Butterflies did not mistakenly mate with worm-like animals to give rise to caterpillars, according to a new report that challenges previous claims to the contrary.
“It’s a nutbar idea,” says study author Michael Hart, an evolutionary biologist at Simon Fraser University in Burnaby, Canada. “There’s just nothing to it.”
In August, the Proceedings of the National Academy (PNAS) published a paper online by Donald Williamson, a retired zoologist at the University of Liverpool, UK, arguing that at some point in the distant past a larva-less insect hybridized with a velvet worm, and the resulting descendants now develop successively through stages that resemble both parents. The study — which was ‘communicated’ by academy member Lynn Margulis, of the University of Massachusetts, Amherst, via the soon-to-be-obsolete ‘Track I’ submission route, which allows academy members to handle the review of a colleague’s manuscript — became embroiled in controversy after questions were raised about the peer review process (see ‘Row at US journal widens’).
The paper’s print publication was suspended for more than a month, but it was ultimately given the green light in mid-October by the PNAS editorial board. The journal also plans to publish a short ‘letter to the editor’ by Gonzalo Giribet, an invertebrate zoologist at Harvard University in Cambridge, Massachusetts, and a response from Williamson. In addition, PNAS today published a full-length counter-argument to Williamson’s hypothesis.
Williamson’s paper offered little direct evidence for hybridization other than physical resemblances between caterpillars and velvet worms. But he predicted that insects with caterpillar larvae should show genetic similarities to the worm-like invertebrates, and he called on genomicists to test his ideas. Now, two non-academy members — Hart and fellow evolutionary biologist Richard Grosberg, of the University of California, Davis — have taken Williamson to task and turned to published data on genome sizes and contents to rubbish his proposals.
To prepare the new paper — a direct, ‘Track II’ submission for which the PNAS editorial board assigns an academy member expert to organize the review process — Hart and Grosberg spent “a few hours trolling the web for data and observations that were directly relevant to [Williamson’s] specific predictions”, says Hart. First, they amassed genome size data based on the total amount of DNA, which is measured by weight in picograms using cytological methods. They showed that insects that lack caterpillar-like larvae have relatively larger genome sizes than insects that have the grub-like stage in their life cycle. (Williamson predicted the opposite trend.)
The authors also note that caterpillar-bearing bugs have fewer genes than other arthropods, and that genome sequence comparisons do not indicate a close evolutionary relationship between velvet worms and metamorphic insects.
Williamson was unconvinced by these arguments. “The results [from DNA weight measurements] seem to bear little or no relationship to the number of genes in base-pairs,” he says. “I hope some geneticist will point out the difficulty of interpreting the picogram figures.” He notes, for example, that a mountain grasshopper has nearly 200-fold more DNA than a hessian fly as measured by weight but it does not have that many more functional genes.
Hart says that Williamson’s defense is hypocritical. He points out that Williamson himself relied on genome weight data as evidence of larval transfer in his paper. “How can [Williamson] reasonably object to our use of the same type of data in response?” asks Hart.
Williamson also discounts the gene number comparisons because all the insect species with sequenced genomes are metamorphic, and there are no comparable sequences for non-metamorphic insects or velvet worms.
Hart is unsurprised that Williamson is sticking to his guns. “I expect he’ll continue to advocate for the idea,” he says. “I don’t suppose this will be the end of it.”
Image: Velvet worm via Wikimedia Commons