Deep time and the search for anthropoid origins

Abstract

Recent fossil discoveries, phylogenetic analyses, revised reconstructions of continental drift, and accumulating molecular evidence have all yielded new information relating to anthropoid origins within the broader context of primate evolution. There is an emerging consensus among molecular studies that four superorders of eutherian mammals can be recognized: Afrotheria, Euarchontoglires (to which primates belong), Laurasiatheria, and Xenarthra. Overall, molecular phylogenies for mammals agree with some statistical analyses of the primate fossil record in indicating an early origin for primates around 85 Ma ago, and the divergence of haplorhines and strepsirrhines at ca. 77 Ma. Such an ancient date for the origin of haplorhines is some 17 Ma prior to the first known possible primate, and some 22 Ma before the earliest fossil evidence of undoubted euprimates. Because anthropoid fossils date back at least to the late Eocene and perhaps to the middle Eocene, and given indications of an early origin for primates, it is unlikely that ancestral anthropoids arose within any other currently known clade of fossil primates (adapiforms, omomyiforms, strepsirrhines, or tarsiiforms). Implications of new molecular, morphological, and biogeographic lines of evidence are explored with respect to the likely time and place of the origin of anthropoids. Four competing, testable hypotheses are reviewed in detail: 1) the Paratethyan hypothesis, 2) the continental Asian hypothesis, 3) the Indo-Madagascar hypothesis, and 4) the African hypothesis. A case is made that current evidence best supports a relatively ancient Gondwanan origin for primates, as well as a Gondwanan (African or Indo-Madagascan) origin for anthropoids at least as old as that of any other currently documented major primate clade. Available fossil evidence at present seems to be most compatible with the African hypothesis, but it is noteworthy that primates are included not in Afrotheria but in Euarchontoglires.