1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

The phylogenetic relationships of extant and extinct Megalyridae are analysed at the genus level. The dataset comprises seven outgroup taxa, all eight extant genera and a number of extinct taxa that have been associated with Megalyridae, including two genera from Maimetshidae, whose affinity with Megalyridae is uncertain. Analytical results are unstable because some of the fossil taxa have many missing entries. The most stable results are produced when the maimetshid taxa and Cretodinapsis are excluded. When included, these taxa fall outside crown-group Megalyridae, the maimetshid taxa being the sister of Orthogonalys (Trigonalidae). Based on the results of our analyses, we synonymize the fossil genera Rubes Perrichot n.syn. and Ukrainosa Perrichot & Perkovsky n.syn. with Prodinapsis, creating the new combinations Prodinapsis bruesin.comb. and Prodinapsis prolatan.comb. When comparing past and present distributions of Megalyridae with the results of the phylogenetic analyses, it is evident that the genera radiated in the Mesozoic, and that the family as a whole was much more widespread then. The present-day distribution is essentially relictual, with range contraction since the early Tertiary probably being the result of climate deterioration, which caused the disappearance of tropical forests throughout the Palaearctic.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Megalyridae is a small family of infrequently collected parasitic wasps currently distributed primarily in the Southern Hemisphere. It comprises 49 described species in eight extant genera. The namesake of the family, the predominantly Australasian genus Megalyra Westwood, is characterized by its members having an extremely long ovipositor (Fig. S1a in Supporting Information), up to eight times the body length (Shaw, 1990a). The genus name is derived from this feature (Shaw, 1990b), perhaps because dry-pinned female specimens with bent and/or twisted ovipositor sheaths (e.g. Figs S1a–c, S2a) can, when observed with the remainder of the ovipositor, resemble a lyre. This artefact is caused by the drying of the transversely subdivided sheaths (Vilhelmsen, 2003a) that in the living wasp protect the ovipositor proper. Because of the long ovipositor, Megalyridae are sometimes referred to as ‘long-tailed wasps' (e.g. Perrichot, 2009), but the length is actually highly variable within the family, in most cases not exceeding the length of the rest of the body. The ovipositor length is thus not exceptional by the standards of other parasitic wasps with external ovipositors (see Vilhelmsen, 2003a), except in the case of Megalyra.

The biology is unknown for most species of Megalyridae. The information available comes primarily from a few members of Megalyra. With a single exception these are parasites of the immature stages of wood-boring beetles of the families Bostrichidae, Buprestidae and Cerambycidae (Shaw, 1990a and references therein); however, M. troglodytes Naumann attacks mud-nesting sphecid larvae (Naumann, 1987). A few instances of oviposition by female Megalyra sp. were observed by Rodd (1951): wasps apparently preferred to insert the ovipositor in pre-existing cracks or frass plugs, but on one occasion the female appeared to be actively drilling. As stated by Rodd (1951), it is indeed ‘not possible to regard a female Megalyrid [sic] without giving some thought to the problem [of oviposition]’.

Megalyridae are characterized by a unique combination of characters that occur individually in some other hymenopteran families: the presence of distinct subantennal grooves, 12 flagellomeres in the antennae, and highly reduced hind-wing venation. A recent morphology-based cladistic treatment of Hymenoptera (Vilhelmsen et al., 2010) included only two megalyrid genera as terminals, Dinapsis Waterston and Megalyra. These were always retrieved as monophyletic, with moderate support. Putative apomorphies identified in this analysis include the presence of a groove posterolaterally on the propleuron forming an articulation with the pronotum, a posteriorly curved profurcal bridge, the presence of pronotal–profurcal muscles, exposed lateral metacoxal articulations, and having the site of attachment of the propodeo-second abdominal sternal muscle dorsally on the metaphragma. Most of these features are internal and all are difficult or impossible to observe without performing dissections, making them unsuitable as diagnostic characters for a taxon that includes a number of extremely rare and extinct members.

The presence of a prominent wasp-waist between the first and second abdominal segments place Megalyridae firmly within Apocrita. Vilhelmsen et al. (2010) retrieved Megalyridae frequently as the sister group of Ceraphronoidea within the larger clade Evaniomorpha s.s., a taxon that includes Evanioidea and Trigonaloidea. Megalyridae share with Ceraphronoidea the presence of an exposed anterior thoracic spiracle surrounded by pronotal cuticle and the absence of an independent prepectus (Gibson, 1985, 1999). Megalyridae display the defining feature of Evaniomorpha, namely the presence of extended condyles for the distally displaced median articulations of the meso- and metacoxae (Rasnitsyn, 1980, 1988).

Shaw (1987, 1988, 1990a) surveyed and revised the majority of extant Megalyridae; Shaw (1990b) summarized the classification, analysed the phylogeny of the extant and extinct members of the family known at the time, and discussed the results in the context of shifting continental configurations from the Mesozoic onwards. The highest species diversity (i.e. about half of the described species) is currently in the Australasian region, with 26 Megalyra spp. (Fig. S1a) and Carminator nooni Shaw (from New Britain). These genera are represented also in the adjacent Indomalayan region [M. tawiensis Petersen (from the Philippines); Carminator Shaw spp. (Fig. S2c)], together with three Ettchellsia Cameron spp. (Fig. S1c). Carminator japonicus Mita & Konishi was described recently from mainland Japan (Honshu); this is the northernmost occurrence and the only record from the Palaearctic of an extant megalyrid species (Mita et al., 2007; Mita & Konishi, 2010). The absence of any known Megalyridae from the Indian subcontinent is curious, but perhaps reflects limited sampling effort.

The diversity of Megalyridae is apparently much lower in the Afrotropical region. Until very recently, only the monotypic genus Megalyridia Hedqvist (Fig. S2b) and six Dinapsis species (Fig. S1b) were known from South Africa and Madagascar (Hedqvist, 1959, 1967; van Noort & Shaw, 2009). However, this does not reflect the true diversity of the Afrotropical fauna, but rather the comparative lack of effort hitherto expended in exploring it. Dinapsis centralis Shaw & van Noort has been reported recently from the Central African Republic. Furthermore, one of the authors of the present paper is working currently to describe c. 20 new Dinapsis spp. from sub-Saharan Africa and Madagascar (S. R. Shaw, unpublished data). This indicates that the megalyrid diversity of the Afrotropics at least equals that of the Australasian region.

Shaw (1987) described three monotypic genera from the Neotropics: Cryptalyra Shaw (not illustrated) from the tropical and subtropical part east of the Andes, and Neodinapsis Shaw (Fig. S2d) and Rigel Shaw (Fig. S2a) from Chile in the temperate Neotropics. In the two decades that have elapsed since then, two additional species of Cryptalyra have been described (Shaw; Azevedo & Tavares). In terms of species, the Neotropical megalyrid fauna seems to be very depauperate. However, the extant generic diversity is on a par with that of the other Southern Hemisphere regions, and, as in the Afrotropics, the poor species turnout might yet prove to be the result of lack of exploration.

The distribution of the extant diversity of Megalyridae is summarized in Fig. 1C. At a glance, the predominance of the family in the Southern Hemisphere is striking, and it is tempting to speculate that the current distribution was created by vicariance patterns caused by tectonic events in the Late Mesozoic, for example in the final phase of the break-up of Gondwana. The persistent failure of Megalyridae to turn up in the Nearctic and the Western Palaearctic, the two most thoroughly surveyed regions on the planet, is very unlikely to be an artefact. However, the current pattern of distribution must be supplemented by information from the fossil record to obtain the full picture.


Figure 1. Distribution of all extant and extinct genera of Megalyridae and other fossil taxa included in the analyses (maps modified from Blakey (2009)). (A) Late Early Cretaceous (c. 105 Ma) map with all Mesozoic Megalyridae and fossils of Maimetshidae (†Maimetsha and †Guyotemaimetsha) and †Cretodinapsis; (B) Middle Eocene (c. 50 Ma) map with all Tertiary records of Megalyridae; (C) Present-day map with distribution of all extant genera represented by triangles, and with the current localization of Tertiary and Cretaceous records represented by red and green crosses, respectively.

Download figure to PowerPoint

The fossil record of Megalyridae is comparatively rich, and currently is known only from the Northern Hemisphere (Fig. 1A, B). Brues (1923, 1933) described two species of Prodinapsis Brues (Fig. S4a, b) from Eocene Baltic amber. Rasnitsyn (1977) described Cretodinapsis Rasnitsyn from Cretaceous Asian amber. Both of these taxa, as well as Maimetsha Rasnitsyn (see below), were included by Shaw (1990b) in his cladistic analysis of the family and were found to be nested well within extant Megalyridae. Perrichot (2009) described new fossil taxa from Cretaceous and Tertiary amber inclusions from the western Palaearctic: the genera Megalava Perrichot (Fig. S3b), Megallica Perrichot (Fig. S3c), Megazar Perrichot (Fig. S3a), Rubes Perrichot, Ukrainosa Perrichot and Valaa Perrichot (Fig. S3d) as well as three new species of Prodinapsis (Fig. S4c, d, g).

Rasnitsyn (1975) described Maimetsha from late Cretaceous Asian amber. This taxon was placed in a separate family, Maimetshidae, and was considered by Rasnitsyn (1975, 2002) to be ‘intermediate’ between Megalyridae (based on shared, possibly plesiomorphic wing venation characters) and Ceraphronoidea (based on the derived structure of the mesosoma–metasoma articulation). Recently, Rasnitsyn & Brothers (2009) described four new taxa in Maimetshidae from the late Cretaceous of Botswana. The putative association of these taxa with at least crown-group Megalyridae seems dubious, as the maimetshid taxa differ from Megalyridae in key diagnostic characters (absence of distinct subantennal groove, antennae with more than 12 flagellomeres, anterior thoracic spiracle concealed, hind-wing venation more complete).

The possible affinity of maimetshids with Ceraphronoidea was disputed by Shaw (1988); indeed, Shaw (1990b) retrieved Maimetsha as deeply nested within extant Megalyridae, in the tribe Dinapsini. This placement was considered by Rasnitsyn & Brothers (2009) to be an artefact caused by the reduction of wing venation characters that might have been occasioned by independent body-size reduction. Evaluation of the phylogenetic position of Maimetsha is confounded by the specimen having been partly destroyed since the original description was made. Perrichot et al. (2004) described Guyotemaimetsha from Early Cretaceous French amber; they reported it to have most of the diagnostic features of Maimetsha and placed it in Ceraphronoidea sensu Rasnitsyn 2002 (i.e. including Megalyridae, Stephanidae, Trigonalidae and the extinct Maimetshidae and Stigmaphronidae). However, they left Guyotemaimetsha unassigned to family, pointing out that it shared a number of features with the Cretaceous trigonalid CretogonalysRasnitsyn, 1977. Since then, additional individuals of Guyotemaimetsha have been discovered in French Cretaceous amber, which allow for a confident assignment of this genus to Maimetshidae, and also indicate the family to be distinct from Megalyridae (V. Perrichot, unpublished data). Another problematic taxon is Cleistogastrinae, a diverse assemblage of late Jurassic and early Cretaceous compression fossils. These were considered by Rasnitsyn (1975) to be a subfamily of Megalyridae, but they do not appear to share any putative apomorphies with extant Megalyridae (Shaw, 1988). Their inclusion in Megalyridae was disputed by Grimaldi & Engel (2005) and Perrichot (2009), and they were excluded from the analysis of Shaw (1990b), a precedent that will be followed here.

Recently, Poinar & Shaw (2007) described Megalyra baltica (Fig. S4h) from Baltic amber. The discovery of a fossil member of an extant genus almost half a hemisphere removed from its closest living relatives cautions against interpreting the current distribution of Megalyridae as resulting primarily from Gondwanan vicariance events. All the fossil megalyrids described up to and including the present paper occur in the Palaearctic region; undescribed taxa from Burmese and New Jersey amber (Grimaldi & Engel, 2005), while considerably extending the geographic range of the fossils, still provide virtually no overlap with the distribution of extant taxa. This curious complementary distribution (Fig. 1C) is an artefact, probably caused by the lack of exploration of relevant fossil deposits in the Southern Hemisphere. However, the extent of the fossil occurrences outside the current range indicates that the distributional history of Megalyridae might be as much influenced by extinction and dispersal as by vicariance, the present-day distribution being essentially relictual.

Since the publication of the synthesis of the evolutionary history of Megalyridae in Shaw (1990b), a number of new extant (He, 1991; Shaw, 2003; Azevedo & Tavares, 2006; Mita et al., 2007; Shaw & van Noort, 2009; Mita & Konishi, 2010) and fossil (Poinar & Shaw, 2007; Perrichot, 2009) taxa have been described. The recent expansion of especially the fossil diversity of the family warrants an update of the analysis by Shaw (1990b). Whereas all the extant species described over the past 20 years have been accommodated in existing genera, the six fossil genera introduced by Perrichot (2009; see above) have not previously been evaluated in a cladistic context. Furthermore, to place the various fossil taxa correctly it is necessary to expand the outgroup to include more representatives of the closest living relatives of Megalyridae. In the present paper, we re-estimate the phylogeny of all extant megalyrid genera and most of the extinct taxa that have been associated with the family in the past. We then discuss the evolutionary history of Megalyridae according to the results obtained.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Taxon sampling

Shaw (1990b) assembled a dataset including all eight extant genera of Megalyridae and the three extinct genera Cretodinapsis, Maimetsha and Prodinapis, all of which were scored to the extent possible for 20 characters for external morphology; the fossil taxa could be scored only for 50–80% of these. A hypothetical ancestor was constructed to polarize characters. We substantially expand this dataset for the present analyses, both to accommodate the new fossil taxa whose position we attempt to decide, and especially to handle the expanded outgroup included here. We prefer the exemplar outgroup approach to implementing a hypothetical ancestor, deeming the former method to be less subjective and more reproducible. As a result, seven outgroup taxa and 29 characters are added to the dataset of Shaw (1990b); see Tables 1 and 2.

Table 1.  Taxa examined for the present study.
TerminalMaterial examinedMissing/inapplicable entries
  1. C, Ceraphronoidea; E, Evanioidea; O, Orussoidea; S, Stephanoidea; T, Trigonaloidea.

  2. SEM denotes taxa that were examined with scanning electron microscopy.

 Orussus (O: Orussidae)Orussus abietinus (Scopoli) 
 Schlettererius (S: Stephanidae)Schlettererius cinctipes (Cresson)0/1
 Orthogonalys (T: Trigonalidae)Orthogonalys pulchella (Cresson)0/1
 Pristaulacus (E: Aulacidae)Pristaulacus strangaliae Rohwer0/1
 Evania (E: Evaniidae)Evania albofacialis Cameron0/1
 Gasteruption (E: Gasteruptiidae)Gasteruption sp. [South Africa]0/1
 Megaspilus (C: Megaspilidae)Megaspilus fuscipennis (Ashmead)0/4
Megalyridae, extant  
 CarminatorCarminator affinis Shaw SEM1/2
 C. ater Shaw 
 C. nooni Shaw 
 CryptalyraCryptalyra depressa Azevedo & Tavares6/1
 DinapsisDinapsis oculohirta Hedqvist SEM 
 D. seyrigi Hedqvist SEM 
 D. turneri Waterston 
 Dinapsis sp. [Tanzania] 
 EttchellsiaEttchellsia piliceps Cameron 
 E. philippinensis Baltazar 
 Ettchellsia spp. [Taiwan, Thailand] SEM 
 MegalyraMegalyra fasciipennis Westwood SEM 
 M. lilliputiana Turner 
 M. longiseta Szepligeti 
 M. rufipes Erichson 
 M. rufiventris Szepligeti 
 M. shuckardi Westwood 
 M. spectabilis Shaw 
 M. tawiensis Petersen 
 M. testaceipes Turner 
 M. wagneri Fahringer SEM 
 MegalyridiaMegalyridia capensis Hedqvist SEM0/1
 NeodinapsisNeodinapsis peckorum Shaw SEM3/0
 RigelRigel chiliensis Shaw SEM 
Megalyridae, extinct  
 Cretodinapsis (Cretaceous)Cretodinapsis caucasica Rasnitsyn25/0
 Megalava (Cretaceous)Megalava truncata Perrichot19/0
 Megallica (Cretaceous)Megallica parva Perrichot5/1
 Megazar (Cretaceous)Megazar elegans Perrichot 
 Prodinapsis janzeni (Tertiary)Prodinapsis janzeni Perrichot1/2
 Prodinapsis minor (Tertiary)Prodinapsis minor Brues1/2
 Prodinapsis oesiensis (Tertiary)Prodinapsis oesiensis Perrichot0/3
 Prodinapsis pumilio (Tertiary)Prodinapsis pumilio Perrichot & Perkovsky9/2
 Prodinapsis succinalis (Tertiary)Prodinapsis succinalis Brues0/1
 Rubes (Tertiary)Rubes bruesi Perrichot0/2
 Ukrainosa (Tertiary)Ukrainosa prolata Perrichot & Perkovsky0/1
 Valaa (Cretaceous)Valaa delclosi Perrichot6/0
 Guyotemaimetsha (Cretaceous)Guyotemaimetsha enigmatica Perrichot et al.0/1
 Maimetsha (Cretaceous)Maimetsha arctica Rasnitsyn15/1
Table 2.  Data matrix for phylogenetic analyses.
  1. P, Prodinapsis.

  2. Numbers in square brackets indicate polymorphic characters. ?, unknown; -, inapplicable.

P. janzeni0111011001011112011001110100200000010-0-0100210?
P. minor0111111001011112011001110100200000010-0-0100210?
P. oesiensis0111101001011112111001110100200000010-0-0100210-
P. pumilio0??11???0101?112111000?1010??00000010-0-01002102
P. succinalis0111[01]1100111111211100111010020000001020-01002102

The choice of outgroup exemplars and character sample for the phylogenetic analyses (see below) reflects the assumption that Megalyridae belongs to Evaniomorpha s.s., a result that was retrieved frequently in the analyses of Vilhelmsen et al. (2010). Stephanoidea has been included previously in Evaniomorpha (Rasnitsyn, 1988), but is usually placed as the sister group to all other apocritan wasps in Vilhelmsen et al. (2010). Stephanoidea are included in the current taxon sample as well as representatives of all superfamilies of Evaniomorpha s.s. Finally, a representative of Orussoidea, the probable sister group to Apocrita, is included to root the analyses.

The fossil taxon sample comprises 14 terminals, all described from amber inclusions. Included are all the taxa described by Perrichot (2009): the monotypic genera Megalava, Megallica, Megazar and Valaa from the Cretaceous and Rubes and Ukrainosa from the Tertiary. In addition, the five described species of Prodinapsis from the Tertiary are included, as is Cretodinapsis from the Cretaceous. Finally, Guyotemaimetsha and Maimetsha are included to test the possible affinity of the maimetshids with the Megalyridae. The Botswanan taxa described by Rasnitsyn & Brothers (2009) are known only from compression fossils and hence could be scored for less than half of the characters in the analyses. Consequently, they were omitted. We have not expanded the sample of extant Megalyridae; that is, we score the extant genera as terminals. Shaw (1990b) provides putative apomorphies for each genus. Of the eight extant genera, three (Megalyridia, Neodinapsis, Rigel) are monotypic, and two (Cryptalyra, Ettchellsia) comprise only three species each (see Perrichot, 2009, appendix 1). The remaining three genera have 7 (Dinapsis), 6 (Carminator) and 27 (Megalyra) described species, respectively. Shaw (1990a) defined species groups within Megalyra and suggested relationships between them. Concerning Dinapsis, this genus is currently under revision (S. R. Shaw, unpublished data) and as a result of this the number of species can be expected to increase substantially.

We consider species-level phylogeny within these more diverse genera to be beyond the scope of the present study. Furthermore, obtaining material of all extant described species would have been prohibitive. All characters have been checked or rechecked for more than one species of each of the non-monotypic genera (except Cryptalyra) (see Table 1). Three of the described species of Carminator have been examined. Ten species of Megalyra representing seven out of the eight species groups defined for the genus by Shaw (1990a) were checked. With regard to Dinapsis, a number of undescribed species (currently deposited in the University of Wyoming Insect Museum) in the custody of Scott Shaw have been examined.

Examination under dissection microscope and imaging

Most characters were checked under a Leica MZ APO dissection microscope, supplemented by scanning electron microscopy (SEM, see below). Habitus and wing illustrations (Figs 5, S1,S2) were compiled from stacks of images produced with a Leica DFC 420 digital camera and integrated with Helicon Pro software.


Figure 5. Wing venation of Megalyridae: (A) Rigel chiliensis; (B) Megalyridia capensis; (C) Ettchellsia sp., (D) Megalyra fasciipennis; (E) Carminator nooni; (F) Prodinapsis oesiensis.

Download figure to PowerPoint

Scanning electron microscopy

Low-voltage SEM of uncoated specimens was performed with a Hitachi tabletop SEM unit, model TM-1000, at an operating voltage of 15 kV (University of Wyoming) or with a Jeol JSM-6335F field emission SEM unit at 2.0 kV (University of Copenhagen).

Cladistic analyses

Analyses were carried out with tnt 1.1 (Goloboff et al., 2000). The following characters were treated as additive: 4, 7, 16, 24, 37, 38, 45, 48. Space for 1 000 000 trees was reserved in memory. Traditional searches in equal-weights analyses and implied-weights analyses (Goloboff, 1993) with the concavity constant k set in turn to 1, 3, 5 and 10 were run to test the stability of clades under different weighting conditions. Analyses were run with collapsing rules set to maximum length = 0. One thousand replications with 1000 trees saved per replication were run.

To evaluate the effect of missing data, fossil taxa with a high proportion of missing entries were excluded from the analyses singly or in combination with other problematic taxa. Four of the fossil terminals have 23–52% missing or inapplicable entries for the characters scored (see Table 1): Prodinapsis pumilio, Maimetsha, Megalava, Cretodinapsis. To examine the possible effect on topologies of these problematic taxa, additional analyses were run with all these taxa excluded, as well as analyses in which the problematic taxa were excluded one at a time. A special problem is posed when the members of Maimetshidae are included (Guyotemaimetsha and Maimetsha) – see the Introduction. These were excluded in some analyses, and also excluded together with Cretodinapsis, the taxon with the most missing entries. Final analyses were run with the extant taxa only to see how the inclusion of fossils affected the topology of the extant taxa. Analyses with the full set of weighting conditions were run for each of these taxon subsets. Because of the poor resolution of the equal-weights analyses, we abstained from calculating support values.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Character list

This is an expanded version of the dataset of Shaw (1990b), which comprised 20 characters. We deleted his character 19 (metasomal shape elongate/compact) as the separation of its states break down when the fossils are included. His other characters are indicated in the list below. Most of the characters added are included to help resolve the relationships among the outgroup taxa and/or the fossil taxa not included by Shaw (1990b).

1. Antennal insertion:

(0) below or level with ventral margin of eyes (Fig. 2A–F)


Figure 2. Head and antenna of Megalyridae: (A–C) anterior view; (D–F) lateral view. (A, G) Carminator affinis; (B) Neodinapsis peckorum; (C, D) Rigel chiliensis; (E) Dinapsis sp.; (F) Megalyra wagneri.

Download figure to PowerPoint

(1) above ventral margin of eyes

State 1 is observed only in the Evanioidea and Trigonalidae included, as well as in Guyotemaimetsha.

2. Subantennal groove:

(0) absent

(1) present (Fig. 2D, E)

A well-developed subantennal groove is observed on all extant and extinct Megalyridae included here, but not in Maimetshidae; among the outgroup taxa, it is observed only in Orussus, and its presence is not a ground-plan feature of Orussidae (Vilhelmsen, 2003b). Many species of Aulacidae, especially Pristaulacus spp. (but not P. strangaliae), also have well-developed subantennal grooves (Turrisi et al., 2009), but it is not a ground-plan trait of this family either.

3. Dorsal carina of subantennal groove (Shaw, 1990b: char. 4):

(0) absent (Fig. 2C)

(1) present (Fig. 2A, B, E)

Inapplicable when the subantennal groove is absent. Shaw (1988) stated in the diagnosis of Carminator that ‘subantennal groove shallow, not bordered by carinae’; nonetheless, in Shaw (1990b)Carminator was scored as state 1. Carminator nooni appears to have a well-developed carina above the torulus, but not further posterior, and two other species described recently from Japan by Mita et al. (2007) show a distinct carina. Accordingly, this character has been scored polymorphic here (state 0 and 1).

4. Number of flagellomeres (ORDERED):

(0) more than 12 (excluding scape and pedicellus)

(1) 12

(2) fewer than 12

The antennae are broken in the holotype of Neodinapsis peckorum, and hence this taxon is scored as unknown. Twelve flagellomeres, in addition to being observed in Megalyridae, are observed in Pristaulacus (only in females, Turrisi et al., 2009) and Gasteruption.

5. Flagellomeres (Shaw, 1990b: char. 5):

(0) elongate

(1) at least some flagellomeres compact (Fig. 2G)

Compact flagellomeres constitute an apomorphy of Dinapsis and Ettchellsia, but are present in many fossil megalyrids as well.

6. Median notch on anterior margin of clypeus:

(0) absent

(1) present

The median notch is observed in Prodinapsis janzeni, P. minor, P. succinalis and Ukrainosa, as well as in Orthogonalys and Schlettererius among the outgroup taxa.

7. Mandible configuration (ORDERED):

(0) two or fewer mandibular teeth present

(1) three mandibular teeth present

(2) four or more mandibular teeth present (Fig. 2A)

State 2 is a potential apomorphy of Maimetshidae, but it is also observed in Megalava and Carminator (Shaw, 1988).

8. Mandibles:

(0) symmetric, same number of teeth on mandibles

(1) asymmetric, different number of teeth on mandibles

The presence of asymmetric mandibles is a putative apomorphy for Maimetshidae and Trigonalidae.

9. Ocellar corona:

(0) absent (Fig. 2A–C)

(1) cuticular teeth around median ocellus present

The ocellar corona consists of a curved row or circlet of discrete teeth extending posteriorly between the median and lateral ocelli (see e.g. Vilhelmsen, 2003b). It is present only in Orussus and Schlettererius among the taxa included here.

10. Median sulcus on vertex (Shaw, 1990b: char. 1):

(0) absent (Fig. 2B, D)

(1) present (Perrichot, 2009: figs 5, 8, 13)

State 0 was scored for Dinapsis in Shaw (1990), but a short crenulate sulcus is clearly visible immediately posterior to the median ocellus in at least D. oculohirta and D. hirtipes (see Similarly, a ‘longitudinal row of small foveae extending between dorsal ocelli’ was observed in Megalyra baltica by Poinar & Shaw (2007) and Perrichot (2009). Thus this character is scored as polymorphic for Dinapsis and Megalyra. The sulcus is present in most extinct Megalyridae, but absent from all extant members of the family except the Dinapsis spp. mentioned above.

11. Setae on eyes:

(0) absent (Fig. 2F)

(1) present, distinct (Fig. 2C, E)

Setae on the eyes are present in Megalyra baltica, but absent in extant Megalyra spp.; hence, this taxon has been scored as polymorphic. The occurrence of setae varies within Megalyridae.

12. Postocular carina (Shaw, 1990b: char. 2):

(0) absent (Fig. 2D, F)

(1) present (Fig. 2E)

There is considerable variation in this character within Dinapsis; some species have the carina not fully developed, whereas others have two fully developed carina present, one behind the other (Fig. 2E). However, all Dinapsis spp. appear to have the carina at least partly developed, so Dinapsis has been scored state 1 for this character.

13. Occipital carina basally (Shaw, 1990b: char. 3) (UNORDERED):

(0) curving towards back of head (Fig. 2D)

(1) branching, anterior branch curving towards mandible (Fig. 2E)

(2) effaced

State 1 is observed in all Megalyridae except Rigel and Megalyridia, which have state 0, and Carminator, which has state 2.

14. Pronotum anteromedially:

(0) well-developed, protruding

(1) reduced (Fig. 3E, F)


Figure 3. Mesosoma of Megalyridae: (A–C) dorsal view; (D–F) lateral view. (A) Megalyridia capensis; (B) Carminator affinis; (C) Dinapsis sp.; (D) Megalyra wagneri; (E) Ettchellsia sp.; (F) Rigel chiliensis.

Download figure to PowerPoint

Having the pronotum reduced anteromedially is a ground-plan feature of Megalyridae, but is also observed in Gasteruption and Pristaulacus.

15. Pronotum posteromedially:

(0) high, well-developed posteriorly of transverse sulcus

(1) short, reduced medially (Fig. 3E, F)

The posteromedially reduced pronotum is a putative apomorphy of Evaniomorpha s.s. (reversed in Evania).

16. Anterior thoracic spiracle (ORDERED):

(0) concealed or not surrounded by sclerotized cuticle

(1) exposed, with posterior part free, not surrounded by ‘pronotal’ cuticle

(2) exposed, entirely surrounded by ‘pronotal’ cuticle (Fig. 3D–F)

State 1 is observed only in Megazar and is probably present in Megalava. Apart from these reversals, state 2 can be regarded as an apomorphy of Megalyridae and Ceraphronoidea.

17. Fore tibial apical stout spines (Shaw, 1990b: char. 14):

(0) absent

(1) present (Fig. 4A)


Figure 4. Leg features of Megalyridae, lateral view. (A, B) Carminator affinis, fore tibia and hind coxa; (C) Ettchellsia sp., hind leg; (D) Rigel chiliensis, hind tibia.

Download figure to PowerPoint

These spines are present in all Megalyridae except Rigel, Megallica and two Prodinapsis spp.

18. Median mesoscutal sulcus:

(0) absent or weakly developed (Fig. 3B)

(1) present, distinct (Fig. 3A, C)

In Carminator, the sulcus is at most weakly visible (see Mita et al., 2007: 202). In the Megalyra minuta species group, the sulcus is reduced, so Megalyra has been scored polymorphic. However, the sulcus is probably present in the common ancestor of Megalyra (Shaw, 1990a), as it is present in all other Megalyridae except Carminator. The presence of this sulcus is probably a retained plesiomorphy, even if it is absent from most other Apocrita (Vilhelmsen et al., 2010).

19. Anterolateral transverse carina on mesoscutum:

(0) absent (Fig. 3B)

(1) present (Fig. 3A, C–F)

Megalyra has a ‘small acute process' (Naumann, 1987: 219) or a ‘sharp spine’ (Poinar & Shaw, 2007: 67) on the anterolateral corners of the mesoscutum. However, the terminology ‘small projection’ should be preferred (Perrichot, 2009), and it is probably equivalent to the carina observed in most other megalyrids, so Megalyra is scored as state 1. The carina is situated more medially in Dinapsis, but is considered to be homologous here, and thus Dinapsis is scored 1. The presence of the carina is a putative apomorphy of Ceraphronoidea and Megalyridae, albeit with reversals within the latter.

20. Notauli:

(0) absent or reduced (Fig. 3A–C)

(1) present

According to Mita et al. (2007), ‘scratches’ of notauli are present in Carminator helios and C. japonicus. However, these structures are more likely to be the parapsides (see the following character), so Carminator has been scored state 0. Hence, the absence of notauli is probably a megalyrid ground-plan feature.

21. Parapsides:

(0) absent

(1) present (Fig. 3A, B)

State 1 is observed only in Megazar and Megalava among the fossils. Parapsides are present in all extant Megalyridae except Megalyra. They are difficult to observe in many of the extant taxa, making it possible that they have been overlooked in some of the fossils.

22. Axillae:

(0) meeting at inner angles for at least a short distance (Fig. 3A, C)

(1) separated by the anterior portion of inner axillar grooves or by large foveae (Fig. 3B)

Carminator has been scored polymorphic for this character. Regardless, the presence of large axillae abutting in the middle is apparently a ground-plan feature of Megalyridae, probably a retained plesiomorphy.

23. Mesocoxal articulations:

(0) short, articulation at base of coxa

(1) elongate, articulation displaced distally on coxa

State 1 is an apomorphy of Evaniomorpha s.s.

24. Mid-tibial spurs (ORDERED):

(0) none

(1) one

(2) two

State 0 is so far only observed in Schlettererius (Stephanidae). State 1 is an apomorphy of Megalyridae, albeit with some fossil taxa reverting to state 2.

25. Posterior mesopleural margin (Shaw, 1990b: char. 6):

(0) straight (Fig. 3E)

(1) sinuate (Fig. 3D)

State 1 is observed in the extant taxa Carminator, Cryptalyra and Megalyra, and in the fossil taxa Megalava, Megazar and Rubes.

26. Row of foveae posterodorsally on mesopleuron:

(0) absent (Fig. 3E)

(1) row of foveae extend along posterodorsal margin of mesopleuron, adjacent to boundary with metapleuron (Fig. 3D)

According to Shaw (1990b), state 0 is an apomorphy of Ettchellsia; however, foveae are also absent in many Dinapsis spp. (see, all Dinapsis spp. illustrated there where this character can be observed have state 0). Hence, Dinapsis has been scored polymorphic for this character. The foveae are absent also in Carminator, Cryptalyra and Megazar.

27. Metapleural pilosity:

(0) metapleuron mostly bare (Fig. 3F)

(1) metapleuron with well-developed pilosity (Fig. 3D, E)

State 0 is a putative apomorphy of Megalyridae, albeit with reversals in Ettchellsia, Megalyra and Neodinapsis.

28. Longitudinal carina laterally on hind coxa (Shaw, 1990b: char. 15):

(0) absent or not visible in lateral view (Fig. 4B)

(1) present (Fig. 4C)

Some taxa (e.g. Carminator) have a carina developed at the dorsal margin of the hind coxa. This is difficult to observe in lateral view if the hind coxa is closely appressed to the metasoma. Accordingly, only taxa (Dinapsis, Ettchellsia) in which the carina is situated on the outer side of the coxa and thus clearly visibly laterally have been assigned state 1. Within Dinapsis, the extent of the carina varies considerably, but it is always present. The presence of the carina is hence an apomorphy for Dinapsis and Ettchellsia.

29. Hind coxal sculpture (Shaw, 1990b: char. 16) (UNORDERED):

(0) smooth (Fig. 4B, C)

(1) minutely shagreened

(2) rugose

This character is highly variable within Dinapsis, which has been scored polymorphic for all three states. It also displays considerable variation across Megalyridae, making it of limited phylogenetic value.

30. Comb-like spines along inner margin of hind femur and tibia:

(0) absent

(1) present (Perrichot, 2009, figs 15.3, 17.2)

State 1 is observed only in Megazar and Megalava; the spines may extend to the hind trochanter in some specimens.

31. Hind tibial setae (Shaw, 1990b: char. 17):

(0) prone (Fig. 4D)

(1) erect (Fig. 4C)

Erect setae are observed in Carminator, Cryptalyra, Dinapsis, Ettchellsia and Megalyra, as well as in Megaspilus.

32. Hind tibial spurs (Shaw, 1990b: char. 18):

(0) two

(1) one

A single hind tibial spur is observed only in Cryptalyra and Megalyra, but probably evolved independently in these two taxa.

33. Relative length, hind tarsomeres 2–4:

(0) elongate, combined length at least 1.5× of tarsomere 5

(1) short, combined length at most slightly longer than tarsomere 5 (Baltazar, 1962, fig. 1).

State 1 is a putative apomorphy for Dinapsis and Ettchellsia.

34. Tarsal pulvilli (= plantulae):

(0) absent in both sexes

(1) present, at least in females

The presence of tarsal pulvilli, a rare occurrence among Apocrita (Schulmeister, 2003), is a putative apomorphy of Maimetshidae and Trigonalidae.

35. Subapical projections on the tarsal claws:

(0) absent or weakly developed

(1) present, distinct

Subapical projections are present on the tarsal claws in Evania, Orthogonalys, Pristaulacus and Guyotemaimetsha.

36. FW, R1 distally of stigma (Shaw, 1990b: char. 8):

(0) absent (Fig. 5E)

(1) present (Fig. 5A)

The absence of this vein is an apomorphy of Carminator and Cryptalyra.

37. FW, RS apically (ORDERED):

(0) absent, at most spectral (Fig. 5B, E, F)

(1) weakly developed, nebulous or infumate (Fig. 5A, D)

(2) well developed, tubular (Fig. 5C)

Following (Shaw, 1990a), Megalyra has been scored polymorphic (states 0, 1 and 2). This character is highly variable throughout Megalyridae and difficult to interpret.

38. FW, RS configuration, when present (ORDERED):

(0) arched towards wing tip (Fig. 5D)

(1) straight (Fig. 5A)

(2) arched towards stigma (Fig. 5C)

The taxa having this vein reduced (see previous character) have been scored inapplicable for this character, making it difficult to optimize unambiguously.

39. FW, RS basally (Shaw, 1990b: char. 10):

(0) absent or reduced (Fig. 5B, D, E, F)

(1) present, branching from Rs + M (Fig. 5C)

The anterior part of this vein seems to be weakly developed in Rigel (Fig. 5A); this taxon was assigned state 0. This vein is apparently absent from the ground plan of Megalyridae, but present in Dinapsis, Ettchellsia and Neodinapsis as well as in most of the Cretaceous fossil Megalyridae and Maimetshidae.

40. FW, Rs + M position:

(0) anterior, submarginal cell less than twice as large as discal cell (Fig. 5A, C)

(1) posterior, submarginal cell at least twice as large as discal cell (Perrichot, 2009, fig. 15.2)

State 1 is observed in Cretodinapsis, Megazar and Maimetshidae.

41. FW, M + Cu:

(0) absent or spectral, colourless (Fig. 5A–C)

(1) tubular or infumate (coloured) (Fig. 5D)

Following (Shaw, 1990a), Megalyra has been scored polymorphic. Otherwise, the vein is at most weakly developed in all Megalyridae except Cretodinapsis, Megalava and Megazar, as well as in Maimetshidae.

42. FW, distal 1a:

(0) absent distally of cu-a (Fig. 5B)

(1) at least darkened line present distally of cu-a (Fig. 5D)

This character is variable across Megalyridae.

43. FW, infumate banding pattern (Shaw, 1990b: char. 11):

(0) absent, wing hyaline (Fig. 5A, E)

(1) present or wing entirely infumate (Fig. 5B–D, F)

The presence of infumate bands in the wing or infumation of the entire wing is a putative apomorphy of Dinapsis and Ettchellsia. However, among Megalyridae it is also observed in Megalyra and Megalyridia.

44. FW, pterostigma (Shaw, 1990b: char. 12):

(0) absent or reduced, at most small swelling at RS junction (Fig. 5B–D, F)

(1) thick and broad beyond Rs (Fig. 5A)

Shaw (1990b) scored the pterostigma as absent in Carminator. Because of this variation within the genus, which was the only one to be scored absent by Shaw (1990b), we have recoded this character in only two states. Given this definition, a well-developed pterostigma is present only in Rigel and Cretodinapsis, as well as in Maimetshidae and all the outgroup taxa.

45. HW, Rs length (Shaw, 1990b: char. 13) (ORDERED):

(0) absent (Shaw, 1987, fig. 5)

(1) short, barely projecting beyond R1 (Shaw & van Noort, 2009, fig. 2f)

(2) medium, reaches middle of wing (Shaw, 1987, figs 4, 6)


Figure 6. (A) Strict consensus of three trees of fit 9.46429 computed by implied-weights analyses with k = 5, only extant taxa included. (B) Strict consensus of 7399 trees of length 187 steps computed by equal-weights analyses, all taxa included. (C) Strict consensus of three trees of fit 15.38203 computed by implied-weights analyses with k = 5, all taxa included.

Download figure to PowerPoint

(3) long, approaching wing margin (Fig. 5D)

According to Shaw (1990a), most Megalyra spp. have Rs very long in the hind wing and those that do not probably have it secondarily reduced; hence, Megalyra has been scored state 3. This character is highly variable within Megalyridae, state 2 probably being the ground-plan state.

46. Propodeal sculpture (Shaw, 1990b: char. 7) (UNORDERED):

(0) smooth (Fig. 3B)

(1) areolate-rugose (Fig. 3D, F)

(2) carinate (Fig. 3E)

According to Rasnitsyn & Brothers (2009), state 1 is the condition in all Maimetshidae except Guyotemaimetsha, which is state 2. Megalyra is scored polymorphic (states 1 and 2) following Shaw (1990a). This character varies considerably across Megalyridae.

47. Metasomal insertion:

(0) posteroventral, not far removed from metacoxal foramina

(1) anterodorsal, at least halfway between propodeal antecostal sulcus and metacoxal foramina

State 1 is an apomorphy of Evanioidea.

48. Ovipositor length (Shaw, 1990b: char. 20) (ORDERED):

(0) more than 2× body length (Fig. S1a)

(1) 0.75–1.25× body length (Figs S1b; S2a, b)

(2) less than 0.75× body length (Figs S1c; S2c, d)

Dinapsis was scored state 1 according to the present definition of the states, not state 2 as in Shaw (1990b). Many Megalyridae have state 2, only Megalyra has state 0, but state 1 is probably the ground-plan condition.


Analysing only the extant taxa produces well-resolved results. Under equal weights, Megalyridae are not monophyletic, Megaspilus being nested inside. However, all the implied-weights analyses produce topologies (Fig. 6A) that are highly congruent with those obtained by Shaw (1990b), except for the placement of Megalyridia when only one tree is retrieved. When all fossils are included, the equal-weights analyses produce a largely unresolved consensus tree (Fig. 6B). The only clades retrieved are Evaniomorpha s.s. and the three terminal pairs (Carminator + Cryptalyra), (Dinapsis + Ettchellsia) and (Megalava + Megazar). The implied-weights analyses result in more resolved topologies, but they vary considerably across k-values, never retrieve Megalyridae as monophyletic, and the topology of the extant taxa is highly incongruent with the results produced when only extant taxa are included. However, the maimetshids are usually placed as the sister group to Orthogonalys, and Cretodinapsis as the sister to Evaniomorpha s.s. (Fig. 6C). When the four problematic taxa mentioned in Material and Methods are excluded, a much more resolved topology is retrieved under equal weights. However, this and the implied-weights analyses still have Megaspilus jumping inside Megalyridae. The positions of the fossil taxa vary according to analytical conditions. The Cretaceous fossils (excepting Cretodinapsis, see below) are often placed as a basal grade outside the crown group of Megalyridae. Megalava and Megazar are always retrieved as sister groups (Fig. 6C). In contrast, the Tertiary fossils are usually placed well inside the extant Megalyridae. Rubes is usually nested within Prodinapsis spp., Ukrainosa sometimes outside.

Excluding the problematic taxa one by one produces similar results to when all are included, except in the case of Maimetsha. The exclusion of this taxon produces a more resolved tree under equal weights (Fig. 7), similar to that retrieved when all four problematic taxa are excluded. Among the other problematic taxa, Megalava is always placed as the sister group of Megazar, whereas Prodinapsis pumilio is usually placed together with the other Prodinapsis spp., Rubes and Ukrainosa, albeit in slightly different positions. In contrast, the position of Cretodinapsis varies according to which taxa are included and what weighting conditions are implemented. When Maimetsha and Guyotemaimetsha are included, Cretodinapsis is often placed as sister to Evaniomorpha s.s. (Fig. 6C). When the maimetshids are excluded, Cretodinapsis at k-values 3, 5 and 10 jumps deep inside Megalyridae (see also Fig. 7), usually as sister to Carminator + Cryptalyra. Given that Cretodinapsis is missing more than half of the entries in the dataset, this is hardly surprising.


Figure 7. Strict consensus of 176 trees of length 183 steps computed by equal-weights analyses. Maimetsha excluded.

Download figure to PowerPoint

When both Maimetshidae and Cretodinapsis are excluded, k-values 3, 5 and 10 retrieve the same well-resolved topology (Fig. 8). It has Megaspilus as sister to Megalyridae, and the rest of the outgroup is resolved in the same way as suggested by Vilhelmsen et al. (2010). The extant taxa of Megalyridae are resolved in a way that is entirely congruent with Shaw (1990b) and mostly congruent with the analyses run here for the extant taxa only, except for the placement of Megalyridia. This genus and Rigel are placed as successive outgroups to the rest of Megalyridae. The next clade to branch off is Megalyra + (Carminator + Cryptalyra). The included fossil taxa are placed in two clades: (i) the Cretaceous fossils (Megallica, Megalava, Megazar and Valaa) are placed as sister to Dinapsis + Ettchellsia, Neodinapsis being the sister to the combined clade; (ii) the Tertiary fossils comprise a monophylum, both Rubes and Ukrainosa being nested inside Prodinapsis spp. This preferred tree will form the basis for further discussion of character transformations and the general evolutionary history of Megalyridae.


Figure 8. Strict consensus of three trees of fit 13.48990 computed by implied-weights analyses with k = 5. Cretodinapsis, Guyotemaimetsha and Maimetsha excluded. For character optimizations on nodes, see Table 3.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Character evolution

Megaspilus is frequently retrieved as the sister group of Megalyridae in the analyses conducted here. This result has also been produced by more comprehensive analyses of hymenopteran relationships with better representation of the Ceraphronoidea (e.g. Vilhelmsen et al., 2010). Putative synapomophies of Megalyridae and Ceraphronoidea are the presence of an anterior thoracic spiracle entirely surrounded by pronotal cuticle (char. 16:2; reversed in a few fossil megalyrids); the presence of an anterolateral transverse carina on the mesoscutum (char. 19:1; reversed in a few megalyrids); the absence of vein M + Cu in the fore-wing (char. 41:0; present in a few megalyrids); the absence of the distal part of vein cu-a in the fore-wing (char. 42:0; present in many fossil and a few extant megalyrids). Of these, the configuration of the anterior thoracic spiracle is the ‘classical’ apomorphy for these two taxa (see Gibson, 1985, 1999) and together with the presence of the anterolateral transverse carina on the mesoscutum provides the most convincing support. In contrast, the putative apomorphies in the wing venation are reductional and homoplasious within Megalyridae. In some analyses, Megaspilus is nested deeply within Megalyridae, in a clade comprising Carminator, Cryptalyra and Megalyra. Character support for this arrangement is provided by the presence of a sinuate posterior mesopleural margin (char. 25:1), the presence of erect setae on the hind-tibia (char. 31:1), and the absence of veins R1 distally of the pterostigma in the fore-wing and RS in the hind-wing (chars 36:0 and 45:0). Several of these characters display considerable homoplasy. The highly reduced wing venation of Megaspilus and all other Ceraphronoidea makes it difficult to establish a firmly supported position for them in hymenopteran phylogeny (see Vilhelmsen et al., 2010).

The monophyly of Megalyridae is weakly supported and not always retrieved (see previous section). The most convincing apomorphy of the family is the presence of well-developed subantennal grooves (char. 2:1; paralleled in Orussus). As for the remaining putative apomorphies, the anteromedially reduced pronotum (char. 14:1) is paralleled in some of the outgroup taxa; the absence of notauli on the mesoscutum (char. 20:0; paralleled in Orussus) is observed in Ceraphronidae, which were not represented in the analyses; the presence of only one mid-tibial spur (char. 24:1) is reversed in some fossil Megalyridae; and the reduction of metapleural pilosity (char. 27:0) is likewise reversed within the family. Apomorphies suggested in earlier studies, such as the presence of 12 flagellomeres in the antenna (char. 4:1) and the reduced hind-wing venation (char. 45:0/1), are not corroborated here. Vilhelmsen et al. (2010) suggested several additional possible apomorphies for Megalyridae (see Introduction), but these have yet to be confirmed for a wider taxon sample within the family.

In the preferred tree (Fig. 8, Table 3), Rigel is the first taxon to split from the rest of Megalyridae, followed by Megalyridia. All Megalyridae except Rigel are supported by the presence of stout apical spines on the fore-tibia (char. 17:1; reversed in a few fossil taxa) and the pterostigma reduced or absent (char. 44:0); all Megalyridae except Rigel and Megalyridia are supported by having the occipital carina basally branched and curving towards the mandible (char. 13:1). When under some analytical conditions the Cretaceous fossil megalyrids (Megalava, Megallica, Megazar and Valaa) are placed basal to the other members of the family, the remaining ‘crown group’ megalyrids (including Megaspilus) are supported by the presence of a carina dorsally on the subantennal groove (char. 3:1; a few reversals do occur). In this case, the topology of the extant genera of Megalyridae differs widely from that proposed by Shaw (1990b) and also frequently retrieved here. Therefore, the topology having the Cretaceous megalyrids basally is considered less corroborated than the alternative preferred here, despite its being more in accordance with the temporal succession of the terminals.

Table 3.  Unambiguous changes on the preferred tree (node numbers as indicated in Fig. 8).
Terminal/cladeCharacter changes
  1. In bold, unique change; R, reversed within clade indicated.

3: Orthogonalys4:0, 6:1, 8:1, 34:1, 48:2
4: Pristaulacus43:1, 46:2
5: Gasteruption35:0, 40:1
6: Evania4:2, 12:1, 15:0, 48:2
7: Megaspilus4:2, 7:0, 21:0, 31:1, 46:2, 48:2
8: Rigel3:0
9: Megalyridia43:1
10: Megalyra21:0, 27:1, 42:1, 43:1, 45:3, 48:0
11: Carminator7:2, 13:2, 19:0, 46:0
13: Neodinapsis27:1
15: Ettchellsia27:1, 42:1
17: Megallica5:1, 17:0, 46:1
19: Megazar26:0, 42:1
21: P. pumilio22:0
22: Rubes3:0, 25:1
24: Ukrainosa12:0, 29:0, 37:1
26: P. minor5:1
27: Evaniomorpha s. str.9:0, 13:0, 15:1R, 23:1, 25:0R, 26:1R, 48:1R
28: Evanioidea + Orthogonalys1:1, 22:1, 35:1R, 37:2
29: Evanioidea47:1
30: Evania + Gasteruption38:2, 45:0
31: Megalyridae + Megaspilus16:2R, 19:1R, 41:0R, 42:0R, 45:2R
32: Megalyridae2:1, 14:1, 20:0, 24:1R, 27:0R
33: Megalyridae (−Rigel)17:1R, 44:0
34: Megalyridae (−Rigel and Megalyridia)13:1
35: Cryptalyrini + Megalyra25:1, 31:1
36: Cryptalyrini5:1, 26:0, 36:0, 45:0
37: Dinapsini + Megazarini + Neodinapsini + Prodinapsini12:1R, 38;2
38: Dinapsini + Megazarini + Neodinapsini37:2, 39:1
39: Dinapsini + Megazarini45:1R, 46:2R
40: Dinapsini28:1, 31:1, 33:1, 43:1
41: Megazarini3:0, 12:0
42: Megallica + Megalava + Megazar19:0, 24:2
43: Megalava + Megazar30:1, 41:1
44: Prodinapsini10:1, 21:0, 22:1R, 42:1
45: P. oesiensis + P. pumilio + Rubes5:1
46: P. succinalis + P. janzeni + P. minor + Ukrainosa6:1
47: P. janzeni + P. minor17:0

The next clade to split off is Megalyra + (Carminator + Cryptalyra). This clade is supported by the presence of a sinuous posterior mesopleural margin (char. 25:1; paralleled in a few fossils) and the presence of erect setae on the hind-tibia (char. 31:1; paralleled in Dinapsis, Ettchellsia and Megaspilus). The sister-group relationship between Carminator and Cryptalyra is well supported by having compact flagellomeres (char. 5:1; paralleled in many fossils), the absence of foveae posterodorsally on the mesopleuron (char. 26:0; paralleled in Ettchellsia, some Dinapsis spp. and Megazar), and the absence of veins R1 distally of the pterostigma in the fore-wing and RS in the hind-wing (chars 36:0 and 45:0; but see above). In addition, the ovipositor in these two taxa is very short (char. 48:2), but this character is highly variable throughout the Megalyridae.

The remaining Megalyridae are placed in a clade comprising all the fossil taxa as well as the extant genera Dinapsis, Ettchellsia and Neodinapsis. The most convincing character supporting this clade is the vein RS in the fore-wing arching towards the pterostigma (char. 38:2); all the extant taxa as well as Prodinapsis and Rubes have a postocular carina (char. 12:1), but this is missing from the remaining fossil taxa. The clade can be split into two subclades: one comprises the extant taxa, as well as the Cretaceous fossil taxa (Megalava, Megallica, Megazar and Valaa); the other consists exclusively of the Tertiary megalyrid fossils (Prodinapsis, Rubes and Ukrainosa). The first subclade is supported by the presence of a well-developed fore-wing vein RS both apically and basally (chars 37:2 and 39:1). The second is corroborated by the presence of a vein 1a distally of cu-a in the fore-wing (char. 42:1). The presence of a median sulcus on the vertex (char. 10:1), the absence of parapsides (char. 21:0) and having the axillae separated medially (char. 22:1) on the mesocutum are characters more variable throughout Megalyridae and are hence less reliable.

Relationships within the clade comprising the Cretaceous fossils and the extant taxa are mostly weakly supported. The best-supported groupings are the sister-group pairs Dinapsis + Ettchellsia and Megalava + Megazar. Dinapsis and Ettchellsia share having a longitudinal carina on the hind-coxa (char. 28:1), the presence of erect setae on the hind-tibia (char. 31:1; paralleled within Megalyridae, see above), reduced hind tarsomeres 2–4 (char. 33:1) and the presence of an infumate banding pattern in the fore-wing (char. 43:1; paralleled in Megalyra and Megalyridia). Megalava and Megazar are united by having distinct bristles on the hind trochanter and femur (char. 30:1) and a coloured vein M-Cu in the fore-wing (char. 41:1), both being unique or almost unique features within Megalyridae. The nodes uniting all the Cretaceous fossils and all except Valaa are corroborated by characters that are interpreted as reversals: the absence of a dorsal subantennal groove and postocular carina (chars 3:0, 12:0; all Cretaceous taxa), and the absence of an anterolateral transverse carina on the mesoscutum (char. 19:0) and the presence of two mid-tibial spurs (24:2), (all except Valaa).

None of the relationships within the clade comprising the Tertiary fossil taxa are strongly supported. The clade consisting of Prodinapsis oesiensis, P. pumilio and Rubes all have compact flagellomeres (char. 5:1), but so has P. minor and some P. succinalis. The clade uniting the remaining Prodinapsis spp. and Ukrainosa share having a median notch in the clypeus (char. 6:1), a unique feature within Megalyridae. Prodinapsis janzeni and P. minor have no stout apical protibial spines (char. 17:0), a reversal within the family.

The fossil taxa that were excluded in the final analyses can tentatively be placed in the phylogenetic scheme for Megalyridae. When included, Maimetshidae (Guyotemaimetsha + Maimetsha) are supported by the presence of at least four mandibular teeth (char. 7:2; paralleled in Carminator and Megazar), the absence of parapsides on the mesoscutum (char. 21:0; absent also in many Megalyridae), having the submarginal cell enlarged (char. 40:1; paralleled in Cretodinapsis, Gasteruption and Megazar), and having a very short ovipositor (char. 48:2; numerous parallelisms). Placing Maimetshidae as sister to Trigonalidae is corroborated by the presence of asymmetric mandibles (char. 8:1); however, this character is not observed in all Maimetshidae (Perrichot, 2009). The presence of more than 12 flagellomeres (char. 4:0) and tarsal pulvilli (char. 34:1) might provide additional support, but could not be scored for Maimetsha. The position of Cretodinapsis as sister to all Evaniomorpha s.s. included is highly tenuous, the latter being supported only by the presence of an apically tubular vein RS in the fore-wing (char. 37:2; reversals within Evaniomorpha). Cretodinapsis could not be scored for the character traditionally cited as supporting Evaniomorpha s.s.: the presence of elongate median mesocoxal articulations (char. 23:1).

Consequences for the classification of Megalyridae

Perrichot (2009: appendix 1) revised and summarized the classification of Megalyridae. Essentially, this is an expansion of the classification of Shaw (1990b), incorporating the many new fossil taxa described in Perrichot (2009). Unlike Shaw's, the Perrichot classification was not based on a cladistic analysis. The most significant difference between the two classifications is that Perrichot (2009) excludes the maimetshids from Megalyridae entirely, recognizing them as a separate family, the Maimetshidae. This is in accordance with the original placement of Maimetsha by Rasnitsyn (1975), and it is also borne out by the results of the present analyses.

Perrichot (2009) recognized a total of 7 tribes and 16 genera in Megalyridae, all of which are represented in the present analyses. Three of the tribes listed in Perrichot (2009) are monotypic. Of the remaining tribes, only Cryptalyrini (Carminator + Cryptalyra) and Megazarini (Megalava + Megazar) are consistently retrieved and can be diagnosed in a recognizable way (see previous section). The Cretodinapsini (Cretodinapsis, Prodinapsis, Rubes, Ukrainosa) and Dinapsini (Dinapsis, Ettchellsia, Neodinapsis, Valaa) sensu Perrichot are never retrieved and appear to be entirely uncorroborated by our results. Cretodinapsis is unlikely to be a crown-group megalyrid, so we propose to exclude it from Megalyridae and for now place it as Evaniomorpha incertae sedis. As for the remaining genera, given the low robustness of our phylogenetic results there is little to be gained by adjusting the tribal classification of Megalyridae any further at present. Revising the classification would entail the recognition of even more monotypic tribes and might not be upheld by future phylogenetic treatments.

Within the clade comprising all the Tertiary fossils, Rubes and Ukrainosa render Prodinapsis paraphyletic. Although both Rubes and Ukrainosa are distinct taxa, they share a number of characters with Prodinapsis spp. (see above). Furthermore, they do not have a geographic or temporal distribution much exceeding that of Prodinapsis, so we propose to synonymize Rubes and Ukrainosa with Prodinapsis. Hence, Prodinapsis has been expanded to include Prodinapsis bruesin.comb. and Prodinapsis prolatan.comb. Revised keys to genera of Megalyridae and to Prodinapsis spp., modified from those of Perrichot (2009), are proposed accordingly (see Appendix).

Evolutionary history of Megalyridae

Shaw (1990b) suggested that Megalyridae arose in the late Early Mesozoic (i.e. late Triassic), the distribution of its extant and extinct members and their phylogenetic relationships indicating a possible correspondence between cladogenesis and major tectonic events correlated with the break-up of Pangaea. The fossil record does not corroborate this hypothesis, as the first true Megalyridae appear in the late Early Cretaceous. Jurassic deposits contain only Cleistogastridae, which might be a stem-group ‘Megalyroidea’, but this was not tested in the present study as too many characters are missing on fossils of cleistogastrids. Nonetheless, the results of the present phylogenetic treatment largely corroborate and supplement the hypothesis that megalyrids arose before the break-up of Pangaea. The distribution displayed by the extant Megalyridae indicates that they are associated with relict primary tropical and austral temperate forests. Their previous distribution should be interpreted in this context.

Several lines of evidence suggest that Megalyridae have a long and complex evolutionary history: (i) the presence of several lineages of fossils of Cretaceous and Tertiary age nested well within extant lineages (e.g. the Cretaceous crown-group megalyrids; Prodinapsis; Megalyra baltica); (ii) current wide separation of sister-group pairs among extant taxa [e.g. Carminator (Australasian-Indomalayan-eastern Palaearctic) –Cryptalyra (Neotropical); Dinapsis (Afrotropical) –Ettchellsia (Indomalayan)]; (iii) adjacent or overlapping distribution between phylogenetically well-separated extant lineages throughout the distribution range (Afrotropical: Dinapsis, Megalyridia; Neotropical: Cryptalyra, Neodinapsis, Rigel; Indomalayan: Carminator, Ettchellsia, Megalyra). This complexity makes it difficult to infer simple vicariance or dispersal scenarios for Megalyridae. However, the widespread distribution within some of the major clades (Carminator + Cryptalyra + Megalyra; Dinapsis + Ettchellsia + Neodinapsis + Cretaceous fossils) over a time span of more than 100 Myr indicates that the diversification of the genera of Megalyridae occurred as early as in the late Jurassic or earliest Cretaceous and was probably influenced by tectonic and climatic events during this time.

The warmer climate in the Late Mesozoic and early Tertiary, and the concomitant wider distribution of tropical forests in, for example, the Palaearctic (Francis & Frakes, 1993; Wilf et al., 2001, 2003; Zachos et al., 2001), is probably the main explanation for the presence of the many fossil megalyrid taxa in the latter region. There is evidence for at least two separate, now extinct, radiations of Palaearctic Megalyridae, in the Cretaceous and early Tertiary, respectively. What caused the change in faunal composition in the 50-Myr interval between these two radiations must remain conjecture for now. However, it is evident that the present-day distribution of the Megalyridae is essentially relictual. Their range contraction during the Tertiary was probably caused by the shifts to cooler climates worldwide (Zachos et al., 2001), perhaps combined with the moderate tectonic movement from lower to higher latitudes of Europe especially during the Tertiary (compare Fig. 1B and C). The climate changes thus induced had deleterious consequences for the tropical forest habitats of Megalyridae. In contrast, the genera restricted to austral temperate forests (Megalyra, Megalyridia, Neodinapsis, Rigel) have managed to survive there. The northward expansion of the range of Carminator into the Palaearctic (Japan) (Mita et al., 2007; Mita & Konishi, 2010) could be a comparatively recent (i.e. late Tertiary/Quaternary) event.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

The following persons kindly mediated access to specimens in their custody: Gavin Broad, Natural History Museum of London, U.K.; Rune Bygebjerg and Roy Danielsson, Museum of Zoology, Lund University, Sweden; Xavier Delclòs, University of Barcelona; Michael Engel, University of Kansas; Brian Fisher, Charles Griswold and Robert Zuparko, California Academy of Sciences; John Huber, Canadian National Collection of Insects, Ottawa; André Nel, National Museum of Natural History, Paris; Evgeny Perkovsky, Schmalhausen Institute of Zoology, Kiev; Mike Reich, University of Göttingen, Germany; and Dave Smith, National Museum of Natural History, Washington D.C.; Celso Azevedo, Federal University of Espirito Santo, Vitoria, Brazil, provided crucial information on and images of Cryptalyra; Simon van Noort, South African Museum, Cape Town, South Africa, provided information on Megalyridia. We thank Guinevere Z. Jones of the Department of Renewable Resources, University of Wyoming, and Dr. Zhaojie Zhang of the University of Wyoming Microscopy Core Facility, for their kind assistance with scanning electron microscopy. Pete Cranston and three anonymous referees provided useful input for the improvement of the paper. This work was supported in part by a 2009 Faculty Grant-in-Aid from the University of Wyoming, Studies of Afrotropical Insects: Megalyrid wasps of Africa and Madagascar (to S.R.S.), and by a 2007 postdoctoral research fellowship from the Alexander von Humboldt Foundation, Germany (to V.P.). This is a contribution to programs AMBRACE (no. BLAN07-1-184190) from the French National Research Agency, HymAToL (EF-1341724 and DEB-0542909) from the U. S. National Science Foundation, and the Madagascar Arthropod Biodiversity Project (NSF-DEB-00-72713).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information
  • Azevedo, C.O. & Tavares, M.T. (2006) A new species of Cryptalyra (Hymenoptera: Megalyridae) from Brazil with a key to species of the genus. Zootaxa, 1259, 5559.
  • Baltazar, C.R. (1962) Ettschellsia philippinensis sp. nov. (Dinapsinæ, Megalyridæ, Hymenoptera). The Philippine Journal of Science, 90, 219220.
  • Blakey, R.C. (2009) Global Paleogeography, Mollewide Globes, Late Early Cretaceous (105Ma) [WWW document]. URL; Eocene (50Ma):; Present: [accessed in February 2010].
  • Brues, C.T. (1923) A fossil genus of Dinapsidae from Baltic amber (Hymenoptera). Psyche, 30, 3135.
  • Brues, C.T. (1933) The Parasitic Hymenoptera of the Baltic Amber. Part 1. Bernstein-Forschungen (Amber Studies), Heft 3. Walter de Gruyter & Co., Berlin.
  • Francis, J. & Frakes, L.A. (1993) Cretaceous climates. Sedimentology Review (ed. by V.P. Wright), Vol. 1, pp. 1730. Blackwell Scientific Publications, Oxford.
  • Gibson, G.A.P. (1985) Some pro- and mesothoracic structures important for phylogenetic analysis of Hymenoptera, with a review of terms used for the structures. The Canadian Entomologist, 117, 13951443.
  • Gibson, G.A.P. (1999) Sister-group relationships of the Platygastroidea and Chalcidoidea (Hymenoptera): an alternate hypothesis to Rasnitsyn (1988). Zoologica Scripta, 28, 125138.
  • Goloboff, P.A. (1993) Estimating character weight during tree search. Cladistics, 9, 8391.
  • Goloboff, P.A., Farris, S. & Nixon, K. (2000) TNT (Tree Analysis Using New Technology) (BETA) Version. 1.1. Published by the authors, Tucumán.
  • Grimaldi, D. & Engel, M.S. (2005) Evolution of the Insects. Cambridge University Press, New York, New York.
  • He, J. (1991) A new species of the genus Ettchellsia Cameron (Hymenoptera: Ichneumonoidea, Megalyridae). Acta Entomologica Sinica, 34, 475477 (in Chinese, with English summary).
  • Hedqvist, K.-J. (1959) Hymenoptera (Ichneumonoidea) Megalyridae. South African Animal Life, 6, 485490.
  • Hedqvist, K.-J. (1967) Notes on Megalyridae [Hym. Ichneumonoidea]. Annales de la Societe Entomologique de France, 3, 239246.
  • Mita, T., Konishi, K., Terayama, M. & Yamane, S. (2007) Two new species of the genus Carminator Shaw from Japan, the northernmost record of extant Megalyridae. Entomological Science, 10, 201208.
  • Mita, T. & Konishi, K. (2010) Phylogeny and biogeography of Carminator (Hymenoptera: Megalyridae). Systematic Entomology, in press.
  • Naumann, I.D. (1987) A new megalyrid (Hymenoptera: Megalyridae) parasitic on a sphecid wasp in Australia. Journal of the Australian Entomological Society, 26, 215222.
  • van Noort, S. & Shaw, S.R. (2009) Megalyridia capensis (Hymenoptera: Megalyridae: Megalyridiini), a relict species endemic to South Africa. African Natural History, 5, 18.
  • Perrichot, V. (2009) Long-tailed wasps (Hymenoptera: Megalyridae) from Cretaceous and Paleogene European amber. Paleontological Contributions, 1, 135.
  • Perrichot, V., Nel, A. & Neraudeau, D. (2004) A new, enigmatic, evaniomorphan wasp from the Albian amber of France (Insecta: Hymenoptera). Journal of Systematic Palaeontology, 2, 159162.
  • Poinar, G. & Shaw, S.C. (2007) Megalyra baltica Poinar and Shaw n. sp. (Hymenoptera: Megalyridae), a long-tailed wasp from Baltic amber. Zootaxa, 1478, 6568.
  • Rasnitsyn, A.P. (1975) Vysshie pereponchatokrylye mezozoya [Hymeoptera Apocrita of the Mesozoic]. Trudy Paleontologicheskogo Instituta Akademii Nauk SSSR, 147, 1132 (in Russian).
  • Rasnitsyn, A.P. (1977) New Hymenoptera from the Jurassic and Cretaceous of Asia. Paleontological Journal, 11, 349357 [English translation from Russian of Paleontologiceskij Zurnal, 3 (1977), 98–108].
  • Rasnitsyn, A.P. (1980) Origin and evolution of Hymenoptera. Transactions of the Institute of Paleontology, 174, 1192 (in Russian; English translation 1984, Biosystematics Research Centre, Ottawa).
  • Rasnitsyn, A.P. (1988) An outline of the evolution of the hymenopterous Insects (Order Vespida). Oriental Insects, 22, 115145.
  • Rasnitsyn, A.P. (2002) Superorder Vespidea Laicharting, 1781. Order Hymenoptera Linné, 1758 (Vespida Laicharting, 1781). History of Insects (ed. by A.P. Rasnitsyn and D.L.J. Quicke), pp. 242254. Kluwer Academic Publishers, Dordrecht.
  • Rasnitsyn, A.P. & Brothers, D.J. (2009) New genera and species of Maimetshidae (Hymenoptera: Stephanoidea s.l.) from the Turonian of Botswana, with comments on the status of the family. African Invertebrates, 50, 191204.
  • Rodd, N.W. (1951) Some observations on the biology of Stephanidae and Megalyridae (Hymenoptera). Australian Zoologist, 11, 341346.
  • Schulmeister, S. (2003) Morphology and evolution of tarsal plantulae in Hymenoptera (Insecta), focussing on the basal lineages. Zoologica Scripta, 32, 153172.
  • Shaw, S.R. (1987) Three new megalyrids from South America (Hymenoptera: Megalyridae). Psyche, 94, 189199.
  • Shaw, S.R. (1988) Carminator, a new genus of Megalyridae (Hymenoptera) from the Oriental and Australian regions, with a commentary of the definition of the family. Systematic Entomology, 13, 101113.
  • Shaw, S.R. (1990a) A taxonomic revision of the long-tailed wasps of the genus Megalyra Westwood (Hymenoptera: Megalyridae). Invertebrate Taxonomy, 3, 10051052.
  • Shaw, S.R. (1990b) Phylogeny and biogeography of the parasitoid wasp family Megalyridae (Hymenoptera). Journal of Biogeography, 17, 569581.
  • Shaw, S.R. (2003) A new species of Cryptalyra from Colombia (Hymenoptera: Megalyridae). Zootaxa, 248, 14.
  • Shaw, S.R. & van Noort, S. (2009) A new Dinapsis species from the Central African Republic (Hymenoptera, Megalyridae, Dinapsini). Zootaxa, 2118, 3036.
  • Turrisi, G.F., Jennings, J.T. & Vilhelmsen, L. (2009) Phylogeny and generic concepts in the parasitoid wasp family Aulacidae (Hymenoptera: Evanioidea). Invertebrate Systematics, 23, 2759.
  • Vilhelmsen, L. (2003a) Flexible ovipositor sheaths in Hymenoptera (Insecta). Arthropod Structure and Development, 32, 277287.
  • Vilhelmsen, L. (2003b) Phylogeny and classification of the Orussidae (Insecta: Hymenoptera), a basal parasitic wasp taxon. Zoological Journal of the Linnean Society, 139, 337418.
  • Vilhelmsen, L., Miko, I. & Krogmann, L. (2010) Beyond the wasp-waist: structural diversity and phylogenetic significance of the mesosoma in apocritan wasps (Insecta: Hymenoptera). Zoological Journal of the Linnean Society, 159, 22194.
  • Wilf, P., Labandeira, C.C., Johnson, K.R., Coley, P.D. & Cutter, A.D. (2001) Insect herbivory, plant defense, and early Cenozoic climate change. Proceedings of the National Academy of Sciences of the United States of America, 98, 62216226.
  • Wilf, P., Johnson, K.R. & Huber, B.T. (2003) Correlated terrestrial and marine evidence for global climate changes before mass extinction at the Cretaceous–Paleogene boundary. Proceedings of the National Academy of Sciences of the United States of America, 100, 599604.
  • Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K. (2001) Trends, rhythms, and aberrations in global climate 65 Ma to present. Science, 292, 686693.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Appendix Revised keys to genera of Megalyridae and Prodinapsis spp.

Key to genera of Megalyridae
  • 1. Fore-wing with vein Rs between Rs+M and r-rs tubular for at least a short distance, apical segment of Rs tubular, arched towards stigma (Fig. 5C).....................2

  • – Fore-wing with vein Rs between Rs + M and r-rs absent, apical segment of Rs absent or spectral (Fig. 5A–B, D–F) ........................ 8

  • 2. Anterior thoracic spiracle posteriorly not surrounded by pronotal cuticle (Perrichot, 2009: fig. 16.3); parapsides present (Perrichot, 2009: figs 16.2, 17.1); metafemur with comb-like spines along inner margin (Perrichot, 2009: figs 15.3, 16.4, 17.2); fore-wing with veins M + Cu and distal segments of Cu present, tubular (Perrichot, 2009: figs 15.2, 17.1)................3

  • – Anterior thoracic spiracle entirely surrounded by pronotal cuticle (Fig. 3D–F); parapsides absent or effaced; metafemur without comb-like spines; fore-wing with veins M + Cu and distal segments of Cu absent or at most spectral (Fig. 5A–C) ........................................... 4

  • 3. Vertex without longitudinal median sulcus; mandibles with four teeth; occipital carina, median mesoscutal sulcus and inner axillar grooves smooth (Perrichot, 2009: fig. 16.2); fore-wing with marginal cell narrow, large submarginal cell entirely closed by tubular veins, and a pre-basal segment Cu1 forming a small pentagonal medial cell (Perrichot, 2009: fig. 15.2)...................................... †Megazar Perrichot (Fig. S3a)

  • – Vertex with a distinct longitudinal median sulcus; occipital carina foveate, median mesoscutal sulcus smooth, and inner axillar grooves crenulate; fore-wing with a small rectangular medial cell, veins Cu and M tubular (Perrichot, 2009: fig. 17)............................ †Megalava Perrichot (Fig. S3b)

  • 4. Vertex without longitudinal median sulcus...............................5

  • – Vertex with a distinct longitudinal median sulcus.................................†Megallica Perrichot (Fig. S3c)

  • 3. Posterior ocular orbits with groove and/or carina present (Fig. 2E)........................................................ 6

  • – Posterior ocular orbits without groove or carina...............................†Valaa Perrichot (Fig. S3d)

  • 6. Posterior orbital groove foveate (Fig. 2E); fore-wing veins 1m-cu and Cu1 absent..................................................................... 7

  • – Posterior orbital groove not foveate; fore-wing veins 1m-cu and Cu1 present (Fig. 5C; Baltazar, 1962: fig. 1)....................................................Ettchellsia Cameron (Fig. S1c)

  • (1) Fore-wing vein A posterior of 1cu-a present at least as a darkened line; metacoxa with a longitudinal carina; metasoma compact (Shaw & van Noort, 2009: fig. 2) .......................................... Dinapsis Waterston (Fig. S1b)

  • – Fore-wing vein A posterior of 1cu-a absent; metacoxa without longitudinal carina; metasoma elongate (Shaw, 1987: figs 3, 6)..............Neodinapsis Shaw (Fig. S2d)

  • 8. Vertex with longitudinal median sulcus distinct or faintly impressed; posterior ocular orbits with groove reduced or absent and carina present, sharp or reduced (Perrichot, 2009: figs 3.5, 6.2, 11.3); ............ †Prodinapsis Brues (Fig. S4a–g)

  • – Vertex without longitudinal median sulcus; posterior ocular orbits without groove, with carina absent or effaced............ 9

  • 9. Dorsal carina of subantennal groove present (Fig. 2A, B, E).................................................................... 10

  • – Dorsal carina of subantennal groove absent (Fig. 2C).........................................................Rigel Shaw (Fig. S2a)

  • 10. Flagellomeres compact (Fig. 2G); fore-wing R1 absent; hind-wing Rs absent (Fig. 5E)....................................................... 11

  • – Flagellomeres elongate; fore-wing R1 present; hind-wing Rs present (Fig. 5B, D).................................................................. 12

  • (2) Pterostigma absent (Fig. 5E); metacoxa shagreened; metatibia with two apical spurs (Shaw, 1988: fig. 1)........................Carminator Shaw (Fig. S2c)

  • – Pterostigma present, only swelling at junction with Rs; metacoxa rugose; metatibia with only one apical spur (Shaw, 1987: figs 2, 5)............................... Cryptalyra Shaw

  • 12. Head and mesosoma shagreened (Fig. 3A; van Noort & Shaw, 2009: figs 1D, 2F, 3E); pronotal spiracle without internal fringe of setae; metacoxa shagreened; metatibia with two apical spurs.................Megalyridia Hedqvist (Fig. S2b)

  • – Head and mesosoma coarsely foveate-reticulate (Figs 2F, 3D); pronotal spiracle with an internal fringe of setae; metacoxa rugose; metatibia with only one apical spur.............................. Megalyra Westwood (Fig. S1a)

Key to species of Prodinapsis

  • 1. Female.................................................................. 2

  • – Male.....................................................................6

  • 2. Antennae thickened at apex, flagellomeres compact (Fig. S4f, g); fore-wing with apical segment of Rs absent (Perrichot, 2009: figs 1.2–1.5, 1.7); body at most 2.30 mm in length (excluding ovipositor)...............3

  • – Antennae filiform, flagellomeres elongate (Fig. S4a, d); fore-wing with apical segment of Rs spectral (Perrichot, 2009: figs 1.1, 1.6); body 2.30 to 4.00 mm in length (excluding ovipositor) .................................. 5

  • 3. Posterior orbital carina small (Perrichot, 2009: fig. 6.2); fore-wing with basal segment of Cu1 present, spectral (Perrichot, 2009: fig. 1.2); body 1.40–2.30 mm in length..................... 4

  • – Posterior orbital carina very weak (Perrichot, 2009: fig. 6.6); fore-wing with basal segment of Cu1 absent (Perrichot, 2009: fig. 1.3); body at most 1.40 mm in length....................................P. pumilio Perrichot (Fig. S4g)

  • 4. Head transversely flat behind eyes; eyes very large, fully covering head length; fore-wing with a prebasal segment Cu1 forming a pentagonal medial cell (Perrichot, 2009: fig. 1.7)............ P. bruesi (Perrichot) n.comb. (SI-fig. IVf)

  • – Head obliquely narrowed behind eyes; eyes not fully covering head length; fore-wing without prebasal segment Cu1, medial cell quadrangular (Perrichot, 2009: fig. 1.2).......................................... P. minor Brues

  • 5. Posterior orbits without groove or carina (Perrichot, 2009: fig. 11.6); fore-wing with basal segment of Cu1 present, spectral; ovipositor as long as body......................... P. prolata (Perrichot & Perkovsky) n.comb. (Fig. S4e)

  • – Posterior orbits with a distinct groove and a sharp carina (Perrichot, 2009: figs 3.5, 4.4); fore-wing vein Cu spectral to nebulous; ovipositor distinctly shorter than body........................... P. succinalis Brues (Fig. S4a)

  • 6. Antennae not extending beyond apex of metasoma; fore-wing with apical segment of Cu1 present, spectral.................7

  • – Antennae extending beyond apex of metasoma; fore-wing with apical segment of Cu1 absent.........................................P. janzeni Perrichot (Fig. S4d)

  • 7. Median mesoscutal sulcus and axillar grooves crenulate..................................... 8

  • – Median mesoscutal sulcus and axillar grooves smooth (Perrichot, 2009: figs 10, 11.4)................................ P. oesiensis Perrichot (Fig. S4c)

  • 8. Fore-wing length low relative to head length (3,5 < FWL/ HL < 5) ........................................... P. succinalis Brues

  • – Fore-wing length high relative to head length (FWL/HL > 5) P. minor Brues (Fig. S4b)

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendices
  10. Supporting Information

Supporting Information

Additional Supporting Information may be found in the online version of this article under the DOI reference: DOI: 10.1111/j.1365-3113.2010.00537.x

Figure S1. Habitus images of Megalyridae. (a) Megalyra lilliputiana Turner, 1916, (b) Dinapsis sp., (c) Ettchellsia sp.

Figure S2. Habitus images of extant Megalyridae. (a) Rigel chiliensis Shaw, 1987 [holotype], (b) Megalyridia capensis Hedqvist, 1959 [holotype], (c) Carminator affinis Shaw, 1988, (d) Neodinapsis peckorum Shaw, 1987 [holotype].

Figure S3. Habitus images of Cretaceous Megalyridae. (a) Megazar elegans Perrichot, 2009 [holotype female], (b) Megalava truncata Perrichot, 2009 [holotype], (c) Valaa delclosi Perrichot, 2009 [holotype male], (d) Megallica parva Perrichot, 2009 [holotype male].

Figure S4. Habitus images of Cenozoic Megalyridae. (a) Prodinapsis succinalis Brues, 1923 [female], (b) P. minor Brues, 1933 [male], (c) P. oesiensis Perrichot, 2009 [holotype male], (d) P. janzeni (Perrichot, 2009) [holotype male], (e) P. prolata (Perrichot & Perkovsky, 2009) [holotype female], (f) P. bruesi (Perrichot, 2009) [holotype female], (g) P. pumilio Perrichot, 2009 [holotype female], (h) Megalyra baltica Poinar & Shaw, 2007 [female].

Please note: Neither the Editors nor Wiley-Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

SYEN_537_sm_FigureS1-4.pdf40551KSupporting info item

Please note: Neither the Editors nor Wiley Blackwell are responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.