Our initial BEAST analysis included additional primate taxa,
included the extinct atelid Stirtonia as a fossil constraint, incorporated
sequence data from A. marginatus, and only enforced monophyly
constraints within the genus Ateles on those clades that were
consistent across all of our phylogenetic analyses (i.e., A. belzebuth
+ A. chamek and A. fusciceps + A. geoffroyi). This analysis estimated
the age of the split between spider monkeys and other
atelids to have occurred 13.4 Ma (95% credibility interval 11.6–
15.3 Ma) (Fig. 4 and Table 6). In the inferred topology, which
matched that recovered in our initial likelihood analysis (Fig. 3A),
A. marginatus diverged from other spider monkeys 6.7 Ma (95%
credibility interval 4.3–9.5 Ma). A clade of A. belzebuth + A. chamek
then diverged from the other species of spider monkey species
4.5 Ma (95% credibility interval 2.7–6.6 Ma). Ateles hybridus split
from other Ateles around 4.0 Ma (95% credibility interval 2.3–
5.8 Ma). The divergence between the trans-Andean forms and A.
paniscus is inferred to date to 3.5 Ma (95% credibility interval
2.0–5.1 Ma). The Mesoamerican forms and Ateles fusciceps are
inferred to share a recent common ancestor 2.2 Ma (95% credibility
interval 1.2–3.5 Ma). Based on these coding regions, Ateles geoffroyi’s
subspecies are inferred to begin to diverge 0.6 Ma (95%
credibility interval 0.2–1.1 Ma), while those of A. fusciceps diverged
1.3 Ma (95% credibility interval 0.6–2.2 Ma) (see also Morales
Jimenez et al., 2015, for further dating analysis using noncoding
mtDNA sequences)