Data from: Reconstructing the demographic history of orang-utans using approximate Bayesian computation

Investigating how different evolutionary forces have shaped patterns of
DNA variation within and among species requires detailed knowledge of
their demographic history. Orang-utans, whose distribution is currently
restricted to the Southeast Asian islands of Borneo (Pongo pygmaeus) and
Sumatra (Pongo abelii), have likely experienced a complex demographic
history, influenced by recurrent changes in climate and sea levels,
volcanic activities and anthropogenic pressures. Using the most extensive
sample set of wild orang-utans to date, we employed an approximate
Bayesian computation (ABC) approach to test the fit of 12 different
demographic scenarios to the observed patterns of variation in autosomal,
X-chromosomal, mitochondrial and Y-chromosomal markers. In the
best-fitting model, Sumatran orang-utans exhibit a deep split of
populations north and south of Lake Toba, probably caused by multiple
eruptions of the Toba volcano. In addition, we found signals for a strong
decline in all Sumatran populations ~24 ka, probably associated with
hunting by human colonizers. In contrast, Bornean orang-utans experienced
a severe bottleneck ~135 ka, followed by a population expansion and
substructuring starting ~82 ka, which we link to an expansion from a
glacial refugium. Therefore, we showed that orang-utans went through
drastic changes in population size and connectedness, caused by the
recurrent contraction and expansion of rainforest habitat during
Pleistocene glaciations, and probably also by the impact of hunting by
early humans. Our findings also emphasize the fact that important aspects
of the evolutionary past of species with complex demographic histories
might remain obscured when applying overly simplified models.