Asked by: Paul Noble, by email
Hmmm, tricky! Not only is it unclear when two groups can be called different species, but recent research shows chance details of genetics and ecology massively affect the process of speciation.
For example, it helps when natural selection is driving two groups in different directions. In this case, the fastest instances can occur without full separation, such as in the cichlid fishes of Lake Victoria. In only 15,000 years, between 500 and 1,000 new species of the fish have arisen, probably through a combination of natural and sexual selection. That implies speciation in a few thousand years: the founding group must already have been pretty diverse. Indeed, existing genetic differences in different areas probably lie behind most speciation.
Broadly, think in the region of tens of thousands or a few million years (or generations). For example, studies of partially interbreeding fruit fly species estimate about three million years for complete separation – or 200,000 years in the rarer cases of speciation with continued interbreeding. Perhaps it’s not surprising that speciation in the lab is rarely observed!