Paromomyid dental characters appear incompatible with the hypothesis of a sister group relationship between paromomyids and dermopterans (the pri-matomorph hypothesis). All hypotheses placing dermopterans or primates nested within plesiadapoid families are dentally quasi impossible.

Paromomyids were not gliders but more probably claw climbers having locomotor and insect capture specializations close to those of Heterohyus and Dactylopsila. Skull characters do not show any paromomyid-dermopteran synapomorphy (Bloch and Silcox, 2001). The Primatomorpha hypothesis, which now lacks postcranial support and cannot be reconciled with dental or cranial evidence, must be abandoned.

The plesiadapiforms are a radiation of forms combining a series of primitive skull characters, small orbits and large olfactory bulbs, and very derived characters of the anterior incisors and muzzle. Despite their dental convergences with primates, early acquired dental specializations exclude them of having primates rooted within them. They are a radiation of clawed arboreal mammals, within which Carpolestes hallucial grasping represents a remarkable convergence with primates. Many of them are North American and they are probably monophyletic (Kay et al., 1992; e.g., all those descended from Purgatorius). An interesting question remains concerning the possibility that plesiadapoids might have an independent Asiatic origin, as raised by some phylogenetic interpretations of Chronolestes dentition; however, this would not alter the broad picture inferred from their cranial and postcranial characters. Hence, Plesiadapiformes are best considered as an order of their own (admittedly having an imprecise content). The evidence favoring a sister group relationship of plesiadapiforms with primates is ambiguous.

The exclusion of plesiadapiforms from primates renders more plausible a series of potential synapomorphies between primates and Tupaiidae. These include characters from the basicranium, the orbit, other parts of the cranium (Kay et al., 1992; Wible and Covert, 1987; Wible and Martin, 1993), and a series of important characters of the astragalo-calcaneum complex. Tupaiidae appear again to be the best available sister group of primates.

Large data sets of morphological characters have proven unsuitable to clearly decipher the sister group of primates. A strategy of pruning the data and adding as much historical information as possible seems more appropriate than simply adding characters to the list. The likelihood of particular reversals should be considered when evaluating alternative hypotheses.

A complex history of arboreal adaptations in the various families of plesiadapiforms and in early archontans is strongly implicated. More comparative work, taking in account Ptilocercus, is needed before phylogenetic signals can be extracted from the postcranial anatomy. However, tarsal characters are sufficiently understood to support the hypothesis of Ptilocercus-primate synapo-morphy for a series of tarsal characters. The author suggests that several tarsal characters of the eosimiids and other Shanghuang primates are primitive for the order. These hypotheses in turn challenge the leaping aspect of the theory of grasp-leaping as the primate morphotype locomotor mode, favoring a mode closer to grasp-quadrupedalism, and not requiring multiple reversals for the origin of simiiform locomotor characteristics.

Small arboreal marsupials such as Marmosa and Caluromys show similarities with primates linked to locomotion on fine branches; however, they do not have nails on all digits. Insect-catching in the arboreal milieu probably better explains the acquisition of primate postcranial characterisics (powerful hallucial grasping, nails, and the peculiar proportions of early primate hands), and visual predation explains the cranial characters as proposed by Cartmill (1972, 1974a,b).

Soft anatomical and molecular characters should be used to test the tupaiid-primate hypothesis. A denser record of tarsal and dental characters also should give decisive confirmation, or refutation, of this hypothesis. The primate dental morphotype must be reassessed after discarding the plesiadapi-forms and including the eosimiids. It is presently far from being established.

Since this essay was completed, several papers related to the subject were published. Among them, only one would chage my view of the plesiadapiform radiation, that is if Dralestes and the Azibiidae were Eocene African plesiadapiforms (Tabuce et al., 2004). However, the lingual part of the upper molars of Dralestes is relatively narrow, suggesting that their posterolingual crest is not a postprotocingulum homologous with that of known plesiadapiforms but more likely represents convergent structure. Tabuce et al. (idem, p 318) recognize that "the phylogenetic position of azibiids needs to be confirmed with more relevant data". The author agrees, doubts their plesiadapiform affinities, and suggests elsewhere that they might be unexpected euprimates.

0 0

Post a comment