Time to pull the plug on Halteria

Halteria is the name for a group of insects defined by a severe modification of flight wings into gyroscopic stabilizing structures called halteres. The group includes all the true flies plus Strepsiptera, an oddball lineage of parasites. This taxonomic scheme is a pleasing arrangement by some counts, as it requires but a single evolutionary origin of a complex structure.

Halteria was named 15 years ago when the first DNA sequences were applied to reconstruct insect evolution. A 1994 paper by Michael Whiting & Ward Wheeler discovered an unexpected relationship of flies to Strepsiptera using a fragment of ribosomal DNA. Previously, Strepsiptera had been considered by morphologists as relatives of beetles.

The new find was exciting. It not only suggested that molecular data would yield novel evolutionary relationships, but also that extensive body rearrangements might be possible via developmental mutations. The halteres of Strepsiptera are fashioned from forewings, while those of the flies from hindwings, implicating a rare segment swap.

There’s one problem with Halteria, though. Beyond the original papers, there’s no evidence it exists.

First came a study by John Huelsenbeck suggesting Halteria might be a statistical artifact. Then came the observation that the fine structure of the appendages did not match, followed by a finding that Strepsiptera lacked the expected sort of hormonal receptor for a fly relative. Last year saw an avalanche of molecular phylogenetic studies (here, here and here), each with considerably more data than the early genetic work. These unanimously concluded that the morphologists- as crusty and out-of-date as they may be- had been correct. Strepsiptera are related to beetles, not flies.

I bring all this up because a new paper by Ishiwata et al adds yet more genetic data in favor of the old Beetle-Strepsiptera connection:

It’s only 3 genes, but they are big ones, and the tree is eminently sensible. The Neoptera (insects that can fold their wings over their backs) and Holometabola (insects with complete metamorphosis) are both supported, and the overall tree topology agrees with that of the larger genomic studies. And Strepsiptera do not group with the flies in any scenario.

I’m not sure how many nails the Halteria coffin needs. But surely that lid is now shut so tight that not even flies can get in.


source: Ishiwata, K. et al 2010. Phylogenetic relationships among insect orders based on three nuclear protein-coding gene sequences. Molecular Phylogenetics and Evolution, online early.

7 thoughts on “Time to pull the plug on Halteria”

  1. I think it might be interesting if you posted some older morphology based phylogenies side by side with the above one. This scheme resembles, in most essential respects, what I remember of the cladogram generated a long time ago by Willi Hennig, with the most significant differences in the orthopteroids (but I’m just going on an old memory).

  2. We will see.

    The morphology based systemitists spent centuries of iterations refining their conclusions. I would expect the genetics based systemitists to go thru similar iterations as well.

  3. Correct me if I am wrong, but as far as I can remember the Strepsiptera+Coleoptera clade was little more than a hunch when suggested by the morphologists. The two groups do not really have any unambiguous synapomorphies. There is, of course, hindwing-powered flight and a hardening of the fore wings, but neither count as conclusive evidence in their own right. Probably the most important initial arguments linking the two groups were the superficial (but almost certainly non-homologous) similarities with the beetle families Meloidae and Ripipiphoridae in terms of larval development. Later, putative synapomorphies were discovered in the wing venation, but these were also found to be ambiguous.
    To me, the most convincing arguments for the Streps+Coleoptera grouping come from molecular data. It is not accurate to state that the morphological tradition has been right all along; although it does give the Streps a correct placement topology-wise, this is based on poor data.

Leave a Reply