We sequenced the complete mitogenome of 39 sloths (19 Bradypus variegatus, 4 B. tridactylus, 1 B. pygmaeus, 1 B. torquatus, 4 Choloepus didactylus, and 10 C. hoffmanni). A Bayesian tree (BI) indicated a temporal split between Bradypus and Choloepus around 31 million years ago (MYA, Oligocene) and the other major splits within each genera during the Miocene and Pliocene. A haplotype network (MJN) estimated a lower temporal split between the sloth genera (around 23.5 MYA). Both methods detected the ancestor of B. torquatus as the first to diverge within Bradypus (21 for BI and 19 MJN), followed by that of the ancestor of B. tridactylus. The split of B. pygmaeus from the common ancestor with B. variegatus was around 12 MYA (BI) or 4.3 MYA (MJN). The splits among the previous populations of B. variegatus began around 8 MYA (BI) or 3.6 MYA (MJN). The trans-Andean population was the first to diverge from the remaining cis-Andean populations of B. variegatus. The genetic differentiation of the trans-Andean B. variegatus population relative to the cis-Andean B. variegatus is similar to that found for different species of sloths. The mitogenomic analysis resolved the differentiation of C. hoffmanni from the C. didactylus individuals of the Guiana Shield. However, one C. didactylus from the Colombian Amazon specimen was inside the C. hoffmanni clade. This could be the first example of possible natural hybridization in the Amazon of both Choloepus taxa or the existence of un-differentiable phenotypes of these two species in some Amazonian areas.
Keywords: Bradypus; Choloepus; mitogenomics; mt Cyt-b; trans-Andean B. variegatus.