Ancestry and Reversion

Lots of good thinking about the evolution of domestic animals this week. Odious started out with a post on Heck cattle, the Hagenbeck brothers' attempt in pre- WW II Germany to breed a reconstructed aurochs. Comments came in from Chas and Doc Hypercube, who was also interested enough to post on Darren's theories about dog origins on Diary of a Mad Natural Historian:

"My thought on all this is that it seems to show that once genetic material is lost in a population (as a result of domestication) it’s gone - all the king’s horses, etc. Perhaps, then, it’s not shocking that dogs revert to pariah dogs in some number of generations - the genes that would allow reversion to a more wolfy form are just gone from the domesticated dog’s gene pool. To extend the title of the post a bit - border collies are to pariah dogs as Holsteins are to criollos/Texas longhorns? Don’t get me wrong - there are a number of other persuasive points in Darren’s post - I’m just not sure about the ‘won’t revert to wolves’ argument at this point."

I reponded with my own theories, which tend to be close to the ones Darren expressed:

"I think there are still mysteries, many mentioned by Darren, and even Dave Mech's correction in his comments don't lay them all to rest for me.

"If you saw the (unusually excellent) [ Only the first half-- the second among other idiocies said salukis were too refined to actually KILL hares--SB] PBS documentary on dog origins last night you will be familiar with Ray Coppinger's "dogs domesticated themselves" theory, Belyaev's fox domestication by selecting only for tameness, Saivolainen's genetics (my dogs are in it!) etc. These rather assume but don't demand a wolf ancestry.

"Darren says: "introgressive hybridization between the two is so limited that it doesn’t pose a threat to the genetic integrity of the wolf. Wolves and domestic dogs are in fact staying distinct. This applies globally as well as locally: despite continuous, near-global sympatry between domestic dogs and wolves, hybridization has hardly occurred and only one mtDNA type is shared". This seems really important to me. Coyotes are the outlier in a dog- wolf- coyote cladogram-- reasonably enough considering their New World origin. But wolves and coyotes hybridize far more often than wolves and dogs or coyotes and dogs. Red wolves are now considered to carry genes from both, and the same for the big wild canids of eastern Canada and adjacent New England. A barrier-- behavioral?-- exists between dogs and other (wild) canids. And as Darren says, the behavioral differences are real (and probably bigger than those between, say, the aurochs and the Spanish fighting bull or longhorn).

"If the split were really old, pre- modern human, what does that imply? Also, on the show last night, Ray Coppinger explained that scavenging wolves without a flight reflex would form the gene pool for dogs. This struck me as WRONG. I have been carrying on an email discussion with the eminent biologist Valerius Geist, who has been studying what interactions lead to wolves attacking humans, and the first sign is that they lose their fear of humans. It is a bad sign! Maybe "wolves" or whatever that show submission, but not lack of fear.

"Also re breeding back-- did you see my comment following yours on O & P about breeding back the quagga from the (conspecific) Burchell's zebra, which was going well last I heard? No Heck cattle they!

"BTW Val will be doing a lot I think to show that the lack of human attacks in North America by C. lupus is a historical accident and artifact and is likely to end bloodily. ( Bad attacks recently in Canada, one ending with someone eaten."

And if this is all not too bloggish in its recursiveness, Odious decided to track down that Quagga link to update on what may be a genetic phoenix, rising again from temporarily hidden genes.