The patterns we observed in this study support the idea that
multiple factors have been important. For example, allopatric speciation
due to the contraction of the forest and the formation of
refugia could have been the cause of speciation for A. belzebuth
and A. chamek as the recent timing of their split (1.8 mya) coincides
with a time during the Cenozoic when ecological conditions
were not stable in tropical latitudes, especially during the early
Pleistocene (Haffer, 2008; Morley, 2000). During this time, alternating
humid and dry climatic periods, as well as warm and cold
periods, led to continuous vegetation changes and allopatric speciation
and subspeciation of other animal populations within
Amazonia due to the contraction of forested and non-forested
areas (Haffer, 2008). The origin of Ateles hybridus, on the other
hand, was more likely due to isolation and divergence caused by
the uplift of the Andes, consistent with a paleogeographic barrier
hypothesis. The uplift of the northern Andes started 23 mya but
had peaks of activity 12 mya and 4.5 mya (Hoorn et al., 2010),and the latter date is close to the estimated time of the divergence
of A. hybridus either from the common ancestor of A. paniscus + A.
geoffroyi + A. fusciceps under the first scenario described or from
the common ancestor of A. belzebuth + A. chamek. At this time,
spider monkey populations located in the Río Magdalena valley
would have been isolated from populations in the Amazon basin
and probably from other spider monkeys in northern Colombia