advantage of modern methods of phylogenetic inference – both maximum likelihood and Bayesian methods yielded similar topol- ogies for the phylogenetic tree of spider monkeys and one which differs markedly from previous studies (Fig. 3). Most significantly, when Ateles marginatus from northeastern Brazil is included in our analysis, we find that this species does not belong in the same clade as A. belzebuth and A. chamek as previously suggested. Rather, it is inferred to be the sister taxon to all other species of spider monkeys. Moreover, several of our analyses place Ateles paniscus in a rather derived position in the Ateles phylogeny, as a sister taxon to a clade containing the trans-Andean forms (A. geoffroyi in Mesoamerica and A. fusciceps from the Pacific coast of South America). Most previous studies (and ours, too, when A. marginatus is not included in our dataset: Fig. 3E and F) have instead proposed A. paniscus to be basal within the Ateles radiation, although one cytogenetic study has placed it in a similar derived position to that
which we reconstruct here (Medeiros et al., 1997). In other respects, our results are concordant with those of Froehlich et al. (1991) and Collins and Dubach (2000a,b) in finding that A. fusciceps and A. geoffroyi shared a relatively recent common ancestor and in suggesting that A. hybridus belongs in a distinct clade from A. bel- zebuth and A. chamek.