Metric variation in <i>Afromeryx</i> and <i>Libycosaurus</i> (Anthracotheriidae: Mammalia) and its utility for biochronology.

Abstract

Anthracotheres, like many other mammals, are useful for biochronology, especially those lineages that increased in dimensions with the passage of geological time. One such lineage comprises the tetracuspidate African anthracotheres of the genera Afromeryx and Libycosaurus, which started out as hog-sized animals in the Early Miocene and ended up larger than hippopotami by the end of the Late Miocene. In mammal groups with few genera at any one time, such as the Neogene anthracotheres of Africa, not only can dental data yield evidence of geological age, but so can postcranial evidence. In the case of Late Neogene suids, which comprised several contemporaneous genera, dental data is useful for biochronology, but only when morphological and metric data are combined. In such groups, postcranial evidence is not as valuable because it is difficult to determine with confidence to which group such bones belong, save in the case of skeletons associated with teeth, of which there are extremely few in the African fossil record. In Africa, in contrast, anthracotheres had low diversity throughout the Neogene, with usually three or fewer taxa present at most levels older than 17.5 Ma, and only one after ca 16 Ma to the end of the Late Miocene. Evidence shows that this lineage increased in size throughout its existence, from which it is inferred that knowing the dimensions of an African Middle or Late Miocene anthracothere provides information of a biochronological nature. The Toros-Menalla (TM) fossil zone in Chad has yielded an impressive sample of Late Miocene anthracotheres that was initially attributed to two species of different sizes (Lihoreau, 2003). The presence of two species in the Chad sample was confirmed by Pickford (2008a, 2008b) who drew out some biochronological information from the sample, showing that the TM sedimentary deposits probably spanned a considerable period of time, from ca 10 Ma to ca 6 Ma. Not only does the dental sample show this, but so does the postcranial one. The material was subsequently pooled into a single species by Lebatard et al. (2008) who deduced a much shorter duration of sedimentation for the TM sequence aged ca 7 Ma. The aim of this contribution is to re-examine the question of variability in the Chad sample of Libycosaurus and to compare it to collections from other African localities in order to determine whether the size differences observed are compatible with a single species (Lebatard et al., 2008) or whether they support the presence of two species in the region (Pickford, 2008a, 2008b). If two taxa occur at Toros-Menalla, then this would have implications for the biochronology of the deposits (Pickford, 2008a, 2008b) whereas if only one highly sexually dimorphic species is present, it would have implications for understanding the ecology of these anthracotheres.