How tiny is tiny? Make it too small (say, the size of a toy poodle) and it probably can’t digest enough natural cellulose in order to sustain nutritional needs.
Stranger
Since it would be a domestic animal (if we bred it), we could probably design “poodephant” food pellets with all the nutrition it would need. No?
there were bones of some super small versions of huge critters on the galapagos werent there?
personally I want to see a bear dog, cant find the relevant link but apperantly there was an animal that was a sort of bear wolf hybrid that went extinct fairly recently.
What about the size of a small cow?
I hadn’t heard about that. Repitles on islands tend to get larger than their mainland relatives, and mammals tend to get smaller (although some very small mammals might get larger). This rule of thumb even seems to apply to humans-- Homo floresiensis.
How recent? I can’t think of anything that would fit that description that has gone extinct in historical times. And “bear wolf hybrid” would have to be a description of the physical characteristics, not an actual hybrid. Ursids and Canids went their seperate evolutionary way at least 20M years ago.
Wow. Got any pictures?
Of the offspring, I mean.
Dwarf elephants were once common on lots of mediterranean islands, all those are now extinct. And yes, they were the size of a cow, or smaller. And there were dwarf mammoths on north american islands.
Also, while bears and dogs split apart 20 million years ago, there used to be a family of bear-dogs that weren’t ursids or canids but amphicyonids. I’m not sure when the last amphicyonid went extinct, but they lived at least until the Pliocence, I haven’t been able to google up a reference to a Pleistocene species.
http://en.wikipedia.org/wiki/Amphicyonid
http://www.amnh.org/exhibitions/permanent/fossilhalls/vertebrate/specimens/amphicyon.php
How about approaching it from the other end? I mean, if you want a ‘dog’ the size of an elephant, would you settle for an elephant that’s been breed to have a personality like that of a dog? How amenable are ‘personality’ traits to selection?
Getting the elephant to have a nice hairy coat should be a snap, it’s lurking somewere in their genes from mammoth and mastodon days.
Sure, you could provide it with some kind of partially digested or easily broken down chow, and I guess the OP is interested only in size, not survivability.
I think–although I can’t find any links offhand–that elephants have been hormonally retarded to that size, albeit with attendant health problems. And the various fossil dwarf elephants/mammoths haven’t been much larger than the average bovine.
But the point is that there are are numerous factors that comprise the optimal and maximum/minimum size of an animal that brute directed breeding for a characteristic–i.e. breeding the largest members of a population together–will not significantly extend.
Stranger
Selectively breeding elephants for anything is going to be a lot harder than it is in dogs or sheep. Elephants have a much longer generation time, they only produce one offspring per pregnancy, and it takes 10 years for the baby to reach reproductive age. Plus each elephant takes literally tons of food, lots of space, and daily care. You could maintain a flock of a hundred sheep more cheaply than you could one elephant.
Even though elephants have been tamed they’ve never been domesticated. Captive elephants generally don’t breed well, and in Asia tame elephants are usually captured young rather than being born in captivity. Which means there’s no chance for humans to control breeding.
Also, mammoths and mastodons are NOT ancestors of modern african or asian elephants, any more than asian elephants are ancestors of african elephants or vice versa. Mammoths and mastodons and all the other Pleistocene proboscideans were relatives of today’s elephants, not ancestors. Today’s species of african elephant (Loxodonta) was present in africa during the ice age, today’s asian elephants (Elephas) were present in asia during the ice age.
However, it is true that breeding for hairyness wouldn’t be that difficult. Baby elephants are covered with coarse hair although they lose most of it, and adult asian elephants are much hairier up close than you might expect.
And on Wrangel Island (off Russia, in the Arctic ocean) as recently as a few thousand years ago-- well into historical times.
BTW, I mispoke above about extant pygmy elephants. There are pygmy hippos, but not any pygmy elephants anymore.
Well, only if there is some reason for such a beast.
Only recently has society become ‘rich’ enough to support pets that serve no functional purpose (companion animals). Before this, dogs served functional purposes as herding, hunting, retrieving, or guarding dogs.
A cow-sized dog would not be significantly better at any of those functions, but would have significantly increased costs in food and shortened life. Not worth it!
P.S. The largest breeds of dogs (Irish wolfhound, Mastiff) are actually larger than some of the current breeds of miniture cows.
What do you mean by"recently"? Certainly the Egyptians (1000 BC) , Romans (1 AD), Chinese (1000 AD), or Europeans (1600 AD) were “rich” enough to have luxury dogs. Most dog breeds are “recent” compared to those timeframes. If a Roman Emperor wanted a certain breed of dog, I’m sure he could found some dog breeder who would be more than happy to try and accomodate him. 
Well, in terms of the thousands or millions of years of evolution of dogs and humans, that 3000 years is “recent”. And dogs were differentiating into breeds before then. There is evidence that the Egyptians were breeding greyhounds for speed back then. But you’re right that many of the modern breeds weren’t ‘breeds’ yet.
I’m sure a Roman Emperor could have ordered this; they could have ordered juat about anything. For some reason, we don’t have any records of them doing such a thing. They seem to have stayed with functional breeds: Egyptians bred for faster greyhounds, Romans bred for bigger & faster guard dogs, etc. Either nobody important had thought of the idea of the toy lapdog type, or they thought it too silly to try to breed for that. Whatever, they didn’t seem to ever try it. Even strange cases like the Emperor Caligula’s pet horse Incitatus, which slept in his bedroom and was appointed or proposed as Consul was still a functional beast – Caligula used him as a riding horse. The whole concept of non-functional ‘pet’ animals seems missing back in history.
Yes and no. Wolves have been around for a long time, but we don’t have clear evidence of domestication of wolves (dogs) until about 14k years ago. So 3k years is still a pretty good chunk of time overall. And we can find relatively wealthy societies further back than 3k years.
Yeah, especially if we think of breeding for “show”. At any rate, we’re getting into GD territory, and I just wanted to point out that lack of wealth probably isn’t the reason no one in antiquity tried to breed extremely large dogs.