It seems important to comment on the postcranial characters used by Beard (1993a) in his phylogenetic analysis. The primatomorph node was considered the most robust, supported by eleven characters: one dental and 10 postcranial. Some of these characters may now appear problematic in their definition and distribution. From the first three of them, the humeral ones, two are mentioned as examples. The lesser tuberosity is considered robust and medially protruding in "primatomorphs" by Beard (1993a). This tuberosity is in fact less protruding in many primates than in Phenacolemur and Plesiadapis (Szalay et al., 1975), and it has been shown to be as robust and as protruding in Ptilocercus, giving the more probable state of the character in primitive scandentians (Sargis, 2002a). The subspheroidal to spheroidal shape of the capitulum is also found in Ptilocercus, and is considered as a likely primitive archontan character by several authors (Sargis, 2002a,c; Szalay and Lucas, 1993, 1996). These two humeral characters have interesting functional implications, but they do not support a primatomorph node.
A complete reappraisal of all the postcranial characters studied by Beard (1993a), Silcox (2001), and Bloch and Boyer (2002) is far beyond the scope of this chapter. Some general comments are provisionally given. All but one of the characters taken by Beard (1993a) to support Primatomorpha are linked to some kind of arboreal adaptation (these characters would increase to 13 with the inclusion of the coxo-femoral characters which are part of the same functional complex, and which also occur in Chiroptera and were listed as primatomorph in Beard, 1991). A number of them appear to be found in Ptilocercus and are considered as probably primitive archontan (Sargis, 2000, 2002a,b,c; Szalay and Lucas, 1993, 1996). Some others, linked to leaping in Tarsius and Hemiacodon, are questionably part of the primate morphotype. In fact, it would seem hazardous to give a list of all the postcranial attributes of the primate morphotype, because this is a debated notion, which depends on one's preferred interpretation of anthropoid characters (e.g., Dagosto, 1990; Ford, 1988; Godinot, 1992). This choice should be justified. Furthermore, the plesiadapiform radiation is a large radiation of presumably mainly arboreal animals, which leads me to suspect a very complex history of arboreal adaptations in the group. Some of the characters studied by Beard show an evolution within plesiadapiforms. For example, plesiadapids have a limited area for insertion of the M. quadratus femoris on the posterior side of the femur. Beard (1993a) interpreted this primitive state as a reversal; however, it could as well indicate that plesiadapids were less specialized than other "primatomorphs" in their arboreal adaptation. This could be confirmed by the variation in the calcaneocuboid articulation alluded to by Beard (1993a). Plesiadapids is taken as having a derived pivot-like calcaneocuboid joint—a putative primatomorph character, whereas Szalay and Decker (1974) carefully described the calcaneocuboid joint as much more primitive in Plesiadapis than in primates, without a real pivot [Wible and Covert (1987) also mention that the calcaneum of Plesiadapis cookei has no pronounced groove for the tendon of M. flexor fibularis. Does this imply intraspecific variability, or more?]. One would then infer that plesiadapids, and other more primitive plesiadapiforms, were probably less specialized than Beard's reconstructed primatomorph morphotype. A complex history of arboreal adaptations in the different families of plesiadapiforms appears likely, implying a correlative history of some anatomical structures. The cal-caneocuboid joint could have undergone several changes in plesiadapiforms, as it has in primates (Gebo, 1988; Gebo et al., 2001). A complex history is already demonstrated by some of the new skeletons, with the remarkable finding of a fully opposable hallux bearing a nail in a carpolestid (Bloch and Boyer, 2001, 2002). In such a context, postcranial characters, which are known to be prone to convergence due to similar functional demands, will probably turn out to be, as the dental characters, very good for intrafamilial phylogeny and locomotor history, but much more difficult to use for the higher level phylogeny. For example, Hamrick et al. (1999) find that a series of phalangeal characters of paromomyids and dermopterans are shared derived. However, including Dactylopsila, Daubentonia, and perhaps Heterohyus would probably have destroyed the support that they found in favor of paromomyid-dermopteran relationships.
It has been realized that eutherians have a history of arboreal adaptation older than what was previously hypothesized, dating back to the Late Cretaceous, that the broad polarity of some postcranial characters of eutherians has become less secure (Godinot and Prasad, 1994; Prasad and Godinot, 1994), and that many more arboreally adapted groups should be taken into account [e.g., mixodectids (Szalay and Lucas, 1996)], apatemyids, possibly also nyctitheriids (Hooker, 2001). With such a complex background, highlevel phylogenetic inferences will require either a more continuous record of morphologies, or enough experience to spot rare characters, unique transformational series. Within the wealth of postcranial characters, there is no reason why some would not turn out to have a high phylogenetic value, as some carpal and tarsal characters have proven to have in primates and beyond. For the time being, the author is doubtful about the carpal characters used by Beard (1993a) because they rely too heavily on the sole triquetrum of Phenacolemur, which is on the one hand not too different from that of Plesiadapis, and on the other so different from that of Cynocephalus that homologies between them are not straightforward (see also Stafford and Thorington, 1998; Szalay and Lucas, 1996). The more complete skeletons found in the Bighorn Basin are very promising (Bloch and Boyer, 2001, 2002). They should help both testing some of Beard's hypotheses and deciphering good phylogenetic signals. On the whole, despite remarkable progress in the knowledge of plesiadapiform postcrania and their functional interpretation, the use of postcranial characters for deciphering early archon-tan phylogeny (excepting the question of the Volitantia) is considered very conjectural and risky. However, tarsal characters will be mentioned again in later section.
Was this article helpful?