Other Characters and Conclusion

Progress in the interpretation of paromomyid cranial characters engendered by the discovery of new and well-preserved specimens has been discussed earlier. The discovery of a tympanic ring and an annular bridge in Ignacius (Bloch and Silcox, 2001) lends considerable support to the link between Plesiadapis and Ignacius advocated by Kay et al. (1992). It is thus very likely that, like paro-momyids, plesiadapiforms (excluding microsyopids) are not closely related to dermopterans. However, the more relevant question is if they have a close relationship to primates. Important non-dental support for this hypothesis came from the initial interpretation of the Plesiadapis bulla as petrosal (Russell,

1959; Szalay, 1969, 1975; Szalay et al., 1987)—a view abandoned by Russell (1964) and considered doubtful by others because bullar sutures can fuse early (Gingerich, 1976; MacPhee et al., 1983; MacPhee and Cartmill, 1986). Given these uncertainties for Plesiadapis, and given its above-mentioned similarity with Ignacius, Plesiadapis probably also had an entotympanic bulla (Kay et al., 1992; Wible and Martin, 1993). The case for primate affinities for the plesi-adapiforms is thus considerably weakened.

Other potential cranial plesiadapiform-primate synapomorphies are also debatable. The central position of the promontorium within the auditory bulla is linked to the medial expansion of the middle ear cavity in paro-momyids, plesiadapids, and adapids (Szalay et al., 1987). However, Kay et al. (1992) suggested that this similarity arose through different developmental pathways: plesiadapiforms [in fact paromomyids and not plesiadapids, Beard and MacPhee, 1994] differ from adapids in having a narrow basisphenoid across which the bullae nearly touch, whereas in adapids the basisphenoid is much broader and the two bullae are widely separated (Kay et al., 1992). Whether this really implies convergence for this character might require further scrutiny. The lateral route of the ICA in plesiadapiforms and primates is unusual and significant (Bloch and Silcox, 2001; Wible, 1993). However, there is some variability within primates: Shoshonius has the lemur-like lateral position of the posterior carotid foramen, which leads us to consider this as the primitive pathway in primates (Beard and MacPhee, 1994; Bloch and Silcox, 2001). However, this implies that a reversion to the more primitive location "occurred in omomyids more derived than Shoshonius" (Bloch and Silcox, 2001). In fact, this hypothesis would require two independent reversals in the North American Rooneyia and Omomys (Ross and Covert, 2000) and in European microchoerids. Such a scenario is questionable, and in any case, it shows that this character either did revert or changed convergently in primates, diminishing its phylogenetic value. Several other potential plesi-adapiform-primates synapomorphies were mentioned and subsequently dismissed. A maxillary-frontal contact in the orbit occurs in plesiadapiforms and primates; however, it is not a convincing synapomorphy as it is not unusual among eutherians (e.g., lipotyphlans, rodents, lagomorphs, Wible and Covert, 1987). The ventral shielding of the fenestra cochleae was listed (Szalay, 1975; Szalay et al., 1987); however, MacPhee (1981) suggested that shielding by a caudal tympanic process of the petrosal is primitive rather than derived. The presence of an annular bridge linking the tympanic ring to the bulla wall was considered an important plesiadapid-primate similarity (Cartmill, 1975; Gingerich, 1976); however, this character is also present in tupaiids, and it is not present in all primates, rendering its polarity problematic. I would provisionally follow Beard and MacPhee (1994), who consider a complete annular bridge as primitive for primates, due to its presence in tupaiids, plesiadapids (+ paromomyids), and omomyids. It appears that most of the proposed synapomorphies between plesiadapiforms and primates either have been refuted or appeared problematic. Basicranial features have been the subject of much attention, including remarkable developmental studies. They provide characters very important for the study of primate phylogeny; however, they proved relatively deceptive in the search for archontan phylogeny (e.g., MacPhee, 1981; Starck, 1975; Wible and Martin, 1993). Recent authors observe that the basicranium is not a taxonomic touchstone (Bloch and Silcox, 2001; Wible and Martin, 1993).

If we cannot rely on decisive characters from the basicranium, what would the rest of the cranium suggest? Until now, the general shapes of known plesiadapiform skulls are very unlike those of primates. The skulls of Plesiadapis and Ignacius are primitive in having small laterally directed orbits, a broad interorbital breadth reflecting large olfactory bulbs, a large infraorbital foramen suggestive of important blood supply to the anterior part of the muzzle (probably with well-developed vibrissae), as was inferred in the beautiful study of the skull of Palaechthon nacimienti by Kay and Cartmill (1977). They seem to have had a small brain case in comparison with primates. From the size of the optic foramen, Kay et al. (1992) inferred that Ignacius had eyes similar in size to those of Erinaceus. These characters reflect the absence of any evolutionary step toward the crucial primate visual apomorphies. One can find some isolated apomorphies (e.g., a relative reduction of the infraorbital foramen seen in Plesiadapis), or the beginning of a postorbital process in Palaechthon (palaechthonid), and in the microsy-opid Megadelphus; however, these seem to be of limited significance, and probably not homologous with primate states. Plesiadapis and Ignacius skulls also have cross specializations, such as a tubular ectotympanic, absent in the ancestral primate morphotype. One could guess that some of the cross-specializations seen in later plesiadapiforms could be absent in earlier forms, and they may not therefore disallow a close phylogenetic relationship between plesiadapiforms and primates (Bloch and Silcox, 2001). However, given the absence of any crucial primate-like characters, as those of the orbit, in later forms, one can infer their absence in primitive plesiadapiforms and conclude that there is no strong cranial evidence in favor of a close plesiadapiform-primate phylogenetic relationship. In fact, there are the specializations of the muzzle, narrow and elongated in correlation with the large anterior incisors, which argue strongly against any plesiadapiform being ancestral to primates. If anything, the known skull evidence does not favor plesiadapiform-primate close relationship.

On the other hand, it is possible that a series of derived cranial characters would well support plesiadapiform monophyly, in addition to the dental ones. A series of cranial synapomorphies of plesiadapiforms was provided by Kay et al. (1992), among which several seem to hold: suboptic foramen present, ossified external auditory meatus, and strong mastoid tubercle. There is also a degree of reduction of carotid blood supply to the brain, and probably other common characters to extract from the specializations of Plesiadapis and Ignacius skulls. When known, plesiadapiforms (with the exception of the hallux of Carpolestes) have claws and not nails. Their tarsals do not show a close approximation with those of primates, contrary to those of tupaiids (see below). Perhaps some plesiadapiform postcranial synapomorphies will be found? As stated above, a number of characters formerly believed to be plesiadapiform-primate synapomorphies are now suspected to be archontan. The history of arboreal adaptations is much more complex than previously thought, requiring further analysis before we can delineate which characters retain a high phylogenetic value.

Beyond the restricted plesiadapiforms considered here, several other families should be taken into account in a search to elucidate archontan phylogeny. Microsyopidae, which have a more primitive auditory region (Szalay, 1969) are sometimes mentioned as a primate or dermopteran sister group; Mixodectidae and Plagiomenidae are more commonly suspected to have dermopteran affinities (Szalay and Lucas, 1996); Apatemyidae, with their dental and arboreal specializations, could pertain to the same broader group. How far should such a group be extended in order to include primates and their close relatives? Probably, as far as the living tree shrews, order Scandentia. The broad and not well-delineated Archonta appears then as a group of arboreal eutherians, whose history is certainly very complex and which might even include the Cretaceous Deccanolestes (Prasad and Godinot, 1994). They may also be related to nyctitheres (Hooker, 2001).

0 0

Post a comment