Ants in Bright Velvet

A paper that I've been intermittently working on for a while now finally saw publication last week. Authored by myself, Mark Murphy, Yvette Hitchen and Denis Brothers, the paper describes four new species of velvet ant from here in Western Australia.

Female Aglaotilla chalcea, photographed by yours truly.

Velvet ants are not actually ants but a distinct group of typically hairy wasps forming the family Mutillidae. They are strongly sexually dimorphic: females are wingless like ants but males have fully developed wings. They develop as kleptoparasites in the nests of other wasps, with the velvet ant larva feeding on the prey left to provision the host and/or on the host larva itself. Taxonomically, velvet ants are perhaps one of the more difficult wasp groups to work with. The high sexual dimorphism means that it is often impossible to match males with females unless one is lucky enough to catch them in the act of mating, and the mesosoma of females is highly sclerotised and fused with many of the characters useful for identifying other wasp groups no longer visible. The taxonomy of Australian velvet ants is particularly uncertain, almost to comical levels. A large number of species (possibly numbering in the hundreds) remain undescribed, and many of those species that have been described are yet not readily identifiable. No extensive survey of the Australian fauna has appeared since 1898 and most Australian species have been placed in a single genus Ephutomorpha. This genus was established by French entomologist Ernest André in 1902 with a definition that can basically be summarised as "Ugh, I can't even right now": it was explicitly intended as a dumping ground for Australian velvet ants that André was unable to sort more appropriately at the time. A vague promise to get onto it later never eventuated. Even at its time of establishment, Ephutomorpha included taxa that had already been designated as type species for genus names Bothriomutilla and Eurymutilla that should have taken precedence.

A few years ago, I was engaged in identifying wasp specimens collected by Mark Murphy as part of his research into pollinator ecology in the Western Australian wheatbelt. For those of you unfamiliar with the area, the Wheatbelt refers to a band of land inland from Perth. Most of the wheatbelt is rolling, semi-arid terrain that has been cleared for the growth of the eponymous wheat, with the indigenous forest largely reduced to isolated stands and reserves. Mark was studying the diversity of pollinator wasps in these remnant stands, most of which are dominated by wandoo Eucalyptus wandoo. As an example of the difficulties I was referring to above, I was able to recognise over two dozen morphospecies of velvet ants among specimens collected by Mark, only a couple of which I was able to even tentatively connect to known species. The specimens which formed the basis of the new publication came from a particular one of Mark's study methods, nest traps. Mark would leave wooden blocks into which holes had been drilled out in the field for a number of months, over which time they would hopefully be colonised by nesting wasps and bees (Mark was visiting traps once a month to check for nests). The holes were lined with paper tubes and if Mark found one that contained a nest, he would slide out the tube and take it back to the lab to be reared to maturity. Emerging wasps and bees were identified to species both by morphological examination and via the extraction of DNA for fingerprinting. Mark also found that he reared a number of parasitoids and kleptoparasites that were treated in the same way.

The male of Aglaotilla chalcea, also by yours truly.

I realised that this gave us an excellent opportunity regarding the mutillids, of which four identifiable species had emerged from Mark's nest samples. Because of Mark's rearing experiments, we had host data for all four species. Because of the use of DNA fingerprinting, we were able to identify both males and females of three of the four species (the fourth was recorded from a single nest that only provided us with female specimens). And at least two of the species appeared to be completely new to science. It didn't hurt that they were also all very attractive animals with brilliant metallic colours. So I prepared a manuscript describing all four species with myself, Mark and Yvette (who had done the DNA sequencing for the specimens) as authors and submitted it to the journal Zootaxa for consideration.

It was rejected.

That, as it turned out, was a good thing. One of the original reviewers was Denis Brothers of the University of KwaZulu-Natal, one of the world's leading authorities on velvet ants. Denis agreed that, while the paper couldn't stand as originally submitted, there was a definite value in what we were presenting. So he offered to help us with the composition. As well as correcting some misunderstandings I was guilty of regarding mutillid morphology (see my earlier comment on the difficulty of identifying features of the female mesosoma), Denis was able to confirm that all four of our species was actually new. He also informed us that they could be placed in a group of species that he had identified as part of as-yet unpublished research on Australian velvet ants and suggested that we establish a new genus for this group. This new genus was named Aglaotilla by Brothers (2018). Denis also added a new section to our manuscript summarising the recorded host data for Australian mutillids.

Aglaotilla species are mostly metallic in coloration, predominantly blue, green or purple (describing the colours of metallic wasps can be a challenge because the exact shade observed depends a lot on the incident lighting). One of our species, A. micra, has the mesosoma reddish with a purple gloss whereas an earlier described species A. discolor has the mesosoma entirely red. Females often have prominent spots or bands of clustered white hairs on the metasoma. Depending on the species, the colour pattern of the sexes may be similar or distinct. One of our new species, A. lathronymphos, has a species name that means 'secretly married' because without the DNA fingerprinting we would have had no reason to associate the bright blue males with the reddish-purple females. Females lack the rake-like spines on the fore legs and flattened plate at the end of the metasoma found in many other female mutillids. This almost certainly relates to their life cycle. Female velvet ants parasitising ground-nesting hosts use their fore legs to dig into the host nest and the terminal plate to tap down the ground after closing it back up. Aglaotilla females, where known, parasitise hosts that nest above ground in holes in trees and so do not need adaptations for digging. Three of the species we described, A. chalcea, A. lathronymphos and A. micra, were reared from the nests of crabronid wasps belonging to the genus Pison. The fourth species, A. schadophaga, was reared from the nests of resin bees. Aglaotilla species are very unusual among velvet ants in that more than one larva may grow to maturity in a single host nest cell; in all other mutillids for which host data is available, only a single individual will ever emerge from a single host.

A likely live female of Aglaotilla in search of a suitable host nest, copyright Mark A. Newton.

The Australian mutillid fauna includes a number of enticing taxa that deserve further examination: the strikingly patterned Australotilla species and the weird ant-associated Ponerotilla are just a couple of examples. Not to mention the hordes of new species that don't even have names yet. I have been pleased to make some contribution to this much-neglected family.

REFERENCES

André, E. 1902 Hymenoptera. Fam. Mutillidae. Genera Insectorum 11: 1–77, 3 pls.

Brothers, D. J. 2018. Aglaotilla, a new genus of Australian Mutillidae (Hymenoptera) with metallic coloration. Zootaxa 4415 (2): 357–368.

Taylor, C. K., M. V. Murphy, Y. Hitchen & D. J. Brothers. 2019. Four new species of Australian velvet ants (Hymenoptera: Mutillidae, Aglaotilla) reared from bee and wasp nests, with a review of Australian mutillid host records. Zootaxa 4609 (2): 201–224.