Mousy blondes: Ready for evolution textbooks?

This paper on mice evolving a new coat colour has been making a big splash in science news. It’s being touted as a new textbook example of evolution. Actually, not just an example, but an “icon.” I’m not sure what to think about that, given that Icons of Evolution is a notorious creationist book. Plus, the last time someone was touting “it’ll be in all the textbooks” were the promoters of the breathlessly over-hyped Darwinius / Ida fossil.

Reading the technical paper is very frustrating. I hate to say, but I don’t think the story is as complete or as impressive – yet – as the press releases indicate.

The press release version of this story is that in Sandy Hills, Nebraska, a region that has very light coloured soil, local deer mice have evolved a light coloured fur coat in a relatively short geological period of time. (In the picture, light deer mice are shown on dark soil, not their matching soil; same for the dark mice: they are shown on light soil.) In fact, the light coat was not present in the original population, and only evolved after the region emerged.

That is an interesting story. On what basis do Linnen and collegues make these claims?

First, they measured and compared the coats of five mice from two locations. Based on these measurements, they show the Sandy Hills mice are more reflective across the light spectrum, and this appear to be due to a pigment called pheomelanin.

In lab mice, a well known gene called Agouti affects coat colour. By breeding laboratory deer mice with mutations related to this gene, they showed these differing coat colours are classic Mendelian genes, where light is dominant over dark. Two dark mice will breed true; two light coloured mice will either breed true, but could have a mix of light and dark offspring (about three light offspring for every one dark offspring).

They also measured the expression of the Agouti gene during development, by tracking mRNA levels for the gene. As expected, they found Agouti more heavily expressed in light deer mice than dark.

They then captured deer mice in the wild, near the edge of the Sand Hills, to see if an Agouti mutation was associated with coat colour. As predicted, it was. In doing so, they found no relationship between where they collected the deer mice and their coat colour. From this, they argue, mice of both colours are interbreeding. I would like to see actual behavioural mate choice tests to support this.

The argument that selection has occurred revolves around the variation of the genome sequence. It’s fairly technical and I am not going to pretend that I fully understand the logic here. But I’ll spot that it’s all correct, and I would still argue that this is the weakest part of the paper. The molecular data are an indirect inference suggesting selection for colour. The field experiments I’d like to see demonstrating colour advantage seem to have been done, but they’re in a fairly obscure journal from the 1940s. Anyone have copies of Contributions from the Laboratory of Vertebrate Biology of the University of Michigan handy?

Similarly, there are no archaeological pelts that they base their arguments for a recent evolution of light coat colour on. It’s all mathematical models that assume population size, relate strength of selection and variability. To be clear, I’m not coming down on using models, just pointing out what kind of evidence is being used to make the case. To give an example, they note Sand Hills is about 8,000-10,000 years old, but they don’t give an approximate age for the origin of the Agouti mutation in years. 8,000-10,00 years is ~0.4 to 0.5 4N generations of deer mice, and they estimate the Agouti mutation arose 0.05 to 0.18 4N generations ago. And I am not going to pretend I understand what those figures mean or how they calculated them.

I could be badly biased here, because I am an organismal biologist, and this is mostly a molecular biology paper. But if I’m having problems understanding the details, I wouldn’t unleash this as a case study in evolution on students who weren’t graduate level or very close to it. The level of complexity a student would have to master to get a firm grasp on the evidence is high enough that to a lot of undergraduate students, this would be a plausible “just so” story. This research would fail as a textbook example right now, fine work though it is. For this to become a textbook example of evolution in action, I do think that there would have to be better explanations of the existing field experiments, and probably a whole lot of new ones, too.

Reference

Linnen, C., Kingsley, E., Jensen, J., & Hoekstra, H. (2009). On the Origin and Spread of an Adaptive Allele in Deer Mice Science, 325 (5944), 1095-1098 DOI: 10.1126/science.1175826