Another Tool for Inferring Function: Maps
In The Heretic in Darwin’s Court, Ross Slotten describes the debate between Darwin and Wallace about the function of zebra’s stripes:
The zebra’s stripes, which Darwin regarded as useless on the open African plains, alerted a straggler to the location of the herd or camouflaged the animal when foraging or resting in the bush. ‘Until the habits of the zebra have been observed with special reference to these points,’ Wallace wrote, ‘it is surely somewhat hasty to declare that stripes ‘cannot afford any protection’”
A new paper, covered in a number of media outlets, vindicates Wallace’s proposal, but the protection is of a different sort from what he had in mind.
Caro et al., in a paper entitled “The function of zebra stripes,” suggest that the zebra stripes are not for protection against large predators, but rather against small ones: insects.
The paper has been covered by any number of media outlets – you know that a story is not news anymore when even Saturday Night Live Weekend Update picks it up (5:08) — so I’ll just comment on one aspect of the work that fits with a theme that I’ve possibly mentioned once or twice before on the pages of this blog, the nature of evidence of function.
A few years ago, for example, I wrote about horses of the sea variety, discussing work that used models of seahorse shapes to infer that the function of their distinctive heads had to do with increasing the range at which they could catch prey. The idea, which I think is an important one, is about the epistemological question of what sorts of evidence can be used to infer the function of a trait, whether physical or behavioral.
This issue arises in evolutionary psychology because Various Critics think to teach us in the field about what sorts of evidence biologists use when they make claims about function. This often has to do with genetics, heritability, phylogeny, and so forth. The point of the seahorse post was to say that while of course other sources of evidence are relevant to the question of inferring function, evidence from computer models also bear on the question. The key is finding or creating data that speak to the design features of the trait from the standpoint of the putative function.
The reason that I like the paper about zebra’s stripes is the evidence that it brings to bear in the service of the aggressive claim in the title of the paper, the function of zebra stripes. The authors reasoned as follows. Suppose zebra stripes, as Wallace seems to have thought, afford protection against large predators, such as hyenas, lions, and tigers. If so, then species that live in places with higher densities of these predators ought to be more likely to have stripes than closely related species who live in places with lower densities of these predators. Similarly, if stripes defend against insects, then species that inhabit regions with more of the insects that prey on zebras should be more likely to have stripes than species that inhabit regions with lower densities of these insects.
To look at this, the authors used maps, measuring the overlap between striped zebras compared to their non-striped relatives and the distribution of predators of the large (hyenas, etc.) and small (insect) variety. (See the accompanying figure.) In part, their results were that “the presence of body stripes was very strongly associated with presence of tsetse flies.” At the end of the first paragraph of the paper, the authors write: “A solution to the riddle of zebra stripes, discussed by Wallace and Darwin, is at hand.” Stripes are for protection form insects.
If that’s the function, what’s the mechanism? It seems to have to do with the perceptual apparatus of the flies. The alternating pattern of black and white might confuse the insects’ system sufficiently to reduce their ability to land on the zebras. Previous work showed that the distinctive zebra striping pattern attracted fewer houseflies (tabnids) by virtue of the fact that “the light and dark stripes of a zebra’s coat reflect very different polarizations of light in a way that disrupts the attractiveness to tabanids” and, further, that “the stripe widths of zebra coats fall in a range where the striped pattern is most disruptive to tabanids.”