Marsupials and Their Relatives
THE TERM METATHERIA IS USED to unite marsupials and their presumed extinct relatives, including the Deltatheroida and the Asiadelphia. Although generally overshadowed by placental mammals, marsupials persist today, and during the Cenozoic they underwent diverse radiations in South America and Australia, where they still predominate. The oldest known metatherians are from the Cretaceous. Deltatheroidans are a largely Asian clade restricted to the Cretaceous. They are generally considered to be the sister group of Marsupialia, or of a clade of marsupials and other primitive metatherians. Alternatively they could be the sister taxon of all living therians (i.e., eutherians + marsupials; Luo et al., 2002). Asiadel-phians, based primarily on the Asian genus Asiatherium, are variously considered to be another branch of metatherians, the sister group of Marsupialia, or a primitive clade of marsupials. All of these groups primitively share the postcanine dental formula of three premolars and four molars and have upper molars with a wide stylar shelf and one or more stylar cusps, but no hypocone. Some later metatherians, however, especially Australian clades, evolved a cusp in the position of a hypocone (probably a displaced metaconule).
The fossil record indicates that metatherian and eutherian mammals had already diverged by early in the Cretaceous (Cifelli, 1993a; Eaton, 1993). Some molecular studies suggest an even earlier split, in the Jurassic. Most of the Mesozoic metatherian clades became extinct by the end of the Cretaceous. A number of recent discoveries have greatly expanded our knowledge of these primitive metatherians.
Based on some molecular studies, it has been suggested that marsupials are the sister taxon of monotremes (e.g., Janke et al., 1997, 2002). These authors resurrected
W K. Gregory's abandoned term Marsupionta (subsequently shown to be based on shared primitive characters) for this supposed clade. However, as detailed by Luo et al. (2002), substantial anatomical evidence and even most molecular data indicate that metatherians are more closely related to eutherians than to any other mammals.
Extant metatherians (marsupials) differ from eutherians in many ways, the most obvious of which concern reproductive anatomy and development: the female reproductive tract is bifid; the gestation period is very short (8-42 days; Moeller, 1990); the young are altricial, and organ systems and limbs may be only partly developed at birth; most species have a pouch or marsupium, where the young complete their development (this external "womb" explains Linnaeus's name for the opossum Didelphis, "double womb," although it might also be a reference to the double uterus); and epi-pubic bones project forward from the pubic bones (associated with locomotion and/or support of the pouch or of the developing young). Since, apart from epipubic bones (which are also known to be present in monotremes, multituber-culates, and basal eutherians), these traits are not preserved in fossils, how can metatherians be recognized in the fossil record? Fortunately, there are several characters of the skull and teeth that separate most early marsupials from placen-tals. They include an auditory bulla composed primarily of the alisphenoid; large openings, or vacuities, in the palate; an inflected angular process of the dentary (possibly a retained primitive trait); more upper than lower incisors; simple tri-bosphenic upper molars lacking a hypocone, but with a wide stylar shelf bearing multiple cusps; three simple premolars followed by four molars, the last premolar being the only tooth replaced; and lower molars often with an unreduced paraconid and twinned hypoconulid and entoconid (see Figs. 2.1, 5.1). But, as in other evolutionary transitions, anatomical features were acquired sequentially, and the most primitive forms, transitional between "tribotheres" and meta-therians, lack some of these "diagnostic" traits. For example, some Cretaceous teeth that are otherwise marsupial-like have poorly developed stylar cusps or lack twinning of the hypoconulid and entoconid.
Dental anatomy plays a prominent role in identifying the oldest metatherians. The disposition of stylar cusps on the upper molars has been considered to be particularly significant. The stylar cusps of metatherians are typically designated by the letters A through E, starting at the front of the molar (Fig. 5.1). As in eutherians, cusp A is identified as the parastyle, cusp C the mesostyle, and cusp E the metastyle. Cusp B, which is joined to the paracone by the paracrista, is also called the stylocone, whereas cusp D has no other designation. It should be realized, however, that there is little evidence that these cusps are homologous with the stylar cusps of eutherians; hence, these designations are primarily topographic. Indeed, some of these cusps probably arose multiple times within marsupials. Cusps B and D are often larger than the others (e.g., in didelphoids), and not all cusps are present in most forms.
hypoconulid LEFT LOWER MOLAR
Fig. 5.1. Terminology for metatherian molars. (From Marshall, 1987.)
Recent discoveries suggest that postcranial features, particularly the carpus and tarsus, may also be important in distinguishing the earliest metatherians and eutherians (Luo et al., 2003).
Many different classifications of marsupials have been proposed over the past 20 years or so, but none seems to have achieved the level of consensus. (The classification used here is presented in Table 5.1.) There is even substantial disagreement as to the higher-level assignments of many non-Australian families. But most current authorities agree that marsupials comprise multiple ordinal-level taxa, in contrast to the single order Marsupialia widely used a generation ago. In addition, there is agreement that, except for some Cretaceous forms (sometimes placed in a separate clade, Al-phadelphia), there is a basic dichotomy between New World and Australian marsupials (cohorts or magnorders Ameri-delphia vs. Australidelphia; Szalay, 1982, 1994). Figure 5.2 depicts one interpretation of relationships that illustrates this dichotomy, but it differs in many details from the classification used in this chapter. It is also evident that Australian marsupials evolved from an American form, most likely a microbiothere. A modified version of McKenna and Bell's (1997) classification is used as a framework here. In this scheme, Ameridelphia includes didelphid opossums and other didelphimorphs, as well as paucituberculates and
Table 5.1. Classification of Metatheria
Was this article helpful?