Did two species mix to make butterflies?

27 August 2009 by Bob Holmes

WHAT child's imagination has not been captivated by the near-magical transformation that caterpillars undergo to become butterflies? This is the result of an ancient hybridisation between an insect and a worm-like animal, according to zoologist Donald Williamson, and now he says there is enough genetic information to test the theory.

Unfortunately for Williamson, now retired from the University of Liverpool, UK, the early returns are not encouraging.

Many insect groups, such as butterflies, bees and wasps, have larval stages that look nothing like the adults. Most biologists believe these evolved gradually, perhaps because natural selection favoured juvenile stages that differed from the adults and thus would compete less with them.

Williamson offers a different explanation. At some point hundreds of millions of years ago a larva-less insect - something like a grasshopper or cockroach, say - hybridised with a velvet worm. Also known as Onychophora, velvet worms are worm-like invertebrates with stubby, leg-like appendages. According to Williamson's theory, the resulting hybrid and its descendants now develop successively through stages resembling both parents.

"Nobody knows where caterpillars came from," says Williamson, who thinks that many other invertebrate groups acquired their larvae in the same way (New Scientist, 24 January, p 34). "It's the only solution that makes sense."

Williamson offers little evidence other than physical resemblances, but he predicts that insects with caterpillar larvae should show genetic similarities to velvet worms (Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.0908357106).

Tommyrot, other biologists snort. For a start, the resemblance between velvet worms and caterpillars is only superficial. "In my view, Onychophora and caterpillars do not resemble each other at all," says Georg Mayer, a specialist in velvet worm development and taxonomy at the University of Jena, Germany.

Moreover, the genetic affinities Williamson predicts do not seem to be there. Geneticists have sequenced the genomes of several insects with caterpillar larvae, including silkworms, fruit flies, honeybees and mosquitoes. Yet there is no indication that any of their genes differ from what would be expected in a typical insect.

"I think it would be fairly obvious if there was lots of non-insect stuff in there," says Max Telford, a zoologist at University College London. For instance, all animals use the Hox family of genes to help them develop a front and a back. If caterpillars and adults derived from two separate ancestors, they would need two sets of Hox genes to guide their distinct developmental processes, but this is not the case.

In fact, caterpillars and their adults do seem to have genetic links. In beetles, for example, the genes that control larval leg development are the same ones that guide leg development in adults, says Nipam Patel, an evolutionary biologist at the University of California, Berkeley.

As appealing to the imagination as Williamson's theory may be, it looks like the evidence is not there to support it.

http://www.newscientist.com/article/mg20327234.900-did-two-species-mix-to-make-butterflies.html?DCMP=OTC-rss&nsref=online-news