably related to 2 Ocelli size. Ocellus diameter < OOD, O. Ocellus character 38, and similarly, the derived state is an diameter> OOD, 1. autapomorphy of the ruta group (Vespula s.str.). 3 Clypeal apex. Narrow, O. Broad, 1. 40 Colony size. 'Small' «2500 cells and 400 workers at peak), O. 'Large' (higher numbers), 1. This is 4 Clypeal teeth. Rounded or truncate, O. Sharp and another variable feature, the ranges in Vespa, pointed, 1. The derived state characterizes the Provespa and Dolichovespula overlapping that in social parasites only. Although autapomorphic in Paravespula (cf. Matsuura, 1983, 1984, 1985; these species, the subgenera from which these Greene et al., 1976; with Akre et al., 1981) - as species are removed if placed in their own sub­ does the range, exceptionally, in the ruta group genera are then characterized only by the absence (MacDonald et al., 1974). It is about the only of this feature (and others associated with parasi­ derived character linking the squamosa and tism, cf. Bischoff, 1931a, b), and are thus para­ vulgaris groups that I have been able to adduce, so phyletic. This character is not used in the I have coded it to reflect similarity between these following analyses. groups. It is weak evidence indeed, as shown by 5 Pronotal carina. Present, O. Dorsally effaced, 1. the results below. Matsuura & Yamane scored Dolichovespula as 41 First pedicel. Pedicel of embryo nest coated with ancestral for this, an interpretation I doubt (see glossy oral secretion, O. Uncoated, 1. See Greene above). Nevertheless, I have used Matsuura & et al. (1976). Yamane's scoring, since a blunt carina is present dorsolaterally at least primitively in the genus. 42 First attachment. Pedicel of embryo nest with simple cylindrical shape, O. Pedicel thin and 6 Coxal carina. Present, O. Absent, 1. This is the twisted, 1. Yamane & Makino (1977) note that interpretation of Carpenter (1981). Matsuura & the plesiomorphic state occurs in Vespula Yamane originally scored this in reverse, but dor­ vulgaris. sal remnants of the carina are present in Pro­ vespa. Using their coding does not affect the 43 Swarming. Colony founded by queen only, O. resulting tree shapes, although it of course lowers Swarming, 1. See Matsuura (1985). the consistency of the trees. Matsuura & Yamane 44 Royal court. No distinct royal court of workers also scored Dolichovespula as having the carina, surrounding queen, O. Royal court of workers although it is reduced to at most traces in the present, surrounding queen, 1. This varies in group (maculata). Vespa (Matsuura, 1984, 1985). 7 Prestigma length. Pterostigma longer than pre­ 45 Prey. Live insects, O. Vertebrate carrion in addi­ stigma, O. Prestigma longer than pterostigma, 1. tion, 1. This varies in the squamosa group (Mac­ 8 Prestigma length. Length of prestigma < 2x r1 Donald & Matthews, 1984) and Dolichovespula vein, O. Prestigma > 2x r1 vein (pterostigma (Greene et al., 1976). extremely short), 1. A way of characterizing the 46 Nocturnal. Diurnal, O. Nocturnal, 1. Vespa extremely elongate prestigma of Vespa. binghami is apparently also nocturnal (van der 9 Base of 1RS cell. M oblique, O. Vertical, 1. Vecht, 1959). Enlarged ocelli (character 10) is presumably an adaptation to nocturnal habits, as 10 Jugal lobe. Present, O. Absent, 1.

enlarged ocelli are a feature of virtually all other 11 Hamuli placement. Beginning distad offork ofRJ nocturnal Hymenoptera. and RS, O. Beginning at this fork, 1. 12 Ovariole number. 3 pairs, O. 2:: 6 pairs, 1. 13 Ovariole number. 6 pairs, O. 7-10 pairs, 1. Iwata Matsuura & Yamane (1984) matrix (Table 2) (1955) reported that the number varied within four species of Vespa, and even from ovary to Matsuura & Yamane presented their data as ovary in individuals. He also reported six pairs or two matrices (1984: Tables 10.la and 10.lb). fewer in three of the species (xanthoptera, crabro The first consisted of characters 1-24, and the fiavofasciata and tropica pulchra). Kugler et al. Phylogenetic relationships and classification of Vespinae 419

TABLE 2. Data matrix from Matsuura & Yamane (1984). Character Vespa Provespa Dolichovespula Vespula V: sq. P. R. 1. Vertex length 1 0 0 0 0 0 0 2. Ocelli size 0 1 0 0 0 0 0 3. Clypeal apex 1 1 1 1 1 1 1 4. Clypeal teeth 0 0 [0] [0] 0 0 0 5. Pronotal carina 0 1 0 1 1 1 1 6. Coxal carina 0 1 0 0 0 0 0 7. Prestigma length 1 1 1 1 1 1 1 8. Prestigma length 1 0 0 0 0 0 0 9. Base of lRS cell 0 1 1 1 1 1 1 10. Jugal lobe 1 1 1 1 1 1 1 11. Hamuli placement 0 1 1 1 1 1 1 '12. Ovariole number 1 ? 1 1 1 1 1 13. Ovariole number 1 ? 0 0 1 0 0 14. Larval cIypeus 0 0 1 0 0 0 0 15. Mandible shape 0 0 0 0 0 0 0 16. Mandible teeth 0 0 1 0 0 0 0 17. Inner processes 0 0 1 0 0 0 0 18. SpiracIe 1 1 1 1 1 1 1 19. Prey 0 0 0 0 0 1 ? 20. Swarming 0 1 0 0 0 0 ? 21. Cocoon cover 0 ? 0 0 0 1 ? 22. First attachment 0 0 1 1 1 1 ? 23. First pedicel 0 ? Ib la ? la ? 24. Embryo nest disc 0 ? 0 1 ? 1 ? 25. Occipital carina 0 0 1 1 1 0 0 26. Malar space 0 0 1 0 0 0 0 27. Labial palpus 0 0 1 1 1 i 1 28. Pretegular car. 0 1 1 1 1 1 1 29. TI 0 1 0 0 0 0 0 30. Aedeagus apex 0 1 0 0 0 0 0 31. Larval labrum 0 0 1 0 0 0 0 ',rV' 32. Collar processes 2 1 2 1 1 0 ? 33. Nest site 0 0 0 1 1 1 ? 34. Colony duration 0 ? 0 0 1 1 ? 35. Hunger signal 0 ? 0 1 ? 1 ? 36. First sheet 0 0 1 1 1 1 ? 37. Expansion 0 ? 1 0 0 0 ? 38. Paper type ?? 1 1 1 0 ? 39. Suspensoria 0 0 1 1 0 0 ? 40. Suspensoria 0 0 1 0 0 0 ? 41. Scallopping 0 0 1 1 1 0 ? 42. Hibernaculum 1 ? 1 ?? 0 ?

(1976) reported intraspecific and intraindividual following Yamane (1976), but bidentate seems to variation in ovariole number in Vespa crabro, describe the state better. mandarinia and tropica trimeres; and in Vespula germanica. Therefore, the homology implied by 17 Inner processes. Small processes not present on Matsuura and Yamane's scoring seems quite inner surface oflarval mandible, O. Processes pre- dubious, and I have not used this character in any sent, 1. analyses. 18 Spiracle. Larval spiracular collar processes 14 Larval cIypeus. Larval frontoclypeal suture absent, O. Collar processes present, 1. developed, O. Larval frontoclypealsuture effaced 19 Prey. Live insects, O. Vertebrate carrion in addi- dorsally, 1. tion, 1. 15 Mandible shape. Larval mandible thick and strong, O. Larval mandible thin and weak, 1. 20 Swarming. Colony founded by queen only, O. Swarming, 1. 16 Mandible teeth. Larval mandible tridentate, O. Bidentate, 1. Matsuuia & Yamane actually used 21 Cocoon cover. Cap consisting of two layers, sim- 'monodentate' to describe the derived state, pIe in structure, no sex or caste differences, O. Cap 420 lames M. Carpenter

consisting of two layers in the <;> and d', and three 33 Nest site. Open space, O. Covered space, 1. in the S?, with complex structure, 1. 34 Colony duration. Early decline, O. Decline in late 22 First attachment. Pedicel of embryo nest with fall, 1. I have maintained Matsuura & Yamane's simple cylindrical shape, O. Pedicel thin and scoring for this feature. However, it actually twisted, 1. Matsuura & Yamane did not code this varies not only within genera in a single locality or the following two characters for Provespa, as (e.g. Matsuura, 1984), but within species this genus is supposed to found nests by swarming latitudinally (cL Akre et al., 1981) and also and so does not have a strictly comparable seasonally (Roush & Akre, 1978; cL Greene, 'embryo' nest. However, Matsuura (1985) indi­ 1984, with Spradbery, 1973, and Edwards, 1980)­ cates that the pedicel ofthe nest in Provespa is the and, of course, does not apply to tropical species. same as the primitive state for an embryo nest, It is thus poorly defined, and apparently simply and so it is coded as 0 here. dependent on climate in many species. Members 23 First pedicel. Final coating on pedicel of embryo of the vulgaris group (Paravespula) may exhibit a nest dry, O. Final coating either adhesive (la) or tendency for later colony decline (Spradbery, lacking (lb). As state Ib is an autapomorphy of 1973; MacDonald et a!., 1974), but Matsuura & Dolichovespula, I have simply scored it as 0 for Yamane's scoring of the derived state also in the state la in the analyses. squamosa group is questionable. Whereas the 24 Embryo nest disc. Lacking, O. Present, 1. southern North' American squamosa has relatively late colony decline, its sister species 25 Occipital carina. Present close to mandible, O. sulphurea has early decline (Wagner, in Akre et Effaced within one third of that distance, 1. Mat­ al., 1981). suura & Yamane originally scored Provespa as 35 Hunger signal. Larvae use consistent cell wall having the apomorphic condition. Indeed, the scraping, O. Larvae use low frequency wall scrap­ carina is absent dorsally in all species of this ing, 1. As this type of signalling does not occur in genus, but is present ventrally in females. Polistinae (Yamane, 1976), the polarity could be Therefore it is clear that the carina reduction in coded in reverse. I have followed Matsuura & Provespa is not homologous with that in other Yamane's coding, as this is the parsimonious vespines, and I have treated it as having the primi­ interpretation of the character for any of the tive condition in analyses of Matsuura & resulting cladograrns. Yamane's matrix. 36 First sheet. Bonding point of first envelope sheet 26 Malar space. Length less than half the length of ofembryo nest directly to the nesting foundation, the eleventh antennal segment, O. Length"" the O. Bonded to pedicel, 1. Matsuura & Yamane did length of the eleventh antennal segment, 1. not score Provespa for this character, because its 27 Labial palpus. Third segment with 1 or 2 strong nests are built by swarms and hence are not true setae, O. Third segment with weak hairs, 1. embryo (built by the queen) nests. But the nests are homologous as initial nests, and according 28 Pretegular carina. Present, O. Absent, 1. to Matsuura (1985), anomala exhibits the 29 TI. Sharply angled between dorsal and anterior plesiomorphic condition in the bonding point. faces, O. Angle indistinct, 1. Considering R. W. Matthews (personal communication) has polistines, this could be coded in reverse, but discovered that squamosa has the apomorphic doing this simply produces an interpretation of state of this character, and so I have scored the reversal in Provespa on the cladograms. squamosa group as 1. Therefore I have used the original scoring for this 37 Expansion. Expansion of envelope of embryo character. nest with sheet after the second or third beginning 30 Aedeagus apex. Rounded, O. Bifurcate, 1. on previous sheet, O. Independent sheet begins Actually, an aedeagus with a divided tip is from the pedicel and nesting foundation, 1. R. W. plesiomorphic in Vespinae (see above) so Vespa Matthews (personal communication) has found and Delichovespula are not coded as derived that squamosa has the pleisomorphic condition of here. However, the very deeply divided tip in this feature, and so I have scored the squamosa Provespa appears autapomorphic. group as O. 31 Larval labrum. Widest basally and emargination 38 Paper type. Brittle brown, O. Grey, pliable, 1. As somewhat deep, O. Greatest width more apical, discussed above, this polarity is not supported by emargination shallow, 1. Yamane (1976), follow­ outgroup comparison. Matsurra & Yamane in ing Short (1952), listed D.sylvestris as having the addition originally had Vespa and Provespa not primitive condition. I have seen the larva of this coded for this character, and Paravespula with species and it has the derived condition. both states. I coded paper type both ways, and as the coding does not affect the results, I used Mat­ 32 Collar processes. Short, simple, O. Long and sim­ suura & Yamane's original coding. ple (1) or long and branched (2). The 'primitive' state, occurring only in Paravespula, is most par­ 39 Suspensoria. Suspensoria between first comb and simoniously inferred to be derived, but I have envelope pillarlike, O. Ribbonlike, 1. followed Matsuura & Yamane's coding here, as it 40 -Suspensoria. All other suspensoria pillarlike, O. simply appears as a reversal. Ribbonlike, 1. Phylogenetic relationships and classification ofVespinae 421

41 Scallopping. Envelope scallopped, O. Envelope Paravespula laminar, 1. This polarity is incorrect considering polistines. I have coded it both ways; it does not affect the resulting tree shapes. I have followed & Matsuura Yamane's original scoring here. Rugovespula 42 HibernacuIum. Using pre-existing cavities, O. The queens hibernate singly, in a hibernaculum which is at least partly dug by the queen herself, 1. .JE:------squamosa group

Results

I first analysed the seventeen informative Vespula morphological characters from Table 1. A branch-and-bound routine (XWAGNER) was used to perform an exact solution for parsimony. The tree of Fig. 1 resulted; it has a length of 22. Dollchovespula The consistency index (the sum ofthestates of all the characters, divided by the length ofthe tree;

Kluge & Farris, 1969) is 0.82 as calculated with Provespa the LFITcommand. Vespa is the sister-group of the remaining Vespinae; Dolichovespula and Vespula are sister-groups, and Vespula s.l. is monophyletic. The squamosa group is the sister­ Vespa group of Vespula s.s., and Rugovespula is the FIG. 2. Strict consensus tree for the six most par­ sister-group of Paravespula. simonious trees for the data ofTable 2. When the seven informative behavioural characters listed in Table 1 were included and the analysis performed again, the same tree resulted. The length is 34, with consistency of Paravespula 0.76. The behavioural data are ambiguous rela­ tive to the morphological data. This was explored further. When the seven behavioural characters were analysed alone, five trees Rugovespula resulted, all of length 10 and consistency 0.8. The strict consensus tree - the tree that has only the groups found on all five of the cladograms squamosa group (Nelson, 1979; calculated with the NELSON command) - was completely unresolved. The trees do not agree, but this is due to the fact that Rugovespula is scored as missing for all of the Vespula characters. The Adams consensus tree, which takes the intersection of groups (Adams, 1972; calculated with the ADAMS command),

Dollchovespula showed that the trees differed only in where Rugovespula was placed. Provespa grouped with Vespa (on the basis of character 44, royal court), and Dolichovespula with Vespula + (the Provespa squamosa group + Paravespula). The latter two taxa were grouped by character 40, 'large' col­ onies. When the five behavioural trees were diagnosed with the morphological data (mean­ ing that the morphological data were fit to the FIG. 1. Cladogram for the subgenera ofthe Vespinae. behavioural trees, using the DIAGNOSE com- 422 lames M. Carpenter mand), their lengths ranged from 32 to 36 with Thus although Matsuura & Yamane's matrix consistencies of0.56-0.5. Conversely, diagnos­ admits of multiple trees, both data sets support ing the tree of Fig. 1 with only the behavioural the cladogram ofFig. 1. ButFig. 1is also the best characters produced a length of 12, consistency choice as the general cladogram for these wasps 0.67. The morphological data better account for when the partly overlapping data sets are com­ the behavioural datathan the reverse. Even dis­ bined. When the characters from Matsuura & regarding the problems in scoring the Yamane's matrix not also listed in Table 1 are behavioural characters (see above), they are less added to that latter character matrix (as coded in informative than the morphological characters. Table 2), the cladogram of Fig. 1 again results. InMatsuura & Yamane's matrix (Table 2), six The converse analysis (that is, simply adding characters are invariant in the subfamily, sixteen characters to Matsuura & Yamane's matrix, are autapomorphies, one (no. 4, clypeal teeth) while leaving the codings intact) also produces applies only to parasitic species, and one (no. 13, this result. The cladogram of Fig. 1 is thus quite ovarioles per ovary) in my view questionably robust. The ambiguity in Matsuura & Yamane's homologizes all numbers >6, as discussed matrix is partly due to arguable characters and above. I have analysed the remaining eighteen codings among their behavioural traits, as dis­ characters. Note that I have corrected the cussed above. This is exacerbated by the paucity polarity for characters 1, 6 and 30 (ocelloccipital of morphological relative to behavioural data in distance, hind coxal carina and aedegus), coded 22 their informative characters. But the additional and 36 (pedicel of embryo nest and bonding point data adduced here obviate this. Fig. 1 is the best of first envelope sheet) for Provespa, scored 36 present hypothesis of the phylogenetic relation­ and 37 (pedicel of embryo nest and expansion of ships among vespine taxa. first sheet) for the squamosa group, and corrected The diagnosis of the cladogram (Farris, 1979, the error in no. 25 (occipital carina). These are the 1980) follows: the apomorphies ofeach terminal only changes in informative characters I have taxon and ancestral node are discussed. The pre­ made. viously proposed subdivisions of each genus are The matrix is ambiguous; there are six distinct also discussed in terms of the evidence for their equally parsimonious cladograms when it is naturalness. analysed using the x:wAGNER routine. The length is 29; the consistency is 0.69. However, the trees only disagree on the interrelationships Cladistic diagnoses of the Vespula subgenera/squamosa group/ Vespa Dolichovespula. The strict consensus tree is given in Fig. 2. Vespula in the broadest sense The monophyly of Vespa is shown by seven (i.e. including Dolichovespula) is unresolved. characters in Table 1: 1, elongate prestigma in The Adams consensus tree differs in having the forewing; 9, elongate ocelloccipital dis­ Vespula and the squamosa group as sister­ tance; 22, aedeagus apex transversely project­ groups. Thus, Vespula in the broadest sense if ing; 32, larval spiracular collar processes monophyletic on every tree. Provespa is always branched (convergent in Dolichovespula); 34, its sister-group, and Vespa is always the sister­ nest paper brittle and brown (convergent in group to this. some species of Paravespula); 35, nest envelope The tree of Fig. 1 accounts for the data of scallopped (variable within Vespa, convergent in Table 2 equally well. When the cladogram is Paravespula); 44, royal court (convergent in diagnosed with the informative characters from Provespa). Table 2, a length of 29 also results. When the Bequaert (1930) divided Vespa into four converse diagnosis is made, the six trees range in species groups. Group 1 (analis, parallela and length from 35 to 42. The tree among the six nigrans) was defined with a character interpret­ which does best has Vespula s.l. as a group, but able as apomorphic: female clypeus toothed resolves it as: Vespula + (squamosa group + medially. However, Bequaert later (1939) (Rugovespula + Paravespula)). Agreement reduced the latter two names as colour forms of between different data sets on one tree many be analis. Group 2 (binghami) also had an apomor­ used to choose among trees even if one or both phy: enlarged ocelli. Group 3 (tropica, affinis, data sets are ambiguous (Schuh & Farris, 1981). basalis and mandarinia are considered valid Phylogenetic relationships and classification ofVespinae 423 now) had the apomorphy of an elongate malar sharply angled) and 30 (aedeagus apex deeply space, but Bequaert observed that the genus divided) from Table 2 are autapomorphies of showed a gradual range ofvariation, and that the this genus. Vespa binghami also has enlarged character varied within species. Van der Vecht ocelli, and is presumably nocturnal, but the (1957) rejected this character as a basis for divi­ ocelli are not nearly as enlarged as Provespa, nor sion of the genus. Group 4, the remaining is the rest of the head similar. . species, was defined only by the absence of the features of the other three groups. Obviously, it Dolichovespula + Vespula is paraphyletic, and this system must be aban­ doned. Van der Vecht (1957) also stated that the The grouping of Vespula and Dolichovespula system was unworkable. is supported by three characters in Table 1: 11, Van der Vecht (1959) described the subgenus loss ofthe strong seta on the third segment ofthe Nyctovespa for binghami, citing its 'isolated labial palpus; 16, reduction of the scutallamella; position' within Vespa. Two of the characters 42, presence of a twisted pedicel in embryo mentioned in the diagnosis, enlarged ocelli and nests. In addition, character 36 of Matsuura and absence of tyloides, are autapomorphies. But Yamane (bonding point of first sheet of embryo the other, male metasomal sterna not emargi­ nest); although not coded for Rugovespula, is nate, appears to be primitive within Vespinae. parsimoniously interpreted as supporting this Consequently, binghami may be the sister-group clade. of the rest of the Vespinae. I have not verified that the apomorphic state characterizes all of the Dolichovespula remaining Vespa, and van der Vecht (1959: 212) only stated that the apomorphic condition was Dolichovespula is also monophyletic, as present 'as a rule'. It thus is presently uncertain established by eight characters in Table 1: 6, whether the division is natural, but in any event, occipital carina ventrally effaced (convergent in recognition of a formal category for one species Vespula + squamosa group); 8, elongate malar which expresses this division scarcely seems space; 18, coxal carina reduced (convergent in justified. I am therefore synonymizing this sub­ Provespa); 25, apical attenuation of the genus with Vespa (new synonymy). aedeagus; 30, larval mandible bidentate; 31, larval frontoclypeal suture dorsally effaced; 32, larval spiracular collar processes branched (con­ Provespa + (Dolichovespula + Vespula) vergent in Vespa); 37, comb suspensoria rib­ bonlike throughout nest. In addition, four The grouping of Provespa with Vespula and characters from Matsuura and Yamane's matrix Dolichovespula is supported by four characters are unique derived features in this genus: 17, in Table 1: 3, vertical orientation of M with development of small processes on the inner respect to m-cul in the forewing; 5, clustering of surface of the larval mandible; 23, final coating the hamuli of the hindwing apically of the fork of of embryo nest pedicel lacking (an unclear char­ Rand RS; 13, reduction of the pronotal carina acter which remains to be reconciled with char­ dorsally; 14, loss of the pretegular carina. acter 41 from Table 1, pedicel not coated with glossy oral secretion, also scored for Vespula); 31, shape of larval labrum; 37, mode of expan­ Provespa sion of the embryo nest. The monophyly of Provespa is shown to be monophyletic by Dolichovespula in the present sense has been eight characters in Table 1: 2, elongate first recognized since Bequaert (1930), but was ques­ submarginal cell; 7, dorsal reduction of the tioned recently by Varvio-Aho et al. (1984). occipital carina; 10, enlarged ocelli (and dorsal However, Carpenter (1987) showed that the approximation of the compound eyes); 18, loss data used by these authors were so ambiguous as of the hind coxal carina (convergent in to be phylogenetically uninformative at this Dolichovespula); 26, loss of the parameral pro­ level. The present results further support that cess (convergent in Vespula s.s.); 43, swarming; conclusion. 44, royal court (convergent in Vespa); 46, noc­ The subdivision of Dolichovespula is a matter turnal habits. In addition, characters 29 (TI not of confusion. This is due primarily to Bliithgen 424 lames M. Carpenter

(1938, 1943). He followed Bischoff (1931b) in the parameral process and basally narrow recognition of a polyphyletic taxon Vespula aedeagus (cf. Archer, 1981c: Fig. 3) as for the social parasites Vespula austriaca, autapomorphies. But the other two subgenera Dolichovespula omissa and D.adulterina (Bis­ are not monophyletic. Boreovespula is mini­ choff incorrectly used the objective junior syn­ mally paraphyletic in terms of Pseudovespula, onym Pseudovespa for this genus). Bliithgen and the latter may be polyphyletic. (1938) raised Dolichovespula to genus, with the Boreovespula is definable only in terms of the subgenus Paravespula described for Vespula ger­ absence of defining features of both Pseudo­ manica, rufa and lJulgaris. Both Dolichovespula vespula and Metavespula. Yamane (1976) and Paravespula were thus paraphyleticin terms treated an 'indistinct' parameral process as an of Vespula sensu Bliithgen. Bliithgen (1943) autapomorphy of this subgenus, but in fact it is described the subgenera Metavespula (for well developed in nearctic species. Thus sylvestris) and Boreovespula (for norwegica and Boreovespula may be paraphyletic in terms of saxonica) in Dolichovespula, and raised Pseudo­ both Pseudovespula and Metavespula. Pseudo­ vespula (following Weyrauch, 1937) and Para­ vespula, including omissa and adulterina, has vespula to generic rank (with the new subgenus several apomorphic traits common to its two Allovespula in the latter for rufa). Thus, all but species, but also shared with arctica. These are four European Vespinae were in a separate the sharply pointed apical teeth, loss of tyloides, genus or subgenus, and Bliithgen had created an dorsal reduction of the occipital carina, and of instability in vespine classification similar to that course the parasitic habits. However, all of these which he created in the Eumeninae (cf. Carpen­ traits are apparently uninformative at this level: ter & Cumming, 1985). His system was even­ the sharp clype"al teeth also occur in Vespula tually widely adopted, even in North America austriaca and so this is correlated with the parasi­ (Wagner, 1978), except that generally the para­ tic habits. Loss of tyloides occurs in other sitic species have been placed in the same sub­ Dolichovespula (sylvestris) , and the develop­ genus as their hosts (Guiglia, 1948). Although ment of the occipital carina varies in the genus. this step removed some obvious paraphyly, this The parasitic species have the apomorphic, has itself resulted in some nomenclatural errors attenuate aedeagus typical of Boreovespula and (viz the use of the name Boreovespula Bliithgen, Metavespula, and so regardless of the relation­ 1943 for a subgenus of Dolichovespula including ships among the parasites, they are a component the species D.adulterina, type species ofPseudo­ ofthis c1ade. I have not investigated the relation­ vespula Bischoff, 1931b; cf. Guiglia, 1948, 1971; ships of the individual parasitic species further, Edwards, 1980; Matsuura & Yamane, 1984). but note that omissa is apparently not closely So the question arises: are the subgenera in related to its host, sylvestris. In addition to not Dolichovespula natural groups? Ifthey are, then sharing the apomorphies of Metavespula noted the question of their formal recognition is a sub­ here, the allozyme data of Varvio-Aho et al. jective matter, but if not, some or all must be (1984) consistently indicated closer relationship sunk. Of the four subgenera, Dolichovespula of omissa to saxonica and norwegica despite including maculata and media, is clearly mono­ multiple trees (Carpenter, 1987). phyletic. Ithas several autapomorphies, includ­ In summary, the monophyly of no more than ing the striate pronotum and propodeum, and two subgeneric groups can be shown, the male genitalia. The aedeagus has a pair of Dolichovespula and a group composed of the medial lobes which are unique in the Vespinae remaining species, for which Pseudovespula is (cf. Bequaert, 1932: Fig. 1). Further, the the senior synonym. I do not consider the recog­ remainder of the genus may form a monophyle­ nition of these subgenera to serve any other tic group in relation to this subgenus. The purpose than to further overburden vespid aedeagus is far more attenuate, and the apical nomenclature, and synonymize them below. division between the rods not as pronounced in the remaining species (see figures in Bequaert, Vespula 1932). These features appear to be apomorphic. Metavespula, including sylvestris and asiatica Vespula is shown to be monophyletic by five (Archer, 1981c; not omissa), is apparently also characters in Table 1: 12, loss of tyloides in the monophyletic, with the distal prolongation of male antennae; 13, complete loss ofthe pronotal Phylogenetic relationships and classification ofVespinae 425 carina; 24, rods forming aedeagus fused apically; Rugovespula. is monophyletic as shown by 33, nesting site in cavity or underground; 41, character 17, propodeum striate in queens. In pedicel of embryo nest not coated with glossy addition, the distally short second submarginal oral secretion. In addition, two characters from cell (see above), and medially constricted Table 2 are autapomorphies of the genus: 24, aedeagus (cf. Yamane & Tano, 1985: Fig. 1c; embryo nest with disc (not scored in Archer, 1982: Fig. 4) are apomorphic. The Rugovespula and the squamosa group); 35, lar­ sharply pointed female svr dorsolateral process val hunger signal low frequency scraping (also may also be apomorphic in relation to Para­ not scored in Rugovespula and the squamosa vespula (Archer, 1982: Fig. 1), but the polarity is group). Matsuura and Yamane's character 23, unclear. final coating on pedicel of embryo nest adhesive Paravespula is monophyletic as shown by five (not scored in Rugovespula and the squamosa characters in Table 1: 23, aedeagus apex spoon­ group) is an apomorphy ofthis clade when Fig. 1 shaped; 27, parameral process fingerlike; 35, is diagnosed with the data from Table 2, but as nest envelope scallopped (convergent in Vespa) , pointed out previously under Dolichovespula, 40, large colonies (convergent in the squamosa its relation to character 41 from Table 1 (pedicel group); 44, use of vertebrate carrion. In addi­ not coated with glossy oral secretion) is unclear. tion, three characters from Table 2 support this Within Vespula, Vespula s.s. and the group: 21, cocoon cover complex; 32, larval squamosa group are sister-groups, sharing three spiracu}ar collar processes short and simple ­ characters: 6, occipital carina ventrally effaced inferred to be a reversal; 34, colony decline in (convergent in Dolichovespula); 28, volsella late fall (convergent in the squamosa group). short; 29, digitUs slender and fingerlike. Archer In relation to the subdivision of Vespula of (1981b) also considered the squamosa group to Yamane et at. (1980), first, contrary to the state­ be close to Vespula s.s., but confused ment of these authors, the recognition of only plesiomorphic and apomorphic character states. the rufa and vulgaris groups is a natural system. Consequently, nearly all of the characters he The rufa group (Vespula s.s.) corresponds to cited to support this are plesiomorphic in the two their groups 5-7, and within this group, the groups. Archer did include the short volsella, squamosa group (their group 5) and the which is a synapomorphy. remainder (groups 6 and 7) are sister-groups. Vespula s.s. is monophyletic, as shown by four However, group 6 (Allovespula) is paraphyletic characters in Table 1: 26, loss of the parameral in terms of austriaca (type of Vespula). The lat­ process (convergent in Provespa); 37, suspen­ ter species was separated categorically because soria of first comb ribbonlike; 38, one worker of its socially parasitic habits and attendant cell comb; 39, worker and queen cells on structural peculiarities (e.g. Bliithgen, 1943). different combs.' These traits are autapomorphies, and further, The squamosa group is also monophyletic, occur also in the parasitic species of and thus squamosa and sulphurea are sister­ Dolichovespula, so they are not even unique. groups. The only character establishing this in Vespula austriaca shares with the species in Table 1 is 40, large colonies; and from Table 2: Allovespula the apomorphies discussed above, 34, colony decline in late fall. Both are con­ but the latter subgenus is definable only in terms vergent in Paravespula. In the squamosa group's of the absence of the derived traits of austriaca. case they are presumably correlated with the Recognition of Allovespula must therefore be southerly nearctic distribution of the group. abandoned, as has been done even by some Other characters not considered in this analysis authors who recognized subgenera in Vespula show that these two species are sister-groups, (Guiglia, 1948, 1971,1972; Edwards, 1980; Mat­ such as the mesoscutal stripes. suura & Yamane, 1984). Bliithgen (1961) Paravespula and Rugovespula are also sister­ defended the categorical separation of social groups, sharing three unique characters in Table parasites from their hosts in Vespidae, citing 1: 19, female svrwith a dorsolateral process; 20, similar practice in Apidae (vii Bombus vs. male TVII disc depressed; 21, male SvrI trans­ Psithyrus). This is unacceptable in a phylogene­ verse. Archer (1982) listed these characters, as tic system (Hennig, 1966); it results in well as the long volsella (a symplesiomorphy), as paraphyly. distinguishing these two groups. As for Paravespula (groups 1-4 of Yamane et 426 lames M. Carpenter at., 1980), Rugovespula (group 4) and the done by the use of species groups (Bequaert, remainder are sister-groups. Their group 1 1932), which do not constitute a nomenclatural (jlaviceps and shidai) is monophyletic, as shown burden, and I am therefore synonymizing all of by theventrally evanescent occipital carina (c.on­ the subgenera of Vespula. vergent elsewhere) and a small ventral proJec­ tion of the paramere (Yamane et at., 1980: Fig. ~nd 20). Group 2 (vulgaris, structor, maculi!rons Classification flavopilosa) may also be monophyletic, havmg the synapomorphy ofthe aedeagus with pointed In the phylogenetic system, cladograms are lateral projections basad of the apex (Bequaert, exactly reflected in classification (Hennig, 1932: Fig. 2). However, similar projections are 1966). Not only are cladistic classifications most found in the flaviceps group, differing only in useful for evolutionary studies, as other types that they are more basally directed (cf. Yamane are misleading (Hennig, 1966), but such et at., 1980: Figs. 7-9). Itis not clear which is the classifications best reflect anagenetic informa­ primitive condition, but presence of the projec­ tion, as shown by Farris (1979). tions is certainly apomorphic. These groups are However, the matter offormal ranking retains thus closely related, which is also shown by a subjective element. Namely, which mono­ the mesally straight third mandibular tooth phyletic groups will be given formal nomina? Dolic.h~­ (generally curved in other Vespula and Recognition ofjustfour genera is most prevelant vespula) and brittle brown nest paper. But It IS in current systems; only a few European authors not clear if the last group (group 3, germanica recognize Paravespula as a genus. This taxon and pensylvanica) is monophyletic- the traits by must include Rugovespula in order for Vespula which it is defined (complete occipital carina, to be monophyletic, and as shown in the 'Cladis­ third mandibular tooth mesally concave, tic diagnoses' these taxa are distinguished by aedeagus without lateral projections, nest paper relatively few apomorphies, which moreover grey and pliable) are plesiomorphic. Thus th.e may be considered relatively less important that naturalness of this subdivision of Paravespula IS the apomorphies distinguishing Vespula s.l. But unclear. as shown in the section on cladistic diagnoses, To sum up, aside from Allovespula, the sub­ recognition of several subgenera in genera recognized within Vespula are natural Dolichovespula and Vespula is consistent with groups. The squamosa group is monophyletic, phylogenetic principles. However, this results in but contrary to MacDonald & Matthews (1975, nomenclatural burden with little discernible 1984), it is closely related to the rufa group. And benefit (cf. Menke & Carpenter, 1984; Menke, although the squamosa group is 'distinct' in rela­ 1985). Therefore in the following classification tion to its sister-group, that is, it has some not only am I deleting all unnatural taxa, I am derived traits, recognition of a new subgenus for sinking all of the subgenera. the two species (Edwards, 1980; Archer, 1982) is The classification is sequenced (Nelson, 1972; scarcely justified. It shares outstanding syn­ Wiley, 1979); each genus is the sister-group of all apomorphies with the rufa group, while it is of the following genera taken as a group. It thus separated from the rufa group primarily by mirrors the cladogram exactly. The resulting plesiomorphies. And while recogniti?n of classification is very similar to that of Bequaert Rugovespula separate from Paravespula IS C?~­ (1930), with Dolichovespula in Bequaert's sense sistent with monophyly, that does not mean It IS raised to generic rank (Duncan, 1939). Full syn­ well conceived. Most of the features by which onymies are given with each genus, as the sink­ Rugovespula was originally distinguished from ing of all the subgenera involves some new Paravespula (Archer, 1982) are plesiomorphies, changes in status. and the diVIsion separates just two species in Rugovespula (or three, Yamane & Tano, 198~) Vespa L., 1758: 343. Type species Vespa crabro from the other eight species in Paravespula. This L., 1758. Designated by Lamarck, 1801:271. sort of oversplitting has rendered the classifica­ Macrovespa Dalla Torre, 1904: 64. Type tion of the Eumeninae chaotic (cf. Carpenter & species Vespa crabro L., 1758. Designated Cumming, 1985), and is resisted here. In fact, in by Bequaert, 1930: 64. my view the subdivision of Vespula is better Nyctovespa van der Vecht, 1959: 210. Type Phylogenetic relationships and classification ofVespinae 427

species Vespa binghami Buysson, 1905. nized; the system was not phylogenetic in Vespa Original designation. syn.n. and is ambiguous in Dolichovespula. Further Provespa Ashmead, 1903: 182. Type species study is required in Dolichovespula, as the nor­ Vespa dorylloides Saussure, 1853 (= Vespa wegica and sylvestris groups sensu Bequaert may anomala Saussure, 1853). Monotypic. yet prove to be monophyletic. But the morpho­ Dolichovespula Rohwer, 1916: 642. Type logical grounds for recognizing the three afore­ species Vespa maculata L., 1763. Original mentioned species groups remain good charac­ designation. ters, and behavioural data gathered since Pseudovespula Bischoff, 1931b: 346. Type Bequaert's time generally confirm the other species Vespa norwegica var. adulterina groups, as shown above. The oversplit classifica­ Buysson, 1905. Original designation. tion produced by Bliithgen merely applies for­ Boreovespula Bliithgen, 1943: 149. Type mal nomina to Bequaert's original groups, with species Vespa norwegica F., 1781. Original the addition of some unnatural taxa. The result designation. has been an inflexible, confusing nomenclature, Metavespula "Bliithgen, 1943: 149. Type for contrary to Edwards (1980), the species species Vespa silvestris [I] Scopoli, 1763. group and subgenera concepts are not Original designation. equivalent. One falls under the rules of the VespulaThomson, 1869: 79. Type species Vespa International Code of Zoological Nomencla­ austriaca Panzer, 1799. Designated by ture. Because some of the groups must be dis­ Ashmead, 1902: 164. pensed with as being unnatural, formal Pseudovespa Schmiedeknecht, 1881: 314. synonymies are required, and must henceforth Type species Vespa austriaca Panzer, 1799. accompany taxonomic works on the subfamily. Monotypic. Continued use ofBequaert's system, as has been Paravespula Bliithgen, 1938: 271. Type done in North America, would not have species Vespa vulgaris L., 1758. Original required these actions when the system was designation. modified. Allovespula Bliithgen, 1943: 149. Type The present arrangement, as a phylogenetic species Vespa rufa L., 1758. Original one, is of particular use in behavioural studies, designation. . upon which much current interest is focused. Rugovespula Archer, 1982: 261, 264. Type This is shown in the character matrices, where species Vespa koreensis Radoszkowski, many of the behavioural features discussed dur­ 1887. Original designation. syn.n. ing the past decade are for the first time separ­ ated into primitive and derived states. Some of As established in the 'Cladistic diagnoses', the doubts expressed by various behaviourists within Dolichovespula two monophyletic sister­ concerning Bequaert's system result from a groups can be recognized: the maculata species failure to consider all of the available evidence group (maculata and media) and a group com­ (viz morphology), but also result from confusion prised of the remaining species, which may be regarding the evolutionary polarity of termed the norwegica group. Within Vespula, behavioural features. For example, both the the austriaca (rufa) group and the vulgaris group rufa group and vulgaris group are seen to exhibit are sister-groups. a mosaic of primitive and derived behavioural traits, and the nature of each trait must be taken Discussion into account when attempting to explain the basis for differences in these traits. Thus, Mac­ The data presented here strongly support the Donald (1977) suggested that the single worker relationships depicted in Fig. 1. The expression cell comb in the rufa group was a consequence of of those relationships by use ofspecies groups in reliance on live prey. The first trait is derived but addition to formal taxa is one similarity to the the second is primitive, shared with most other system of .Bequaert (1930, 1932). Another is vespines, and so clearly is not a cause of the some of the groups themselves: the maculata single worker comb. Consideration of such feat­ group in Dolichovespula, the austriaca (rufa) ures in the context of the phylogenetic system is and vulgaris groups in Vespula are natural. This required for critical understanding of the evolu­ is not true of all of the groups Bequaert recog- tionary basis of behaviour. 428 lames M. Carpenter

The Vespinae seem to be a particularly tract jackets. Economic Impact and Control of Social able group for further investigation on species­ Insects (ed. by S, B, Vinson), pp, 353-412, level relationships. This should be very reward­ Praeger, New York. Archer, M.E. (1981a) A new species of ing especially for the study of the biogeography Dolichovespula and subspecies of D. pacifica of these wasps. A Southeast Asian centre of (Hyipenoptera: Vespidae) from China. Entomon origin has generally been assumed for Vespinae 5,341-344 (1980). ' (van der Vecht, 1964), but Carpenter (1981) Archer, M.E. (1981b) the Euro-Asian species of the V~pula r~fa. group (Hymenoptera, Vespidae), pointed out that this was based on fallacious With descnptions of two new species and one new grounds. Carpenter (1981) concluded that the SUbspecies. Kontya, 49, 54-64. distribution ofthe subfamily as a whole as laura­ Archer, M,E. (1981c) Taxonomy of the sylvestris sian (Bequaert, 1932, also inferred a 'northern group (Hymenoptera: Vespidae, Dolichovespula) common origin' for the subfamily), and the pre­ with the introduction of a new name and notes on distribution. Entomologica Scandinavica 12 187- sent system does not permit any more precise 193. ', inferences. This is because the connections Archer, M.E. (1982) A revision of the subgenus between Eurasia and North America are within Rugovespula novo of the genus Vespula the species groups in both Vespula and (Hymenoptera, Vespidae). Kontya, 50, 261-269. Dolichovespula, and the connections between Ashmead, W.H. (1902) Classification of the fossorial, preda~eous and parasitic wasps, or the superlamily Eurasia and the Oriental region are completely VespOIdea. Paper no, 6. Family XXVIII. unclear in the absence of a species level hypo­ Vespidae. Canadian Entomologist, 34, 163--166. thesis in Vespa. Inference of patterns in histori­ Ashmead, W.H. (1903) Provespa, a new genus in the cal biogeography is a problem in character state Vespidae. Entomological News, 14, 182. Bequaert, J. (1930) On the generic and subgeneric optimization (Mickevich, 1981), and because divisions of the Vespinae (Hymenoptera). Bulletin the relevant area transformations occur within ofthe Brooklyn Entomological Society, 25,59-70. the groups established here, the state assign­ Bequaert, J. (1932) A tentative synopsis ofthe hornets ments are ambiguous. Elucidation of the and yellow-jackets (Vespinae; Hymenoptera) of America. Entomologica Americana (n.s.), 2, phylogenetic relationships among all the species 71-138. is thus necessary for further progress in under­ Bequaert, J. (1939) The Oriental Vespa analis standing of vespine biogeography. Fabricius and its color fonus, with a note on the synonymy of Vespa esakii Sonan and Vespa /or­ mosa,na Sonan. Transactions of the American Acknowledgments Entomological Society, 65, 37-42. Birula, A. (1930) Uber die russischen Wespen und ihre geographische Verbreitung. Annuaire du This study was begun while the author had a Musee Zoologique de l'Academie des Sciences de Smithsonian Postdoctoral Fellowship. I thank l'URSS, 31,291-339. A. S. Menke, K. V. Krombein and the Ashmead Bischoff, H. (1931a) Eine bisher verkannte und eine neue Art der Gattung Vespa aus der mit­ Club for assistance while working at the U.S. teleuropaischen Fauna (Hym. Vesp.). Mit­ National Museum. I am grateful to A. S. Menke, teilungen der Deutschen Entomologischen K. G. Ross, R: W. Matthews, and especially J. Gesellschaft, 2,6-7. Kojima, for reading and commenting on the Bischoff, H. (1931b) Zur Kenntnis der Gattung manuscript. Mie Norton translated the Japanese Pseudovespa. Sitzungsberichte der Gesellschaft Naturforschender Freunde, 329-346 (1930). data tables. Use of the computer was supported Bliithgen, P. (1938) Systematiches Verzeichnis der by NSF Grant BSR-8508055 to the author. Faltenwespen Mitteleuropas, Skandanaviens und Englands. Konowia, 16, 270--295. Bliithgen, P. (1943) Taxonomische und biologische References Notizen iiber palaarktische Faltenwespen (Hym. Vespidae). Stettiner Entomologische Zeitung, 104, Adams, E.N. (1972) Consensus techniques and the 149-158. comparison of taxonomic trees. Systematic Zool­ Bliithgen, P. (1961) Die Faltenwespen Mitteleuropas ogy, 21,390--397. (Hymenoptera, Diploptera). Ahandlungen der Akre, R.D.,'Greene, A., MacDonald, J.F., Landolt, Deutschen Akademie der Wissenschaften zu Berlin P.J. & Davis, H.G. (1981) The yellowjackets of (Klasse fiir Chemie, Geologie und Biologie), 2, America north of Mexico. U.S.D.A. Handbook, 1-250, 552,1-102. Buysson, R. du (1905) Monographie des guepes ou Akre, R.D. & MacDonald, J.F. (1986) Biology, Vespa. Annales de la Societe Entomologique de economic importance and control of yellow- France, 73,485--556 (1904). Phylogenetic relationships and classification ofVespinae 429

Carpenter, J.M. (1981) The phylogenetic relation­ tentrionale. Faune de ['Europe et du Bassin ships and natural classification of the Vespoidea Mediterraneen, 6, 1-181. (Hymenoptera). Systematic Entomology, 7, 11-38 Hennig, W. (1966) Phylogenetic Systematics. Univer­ (1982). sity of Illinois Press, Urbana. Carpenter, J .M. (1987) On 'The evolutionary genetics Iwata, K. (1955) The comparative anatomy of the of social wasps' and the phylogeny ofthe Vespinae ovary in Hymenoptera. Part I. Aculeata. Mushi, (Hymenoptera, Vespidae, Vespinae). Insectes 29,17-34. Sociaux (in press). Kluge, A.G. & Farris, J.S. (1969) Quantitative phyle­ Carpenter, J .M. & Cumming, J .M. (1985) A character tics and the evolution of anurans. Systematic Zool­ analysis of the North American potter wasps ogy, 18, 1-32. (Hymenoptera: Vespidae; Eumeninae). Journal of Kugler, J., Orion, T. & Ishay, J. (1976) The number of Natural History, 19, 877-916. ovarioles in the Vespinae (Hymenoptera). Insectes Dalla Torre, K.W. (1904) Vespidae. Genera Insec­ Sociaux, 23, 525-533. torum, 19, 108 p. Lamarck, J.B.P.A. de Monet (1801) Systeme des Duncan, C.D. (1939) A contribution to the biology of Animaux sans Vertebres, I, Paris. North American vepsine wasps. Stanford Univer­ Linnaeus, C. (1758) Systema Naturae, I, 10th edn. sity Publications, Biological Sciences, 8, 1-272. Laurentii Salvii, Holmiae. Eck, R. (1981) Zur Verbreitung und Variabilitat von MacDonald, J.E. (1977) Comparative and adaptive Dolichovespula norwegica (Hymenoptera, aspects of vespine nest construction. Proceedings Vespidae). Entomologische Abhandlungen, 44, of the Eighth International Congress of the 133--152. 1. U.S.S.1., Wageningen, pp. 169-172. Eck, R. (1984a) Uber die Verbreitung von MacDonald, J.E., Akre, R.D. & Hill, W.B. (1974) Dolichovespula pacifica (Birula) und den Status Comparative biology and behavior of Vespula von Dolichovepula pacifica xanthicincta Archer atropilosa and V.pensylvanica (Hymenoptera: (Hymenoptera, Vespidae). Entomologische Vespidae). Melanderia, 18, 1-66. Abhandlungen, 47, 195-199. MacDonald, J.E. & Matthews, R.D. (1975) Vespula Eck, R. (1984b) BestimmungsscWiissel fur die Acten squamosa: a yellow jacket wasp evolving toward der Gattung Dolichovespula Rohwer, 1916. parasitism. Science, 190, 1003--1004. Entomologische Abhandlungen, 48, 35--44. MacDonald, J.E. & Matthews, R.D. (1984) Nesting Edwards, R. (1980) Social Wasps. Rentokil, East biology of the southern yellowjacket, Vespula Grinstead. squamosa (Hymenoptera: Vespidae): Social para­ Farris, J.S. (1970) Methods for computing Wagner sitism and independent founding. Journal of the trees. Systematic Zoology, 19,83-92. Kansas Entomological Society, 57, 134-151. Farris, J .S. (1979) The information content of the Matsuura, M. (1983) Some biological aspects of the phylogenetic system. Systematic Zoology, 28, 483-­ nocturnal vespine genus Provespa. Ecological 519. Study on Social Insects in Central Sumatra with Farris, J.S. (1980) The efficient diagnoses of the Special Reference to Wasps and Bees, pp. 27-29. phylogenetic system. Systematic Zoology, 29, 386­ Sumatra Nature Study (Entomology), Kanazawa 401. University. Farris, J.S. (1982) Outgroups and parsimony. Sys­ Matsuura, M. (1984) Comparative biology of the five tematic Zoology, 31, 328--334. Japanese species of the genus Vespa (Hymenop­ Greene, A. (1979) Behavioural characters as indica­ tera, Vespidae). Bulletin ofthe Faculty ofAgricul­ tors of yellowjacket phylogeny (Hymenoptera: ture, Mie University, 69, 1-131. Vespidae). Annals ofthe Entomological Society of Matsuura, M. (1985) Life history of the nocturnal America, 72,614-619. vepsine, Provespa anomalla [I]. Evolutionary Greene, A,. (1984) Production schedules of vespine Ecology ofInsects in Humid Tropics, Especially in wasps: An empirical test of the bang-bang optim­ Central Sumatra, pp. 27-36. Sumatra Nature Study ization model: Journal ofthe Kansas Entomologi­ (Entomology), Kanazawa University. cal Society, 57,545-568. Matsuura, M. & Yamane, Sk. (1984) Comparative Greene, A., Akre, R.D. & Landolt, P. (1976) The Ethology ofthe Vespine Wasps. Hokkaido Univer­ aerial yellowjacket, Dolichovespula arenaria sity, Sapporo [in Japanese]. (Fab.): Nesting biology, reproductive production, Menke, A.S. (1985) Subgenera vs. species groups. and behavior (Hymenoptera: Vespidae). Melan­ Sphecos, 10, 11-13. deria, 26, 1-34. Menke, A.S. & Carpenter, J. (1984) Nuclearbombus, Guilia, D. (1948) Le Vespe d'Italia. Memorie della new subgenus (or how to eliminate bumblebee sub­ SocietiJ. Entomologica Italiana, 27, Supplemento, genera and learn to love the Bombus). Bulletin of 5-84. the Entomological Society ofCanada, 16, 130. Guiglia, D. (1971) A concise history ofVespidae sys­ Mickevich, M.F. (1981) Quantitative biogeography. tematics in the Old World (Hymenoptera). Advances in Cladistics: Proceedings of the First Entomological Essays to Commemorate the Retire­ Meeting ofthe Willi Hennig Society (ed. by V. A. ment of Professor K. Yasumatsu, pp. 113-117. Funk and D. R. Brooks), pp. 209-222. New York Hokuryukan Publishing Company, Tokyo Botanical Garden, Bronx Guiglia, D. (1972) Les guepes sociales (Hymenoptera Nelson, G.J. (1972) Phylogenetic relationship and Vespidae) d'Europe Occidentale et Sep- classification. Systematic Zoology, 21, 227-231. 430 lames M. Carpenter

Nelson, G. (1979) Cladistic analysis and synthesis: Yamane, Sk. & Tano, T. (1985) Supplements to the Principles and definitions, with a historical note on Vespula fauna of Taiwan (Hymenoptera, Adanson's 'Familles des Plantes' (1763-1764). Sys­ Vespidae). Kontyil, 53,420-425. tematic Zoology, 28, 1-21. Yamane, Sk., Wagner, R.E. & Yamane, S. (1980) A Rohwer, S.A. (1916) Vespidae. In: Guide to the tentative revsion of the subgenus Paravespula of Insects of Connecticut. Part Ill. The Hymenop­ Eastern Asia (Hymenoptera: Vespidae). Insecta tera, or Wasp-like Insects of Connecticut. Bulletin Matsumurana (N. S.), 19, 1-46. of the Connecticut State Geological and Natural History Survey, 22, 640-644. Accepted 20 February 1987 Rousch, C.F. & Akre, R.D. (1978) Nesting biologies and seasonal occurrence of yellowjackets in northeastern Oregon forests (Hymenoptera: Vespidae). Melanderia, 30,57-94. Appendix Schmiedeknecht, O. (1881) Uber einige deutsche Vespa-Arten. Entomologische Nachrichtenblatt, 7 List oftaxa examined (22), 313-318. Schuh, R.T. & Farris, I.S. (1981) Methods for The status of some of the names listed here as investigating taxonomic congruence and their species is a matter of considerable current dis­ application to the Leptopodomorpha. Systematic Zoology, 30, 331-351. agreement (cf. Archer, 1981a, 1982; Eck, 1981, Short, I.R. (1952) The morphology of the head of 1984a, b; Matsuura & Yamane, 1984; Yamane larval Hymenoptera with special reference to the & Tano, 1985; Yamane et at., 1980). I have not head of Ichneumonoidea. Transactions of the followed any of these authors in all respects Royal Entomological Society of London, 103, regarding classification; many of their tax­ 27--66. Spradbery, J.P. (1973) Wasps. University of onomic decisions are in my view insufficiently Washington Press, Seattle. supported. All specimens listed are deposited in Thomson, C.G. (1869) Opuscula Entomologica, 1,78­ the Museum of Comparative Zoology and the 82. V.S. National Museum. Larvae are in the collec­ van der Vecht, J. (1957) The Vespinae of the Indo­ Malayan and Papuan areas (Hymenoptera, tion of the British Museum. Vespidae). Zoologische Verhandelingen, 34, 1-83. ~<;?d' van der Vecht, 1. (1959) Notes on oriental Vespinae, Vespa affinis (L.) (9 subspecies) including some species from China and Japan analis F. (8 subspecies) ~<;?d' (Hymenoptera, Vespidae). Zoologische basalis Smith ~ <;? d' Mededelingen, 36, 205-232. bellicosa Saussure <;? van der Vecht, J. (1964) The geographical distribution bicolor F. (2 subspecies) ~ <;? of the social wasps (Hymenoptera, Vespidae). ~ Proceedings ofthe Twelfth International Congress crabro L. (8 subspecies) <;? d' of Entomology, London, pp. 440-441. dybowskii Andre ~ Varvio-Aho, S.L., Pamilo, P. & Pekkarinen, A. fervida Smith ~ d' (1984) Evolutionary genetics of social wasps luctuosa Saussure (3 subspecies) ~ <;? d' (Hymenoptera, Vespidae, Vespula). Insectes ~ Sociaux, 31, 375--386. mandarinia Smith (5 subspecies) <;?d' Wagner, R.E. (1978) The genus Dolichovespula and mocsaryana Buysson ~ <;? an addition to its known species of North America. multimaculata Perez (2 subspecies) ~ <;? Pan-Pacific Entomologist, 54, 131-142. orientalis L. (3 subspecies) ~ <;? d' Weyrauch, W. (1937) Zur Systematik und Biologie philippinensis Saussure ~ d' der Kuckuckswespen Pseudovespa, Pseudovespula ~ und Pseudopolistes. Zoologische lahrbilcher simillima Smith (2 subspecies) <;? (Abteilung Jilr Systematik, Okologie und tropica (L.) (12 subspecies) ~ <;? d' Geographie der Tiere), 70, 243-290. variabilis Buysson (2 subspecies) ~<;?d' Wiley, E.O. (1979) An annotated Linnaean hierarchy, velutina Lepeletier (11 subspecies) ~ <;? d' with comments on natural taxa and competing sys­ tems. Systematic Zoology, 28, 308-337. walkeri Buysson <;? Yamane, Sk. (1976) Morphological and taxonomic wilemani Meade-Waldo ~ studies on vespine larvae, with reference to the (Nyctovespa) binghami Buysson <;? phylogeny of the subfamily Vespinae (Hymenop­ ~ tera: Vespidae). Insecta Matsumurana (N.S.), 8, Provespa anomala (Saussure) <;? d' 1-45. barthelemyi (Buysson) ~ <;? d' Yamane, Sk. & Makino, S. (1977) Bionomics ofVespa nocturna Vecht ~ <;? d' analis insularis and V. mandarinia latilineata in ~ Hokkaido, Northern Japan, with notes on vespine Dolichovespula macultata (L.) <;? d' embryo nests (Hymenoptera: Vespidae). Insecta media (Retzius) ~<;?d' Matsumurana (N.S.), 12, 1-33. (Metavespula) sylvestris (Scopoli) ~ <;? d' Phylogenetic relationships and classification ofVespinae 431

(Pseudovespula) adulterina (Buysson) ~d' Vespula austriaca (Panzer) ~d' omissa (Bischoff) ~d'* (Allovespula) acadica (Sladen) ~ ~ d' (Boreovespula) albida (Sladen) ~~d' atropilosa (SIaden) ~ ~ d' alpicola Eck ~ ~ d' consobrina (Saussure) ~ ~ d' arctica (Rohwer) ~d' intermedia (Buysson) ~~d' arenaria (F.) ~ ~ d' kingdonwardi Archer ~ loekenae Eck d' rufa (L.) (2 subspecies) ~ ~ d' norvegicoides (Sladen) ~ ~ d' schrenckii (Radoszkowski) ~ norwegica (F.) ~ ~d' vidua (Saussure) ~ ~ d' pacifica (Birula) ~ 'squamosa-group' panda Archer ~ squamosa (Drury) (2 subspecies) ~~d' saxonica (F.) ~~d' sulphurea (Saussure) ~ ~ d' (Paravespula) flaviceps (Smith) (2 subspecies) *The name ingrica was originally proposed as an ~~d' infrasubspecific name ('Vespula n.saxonica morpha jlavopilosa Jacobson ~ ~ d' ingrica nov.'; Art. 45f(iii» by Birula (1930: 309). It germanica (F.) ~ ~ d' therefore must be considered to have been elevated maculifrons (Buysson) ~ ~ d' (Art. lOc), and to compete in priority from the date ~ ~ upon which it was elevated (Art. 23j). The earliest this pensylvanica (Saussure) d' could be considered to have been done is Bequaert shidai Ishikawa, Yamane and Wagner ~ ~ d' (1932: 89); Bequaert is therefore the author of this structor (Smith) ~ species (Art. 50c). Vespula ingrica Bequaert, 1932 is vulgaris (L.) (2 subspecies) ~~ d' thus a junior synonym of Vespa omissa Bischoff, (Rugovespula) arisana (Sonan) ~ 1931a. Edwards (1980) and Matsuura & Yamane ~ ~ (1984) used the name ingrica foi this species, but Eck koreensis (Radoszkowsi) (2 subspecies) (1984b) correctly used omissa. orbata (Buysson) ~