Abstract

The debate regarding the evolutionary relationships of the extinct South American native ungulates (SANUs) to the major placental clades Afrotheria and Boreoeutheria is exciting and has profound implications for our understanding of their early diversification and paleobiogeography. Although this controversy has not yet proven resolvable using morphological evidence, paleoproteomic and ancient DNA analyses support that at least some SANUs (i.e., Litopterna and Notoungulata) are members of Boreoeutheria, closely related to Perissodactyla (the Panperissodactyla hypothesis). Here we present a critical assessment of a recently published morphology-based study that claims that: (1) some SANUs (i.e., Notoungulata, Astrapotheria, Pyrotheria, and Xenungulata) represent a monophyletic supraordinal group, the Sudamericungulata, closely related to the Afrotherian hyracoids; and (2) the remaining SANUs (i.e., Litopterna and Didolodontidae, placed in a separate taxon, Panameridiungulata) are boreoeutherian in origin. Because this proposal (hereafter, the Sudamericungulata - Panameridiungulata or S-P hypothesis) is based on an incongruously reduced sample of boreoeutherians (including only a single perissodactyl) and inadequate character sampling restricted to dental and mandibular traits, it cannot be regarded as a satisfactory test of SANU relationships. Moreover, the S-P hypothesis fails to recover monophyletic Boreoeutheria and/or Afrotheria, making it incompatible with all well-established hypotheses of placental diversification. We find that the introduction of molecular constraints forcing the monophyly of Boreoeutheria and Afrotheria produces new trees, all recovering Sudamericungulata and Panameridiungulata nested within Boreoeutheria. These results are consistent with our analyses using a corrected version of the S-P matrix. Although we acknowledge that boreoeutherian affinities have still not been conclusively demonstrated for all nominal SANUs, it is beyond argument that any further credible testing must be based on much more exhaustive surveys than are currently available.