A series of tupaiid tarsal characters were interpreted by Szalay and Drawhorn (1980) as derived (as opposed to primitive archontan retentions): astragalus with the groove for the tendon of M. flexor (digitorum) fibularis aligned parallel to the long axis and located upon a ventrally projecting medial body; length of the astragalus relatively large in contrast to the squat bones in Paleogene archontans; sulcus astragalus not approaching the trochlear groove for the tendon of M. flexor fibularis (this may be related to preceding character); calcaneum having lost the primitively large peroneal process found in early archontans; on calcaneum, the anterior plantar tubercle is greatly reduced and has receded more distally [sic, they meant proximally] than it probably was on the ancestral archontan. In fact, these characters and some others can be reinterpreted as potential synapomorphies of tupaiids and primates.
The greater length of the astragalus in comparison with that of plesiadapiforms makes it primate-like (Figure 6). This neck elongation has been linked by Dagosto (1988) with the elaboration of subtalar motion. The most striking difference between the Ptilocercus astragalus and those of all plesiadapiforms (Paleocene "primates") described by Szalay and Drawhorn (1980) is that the latter have a body that is lower medially than laterally; the medial trochlear crest is lower and less salient—a primitive state. In contrast, Ptilocercus has a body that is almost as high medially as laterally, and the two rims of its trochlea are more similar (the medial rim is still more rounded than the lateral one). These derived features make the astragalus of Ptilocercus much more primate-like than all the plesiadapiform astaragali described so far (from the first drawings of Carpolestes published by Bloch and Boyer, 2002, it seems that this fossil too has an astragalus more plesiadapiform-like than primate-like). Why should these characters not be considered shared derived
Figure 6. Tarsals of archontans drawn at the same antero-posterior length. Calcanei in dorsal views (A-D) and astragali in dorsal (E-H) and ventral (I-L) views. Nannodectesgidleyi (A, E, and I), Ptilocercus lowii (B, F, and J); eosimiids (C, G, and K), Notharctus (D), and Teilhardina (H and L). Redrawn from different authors and not entirely accurate (different orientation in g, some facets partially drawn because not clearly delineated on photographs). Note especially that Ptilocercus (B) has a peroneal tubercle in a proximal position as in primates, a high and wedge-shaped astra-galar trochlea (F), and a long neck (F and J), whereas Nannodectes has a short astragalar neck (E), and a sustentacular facet distinct from the navicular (I).
with primates? In Ptilocercus, the astragalar sustentacular facet is continuous with the navicular facet, whereas on several plesiadapiforms there is a clear discontinuity between the facets (on others this continuity does occur). Another interesting fact is that in distal view, the astragalar head is mediolaterally elongated in several plesiadapiforms; it is less elongated, more ovoid, in some other plesiadapiforms and Ptilocercus. Whereas the astragalar head in Ptilocercus is less spherical than in primates, its morphology appears intermediate in form [this would deserve quantification; I note that some of the Shanghuang astragali have heads that are less spherical than in other primates (Gebo et al., 2001; and this could be a primitive state)]. The last intriguing character seen on Szalay and Drawhorn's Figure 9 (1980), and which was observed on one specimen, is a clearly wedge-shaped astragalar trochlea in dorsal view. This character was considered primate-like by Szalay and Lucas (1996), who proposed its interpretation in the framework of the grasp-leaping theory. The distal broadening of the trochlea would be well-suited to the transmission of stress during landing after a leap, with dorsiflexed feet. However, Ptilocercus is not a leaper, and there is no reason to believe that it could have inherited the character from a leaping ancestor. Gebo et al. (2001) link a wedge-shaped trochlea with enhanced dorsiflexed foot positions and greater use of vertical supports, something which is more in line with the behavior of Ptilocercus, which "spend relatively more time [than other sym-patric tree shrews] on large vertical supports" (Emmons, cited in Stafford and Thorington, 1998). The wedge-shaped trochlea could also be related to some foot rotation in conjunction with flexion-extension, as is recognized for monkeys having such a trochlea, and linked with Ptilocercus' frequent use of hal-lucal opposition (Sargis, 2001). In any case, for this character as well, Ptilocercus is remarkably primate-like. There are still differences between Ptilocercus and primates, including an astragalus with a less spherical head, with a lower medial side of the body, possibly a more elongate ventral groove for the tendon of M. flexor fibularis, and other differences that might be more or less accentuated depending on how one reconstructs the primate morphotype (shallow astragalar facet, small or absent posterior trochlear shelf, etc.; see below).
On the calcaneum of Ptilocercus, the peroneal process is reduced in comparison with that of early archontans (Szalay and Drawhorn, 1980). It is also more proximally placed, being at the level of the posterior astragalar facet (Figure 6). This position is remarkably primate-like and unlike the distal position of the peroneal process of plesiadapiforms and many other mammals. Dagosto (1988) noted that "The change in placement of this tubercle has been related to the elongation of the tarsus (Decker and Szalay, 1974), and the shift in the primary function of this muscle from a foot evertor to an hallucal adductor (Gebo, 1986; PhD), but there is as yet no satisfactory explanation for its reduction in size [in primates]." The presence of a reduced proximal process in Ptilocercus suggests that the correlation with tarsal elongation is faulty. Given the hallucial grasping behavior of Ptilocercus (Sargis, 2001), it appears that M. peroneus longus may be a hallucal adductor in these animals. The reduction in the size of the peroneal process in primates may well be partly inherited from a ptilocercine-like ancestor; however, further reduction was manifested in the subsequent primate ancestral morphotype (here too it is interesting to note that some of the Shanghuang calcanei attributed to a new taxon of protoanthropoids have a peroneal tubercle varying from small and moderate (mostly) to "prominent" in one specimen: a primitive retention?). The calcaneocuboid joint of Ptilocercus is a circular pivot (Sargis, 2002b; observed by the author). Because primates typically have a pivot joint, this provides a tempting Ptilocercus-primate synapomorphy. However, we must remember that the calcaneocuboid joint has changed a lot during its evolutionary history, and therefore, a more detailed analysis of the relevant morphology is required before homologous stages can be infered (direct historical evidence might be necessary).
The above list of derived similarities of ptilocercine tupaiids with primates is impressive. Several of them either appear related to enhanced subtalar mobility or to enhanced hallucial opposition, which makes Ptilocercus functionally intermediate with primates (Sargis, 2001). In view of the value of tarsal characters for phylogeny reconstruction, the above-mentioned derived similarities of Ptilocercus and primates make a strong case in favor of their close phylogenetic affinity, to the exclusion of known plesiadapiforms. A similar conclusion was drawn by Hooker (2001) in his analysis of the tarsal characters of the archontans, Deccanolestes, and nyctitheriids; however, this author lost this signal by introducing dental characters in the same analysis. The lengthening of the astragalus, associated with a tall medial body and concomitant reduction of the peroneal tubercle signals an interesting tarsal transformation, possibly associated with frequent hallucal opposition (still far from the powerful hallucal-grasping of primates). This set of characters deserves further functional scrutiny.
Was this article helpful?