Didelphodonta And Other Primitive Cimolesta

The most plesiomorphic group of the Cimolesta, Didelphodonta (as employed by McKenna and Bell, 1997) comprises the single family Cimolestidae, which is widely considered on dental criteria to include the ancestors or sister taxa of most or all of the other groups these authors assigned to Cimolesta. Didelphodonts also seem to be related to the early Tertiary families Palaeoryctidae and Wyolestidae, as well as to the origin of creodonts and carnivores. These inferred relationships are reflected by

Table 7.1. Synoptic classification of Cimolesta

Superorder FERAE

Mirorder fCIMOLESTA

fDidymoconidae1 fWyolestidae1 Order fDIDELPHODONTA f Palaeoryctidae1 fCimolestidae Order fAPATOTHERIA

fApatemyidae Order fTAENIODONTA fStylinodontidae Order fTILLODONTIA

fEsthonychidae (=fTillotheriidae) Order fPANTODONTA fHarpyodidae fBemalambdidae fPastoralodontidae fPantolambdidae fTitanoideidae fBarylambdidae fCyriacotheriidae fPantolambdodontidae fCoryphodontidae Order fPANTOLESTA fPantolestidae fPentacodontidae fParoxyclaenidae fPtolemaiidae Order PHOLIDOTA fEomanidae fPatriomanidae Manidae ?Order PHOLIDOTA

Suborder fPALAEANODONTA fEscavadodontidae fEpoicotheriidae fMetacheiromyidae Suborder fERNANODONTA fErnanodontidae

Notes: Modified after McKenna and Bell (1997). McKenna and Bell consider Cimolesta to be an order, and the orders listed here to be suborders. The dagger (f) denotes extinct taxa. Families in boldface are known from the Paleocene or Eocene.

1 Relationships uncertain.

grouping all these taxa in the grandorder Ferae (McKenna and Bell, 1997).

Traditionally the genera now assigned to Cimolestidae were included in a broadly conceived family Palaeoryctidae, but current usage usually restricts the latter family to the most specialized zalambdodont forms. Members of both families, sometimes loosely called "palaeoryctoids," are generally characterized by cheek teeth with high, sharp cusps; transversely wide upper molars with a broad stylar shelf but lacking pre- and postcingula and hypocones; lower molars with tall trigonids and reduced talonids; and an emphasis on transverse shearing of the front and back of the trigonid against the posterior and anterior crests, respectively, of successive upper molars (Novacek, 1986a; Figs. 7.1, 7.2). Cimo-lestids tend to have a primitive dental formula of 3.1.4.3/ 3.1.4.3, whereas palaeoryctids lack the first premolar, so the dental formula is typically 3.1.3.3/3.1.3.3. Late Cretaceous

Asioryctes, which was originally assigned to the Palae-oryctidae, resembles palaeoryctoids in some dental and cranial features, but most of these features are plesiomorphic. Asioryctes is currently placed in the basal eutherian taxon Asioryctitheria.

Cimolestids are thought to retain one of the most ple-siomorphic dentitions among placentals. Thus it is reasonable to place them in a basal position to many other groups; nevertheless, few synapomorphic features link them specifically with any later taxa. Most cimolestids were approximately the size of living hedgehogs. Unfortunately, little is known about them beyond the dentition. Cimolestes, best known from the latest Cretaceous and early Paleocene of western North America, is generally considered the most primitive and is the only Mesozoic cimolestid genus. Its

Hedgehog Dentition
Fig. 7.1. Dentitions of didelphodonts: (A) Procerberus, left P4-M3; (B) Cimolestes, left P4-M2 and P2-M3; (C) Didelphodus, left upper and lower dentitions. (A from Lillegraven, 1969; B from Clemens, 1973; C from Matthew, 1918.)
Palaeoryctes
Fig. 7.2. Palaeoryctid dentitions: (A) Aaptoryctes, right upper and lower dentitions; (B) Palaeoryctes skull (ventral view); (C) Palaeoryctes right lower dentition. (A from Gingerich, 1982; B from McDowell, 1958; C from Bloch et al., 2004.)

dentition is characterized by simple premolars, the lower ones dominated by one main cusp, and molars with a generalized tribosphenic pattern. The lower molars have tall trigonids and small, shallowly basined talonids with three cusps. The uppers have three main cusps, small conules, no hypocone, and a broad stylar shelf with few cusps. Based on these dental traits, Cimolestes has been considered to be the sister group and possible ancestor of both creodonts and carnivorans (see Chapter 8). Therefore it is almost certainly paraphyletic, and the variation encompassed by its several species might be more appropriately separated into more than one genus.

Several other cimolestid genera are known from the Paleocene and Eocene of North America and Europe, as well as the late Paleocene of Africa and the early Eocene of Asia. Early Paleocene Procerberus of western North America is currently considered a cimolestid because of various dental features shared with Cimolestes (Lillegraven, 1969; McKenna and Bell, 1997). But it also resembles leptictids in having sub-molariform last premolars and in its molar wear pattern, hence its family assignment is debatable (Novacek, 1986a). One of the best-known cimolestids is Didelphodus (early to middle Eocene of North America and Europe), which differs in only minor ways from Cimolestes (e.g., having somewhat lower trigonids and broader stylar shelves).

Palaeoryctidae, known from the Paleocene and Eocene of western North America and possibly northern Africa and Eurasia, were tiny shrew- to mouse-sized insectivores with zalambdodont or nearly zalambdodont molars (Figs. 7.2, 7.3). Current usage restricts the family to Palaeoryctes, Eoryctes, Aaptoryctes, Ottoryctes, and perhaps one or two other genera (MacPhee and Novacek, 1993; McKenna and Bell, 1997; Bloch et al., 2004). Palaeoryctes minimus from the late Paleo-cene of Morocco is among the smallest known mammals, with molars less than 1 mm long (Gheerbrant, 1992). The premolars of palaeoryctids are typically simple, tall, and sharp, except in Tiffanian Aaptoryctes, in which the last pre-molars are enlarged and swollen and have blunt cusps for crushing (Gingerich, 1982). Palaeoryctid upper molars are strongly transverse, with connate (twinned) paracone and metacone usually set well in from the buccal margin; a high, constricted protocone; and often a prominent parastyle. There are deep embrasures between the upper cheek teeth, which accommodate the very high trigonids of the lower molars; the talonids are typically small or vestigial. These features were presumably associated with a diet of insects and other small invertebrates.

Several well-preserved skulls of palaeoryctids are known, which show that palaeoryctids had rather short snouts, a bony auditory bulla, and poorly developed zygomatic arches or none at all. The squamosal in Eoryctes and Ottoryctes is expanded on each side into a broad lambdoidal plate similar to that in apternodontids (Bloch et al., 2004; see the section on Soricomorpha in Chapter 9). The bulla consisted of the ectotympanic in Palaeoryctes (Butler, 1988) but was possibly derived from the petrosal and/or the entotympanic in Wasatchian Eoryctes (Thewissen and Gingerich, 1989). In either case it differs from lipotyphlans, which typically have a basisphenoid bulla or lack an ossified bulla (Novacek, 1986a). In Eoryctes and Ottoryctes, the promontorial and stapedial branches of the internal carotid artery (inside the auditory bulla) were enclosed in bony tubes, whereas in Palaeoryctes grooves (but no tubes) were variably present (McDowell, 1958; Van Valen, 1966; Thewissen and Gingerich, 1989; Bloch et al., 2004). The significance of these arterial tubes, however, is unknown.

Palaeoryctids have been considered to be related either to genera now classified as cimolestids (Van Valen, 1967; Butler, 1988; McKenna and Bell, 1997) or to lipotyphlans (Lillegraven et al., 1981), specifically, soricomorphs (McKenna et al., 1984). Association with soricomorphs is based partly

on mutual presence of a piriform fenestra in the skull (an opening in the bony roof of the middle ear), but this feature is of questionable significance (Butler, 1988; MacPhee and Novacek, 1993). Palaeoryctids share certain other cranial features with various lipotyphlans but differ from them in other ways (as noted above), leaving their relationship to Lipo-typhla uncertain: they are either related or evolved conver-gently, probably from a Cimolestes-like form (Butler, 1988). Although palaeoryctids are assigned here to Cimolesta, following McKenna and Bell (1997), a relationship to Lipo-typhla seems equally likely, based on current understanding.

Sometimes associated with palaeoryctids are the Early Tertiary Micropternodontidae. They are alternatively grouped with soricomorph lipotyphlans, and in the present account they are covered under that heading in Chapter 9.

Knowledge of cimolestids and palaeoryctids is largely limited to teeth and a few partial skulls; only very fragmentary or isolated postcrania (often attributed without dental association) are known. Unfortunately, this paucity of material severely limits our ability to reconstruct their behavior. A fragmentary humerus associated with Palaeoryctes suggests that it was an efficient digger (Van Valen, 1966). Isolated ankle bones attributed to Cimolestes and Procerberus (Fig. 7.4) share derived features with leptictids that imply abilities for plantar flexion and eversion typical of terrestrial mammals (Szalay and Decker, 1974). These features include an oblique and inclined posterior calcaneoastragalar facet and a distally placed peroneal tubercle on the calcaneus, an astragalus with a moderately long neck, grooved trochlea, isolated sus-tentacular facet, and no dorsal astragalar foramen. However, this interpretation should be viewed cautiously until confirmed by associated skeletons.

0 0

Post a comment