A new approach to an old problem: homology of postcanine teeth in therian mammals

Since at least Richard Owen [1], drawing inferences about the homology or identity of teeth across all therian mammals has presented a thorny problem: it has never been clear how the postcanine teeth of marsupials relate to those of placental mammals.Deciding which tooth to compare with which is an important issue for anyone combining observations on the teeth of marsupial and placental mammals in the same analysis, no matter whether one's questions are matters of phylogeny, comparative biology or transformative evolution.Generally, in mammals, teeth possess traits that make them highly individualizable, making them nearly ideal structures for studies of homology or for recognizing the same tooth in different species or even across gulfs of time.Mammalian teeth are readily classified by position (locus), and the replacement of teeth is limited to at most two generations per locus, resulting in there being fewer teeth in most mammals over a lifetime to consider.Teeth in mammals are organized into classes (incisors, canines, premolars and molars), which can be related to well known landmarks independent of the teeth themselves.For example, upper canines are generally the first, or most anterior, teeth to be at least partially within the maxilla, as opposed to the premaxilla.Lower canines generally are corresponding teeth in the dentary that occlude with the front of the upper canines.Even in cases where the lower canines themselves are reduced and no longer the most 'caniniform' teeth in the dentary, they can still be recognized by their position [2].Incisors are simply located in front of their respective canines.Among postcanines, premolars and molars differ in that replacement occurs at premolar loci but not at molar loci.Molars and deciduous premolars are often more complicated (i.e. with multiple cusps and precise occlusion necessary for chewing) than replacement premolars, which are often simpler (suitable for puncturing and crushing).With premolars and molars defined by replacement, extant placental mammals generally possess four premolars and three molars per jaw quadrant or at least are derived from an ancestral condition with those counts, whereas marsupials possess three premolars and four molars, for the same total count of seven postcanines in both groups.Owen noted this shared similarity in postcanine tooth number and suggested a hypothesis of homology that would account for this shared total and at the same time explain the different numbers of premolars and molars in each group.Recognizing that molars and the deciduous teeth compose the primary generation, molars can be seen as a continuation of the same series as the deciduous premolars, which they often resemble in form and function.If the last deciduous premolar were to be retained without replacement, it would then be indistinguishable from a molar in Owen's view.These propositions led to Owen's suggestion that the first molar of marsupials may be homologous to the last deciduous premolar of placental mammals, a proposal taken up by later authors [3,4] to the point where it has now become almost canon.It is now known that stem eutherians of the Early Cretaceous possessed five premolars for a total count of eight postcanines, lessening the