Going to the Dogs…

By Peter Dearden 03/09/2010 2

Peter K. Dearden. Laboratory for Evolution and Development, Genetics Otago and the National Research Centre for Growth and Development, University of Otago

I own a dog. A nice one called Eddy, whose aim in life seems, now he is getting on a bit, to eat, and go for walks briefly, between long periods of sleeping. To quote Ted Hughes ‘he flops flat, and digs down deep, like a miner, into sleep.’ Eddy is a very domestic dog, and he has very floppy ears.
I mention these things because they are the subject of a very cool paper, recently published in PLOS Biology, that looks at the genetics of traits such as size, ear floppiness and head shape in dogs.
Why? Dogs come in a remarkable diversity of shapes and sizes unmatched, in my opinion, by any other domestic animal. In this paper a large group of American scientists try to map genetic diversity in a bunch of dogs, and see how it correlates with these different shapes.
They start with a whole pile of dogs, 915 to be exact, from 80 dog breeds, plus 83 wild dog shaped things (wolves, jackals and coyotes) and 10 African shelter dogs. Now I am not quite sure what an African shelter dog is, but the paper says they are domestic dogs, but out-bred and not recently selected by humans.
Having collected the dogs, they extract DNA and, using the sequenced dog genome, map genetic variation in all these dogs, and correlate that variation with the traits they are interested in. In doing so they find some very unusual things.
If you do similar sorts of studies with human populations, looking at height, or BMI, or a complex disease trait such as type II diabetes, you find hundreds of bits of the genome that, added together, seem to influence the trait; implying that lots and lots of genes, each having a very small effect, add together to produce the final phenotype. This is true in many other species, but not in dogs. In dogs, even given the vast variation in dog forms, only a few genes, each having a large effect, give rise to those huge differences in doggy shape.
Why? Why is variation in dogs produced by a simple architecture, when such variation is usually complex? The authors suggest that this is because of the strange history of dogs. Their maps of variation in dogs show that the genome has been through two periods of strong selection, one when dogs were domesticated, missing from wild dogs, and one very recent one, linked to the development of breed standards in the 19th century, missing from the African shelter dogs. This second period of strong selection is very unusual, being characterised by very small population sizes in each breed, and very strong selection for novelty.
I find it fascinating that a few small changes in genes can have such profound effects on the shape of an animal. Perhaps even more interesting is the idea that the way that an animal has been domesticated, or even its evolutionary history, might affect how morphological traits vary.
It is sobering to know that the domestication history of the large dog, now snoring loudly beside me, has affected not just its shape, but also the very way that his genes made that shape. Dogs have been remarkably changed through becoming our best friends.

2 Responses to “Going to the Dogs…”