The myth of the hippo-like anthracothere: The eternal problem of homology and convergence.

Resumen

The notion that anthracotheres had hippo-like body proportions, locomotion and lifestyles has been in the literature for so long, and has been repeated so many times, that it has taken on the aura of unquestionable truth. However, right from the beginning of studies into hippo-anthracothere relationships over a century and a half ago, observations were made that revealed the existence of fundamental differences in dental, cranial and post-cranial anatomy in the two groups. From 1836 to 1991 two skeletal characters (a descending plate at the angle of the mandible, and raised orbits) have overshadowed all others in suggesting close relationships between hippos and a single anthracothere genus (Merycopotamus) later to be joined by a second genus, Libycosaurus, in 1991 for the descending angle, and 2003 for the raised orbits (Lihoreau, 2003; Pickford, 1991). Close examination of these structures reveals that they are not homologous in the two groups, yet they have played an inordinately stubborn role in interpretations of the relationships between them, featuring in papers as recently as 2005. The rest of the skeleton and many cranio-dental features revealed, as early as 1836, that anthracotheres did not look particularly similar to hippos, either in gross body plan, or in details of the skeletal anatomy, observations that have been confirmed at irregular intervals ever since. Yet, despite the divergent morphology, most authors continued to attribute hippo-like locomotion, behaviour and ecology to the anthracotheres that they studied, whether anthracotheriines or bothriodontines. Two broad themes have run side by side in the long history of study of hippo-anthracothere relationships, «homology versus convergence» and «early versus late divergence», early divergence implying the existence of a ghost proto-hippo lineage of some 30 million years duration. Indeed these two themes are linked together, in the sense that proponents of early divergence have tended to interpret the similarities between hippos and anthracotheres as convergences, whereas those who have proposed late divergence usually took the perceived similarities to represent homologies. All these interpretations were played out within the context of a much broader background debate about monophyly or paraphyly of the artiodactyls, which was actively discussed in the Victorian era just as it is today. More recently, molecular studies have altered the scope of the debate, principally by indicating closer affinities between whales and hippos than between hippos and other artiodactyls. In the search for the ghost lineage that should link hippos to whales, some authors have recently suggested that anthracotheres fill the role «robustly», whereas others have suggested that anthracotheres are not closely related to hippos, whilst yet others have proposed that palaeochoerids or cebochoerids may represent the missing lineage. The aim of this paper is to review the possible role of anthracotheres in the evolution of hippopotamids. It is concluded that they played no part in it, whereas palaeochoerids could well represent the ghost lineage that has evaded scientists for more than a century.