Jared Diamond notes (p 161) that the wild ancestors of domesticated animals are spread unevenly – only two are in South America, while none come from North America, sub-Saharan Africa, or Australia. He then argues that only a few special species are good candidates for domestication, and there just weren’t any in these regions.
The world’s climate was similar to today’s, back the Eemian, the last interglacial (before this one). Every animal species that humans have ever domesticated existed back then, along with many other now extinct. Yet none were domesticated. As far as we know, no plants were domesticated, either. Why not? I’d guess it was because people were different back then. So that’s a possible reason. We could push this argument further, back to earlier interglacial periods, or even back before the Pleistocene: people must have been different.
I could believe this for Australia. Marsupials are significantly different from placental animals – dumber, for one thing – and most of the large ones in Australia had been wiped out a long time ago. Do I believe Diamond’s ” no good candidates” line for North America or Africa? No, I don’t. Zebras might have a lousy disposition, but then so did aurochsen. While eland are easy to domesticate. Tarpans, the wild ancestor of horses, only died out in the 19th century. We have descriptions: they were “absolutely untameable”. Yet here we are.
Lots of domestic animals started out with difficult behavior (aurochsen, tarpans, wild boars): domestication is always first and foremost for tameness. Selection can change behavior just as it changes size and speed and milk production. That’s selection on animal behavior, of course. There’s a special principle that keeps selection from changing behavior in humans.