Second triconodont dentary from the Early Cretaceous of Maryland

Abstract

Geosciences Department, Towson University, Towson, Maryland 21252Previous authors have noted the general scarcity of Early Cretaceousvertebrate fossils, and mammals in particular (e.g., Clemens et al.,1979). The Arundel Clay facies of the Patuxent Formation (PotomacGroup), Maryland, is one of a small number of Early Cretaceous unitsthat has produced a vertebrate assemblage, and is the only vertebrate-producing unit of this age in eastern North America (Kranz, 1996,1998). Isolated bones and teeth of dinosaurs have been known from theArundel for more than a century (e.g., Marsh, 1888; Lull, 1911; Gil-more, 1921; Lipka, 1996, 1998; Chinnery et al., 1998), but only veryrecently did the deposit yield mammalian remains, a dentary of a newtriconodont, Arundelconodon hottoni (Cifelli et al., 1999). Here we re-port a second mammal specimen from the Arundel Clay. The mor-phology of the dentary and its alveoli indicate that it also represents atriconodont, and it is tentatively referred to A. hottoni for reasons dis-cussed below.Triconodonts are primitive mammals distinguished by their simple,laterally compressed molars with three mesiodistally aligned primarycusps. This general molar structure provided the basis for recognitionof the Order Triconodonta (Simpson, 1928a). The serially tricuspatemolar pattern, long known for Amphilestidae and Triconodontidae (pre-viously placed in the same family; Simpson, 1945), later proved to beshared by other Mesozoic mammal groups, such as Morganucodontidae,Sinoconodontidae, Gobiconodontidae, and Austroconodontidae, whichwere accordingly accommodated into the Triconodonta (Patterson andOlson, 1961; Kermack et al., 1973; Jenkins and Crompton, 1979; Bon-aparte, 1990). However, this cusp pattern is plesiomorphic, being sharedby certain non-mammalian cynodonts (Sigogneau-Russell and Hahn,1994). Sinoconodontidae and Morganucodontidae are now widely re-garded to be among the most plesiomorphous mammals or near mam-mals (mammaliaforms) (Rowe, 1988; Crompton and Luo, 1993; Mc-Kenna and Bell, 1997). On the other hand, cranial and postcranial anat-omy of some later triconodonts is rather advanced. Gobiconodon, Je-holodens, and an undescribed triconodontid, for example, haveunexpectedly derived, therian-like features of the shoulder girdle (Jen-kins and Crompton, 1979; Jenkins and Schaff, 1988; Ji et al., 1999).Recent analyses suggest that the expanded Triconodonta is based onplesiomorphy; and even with the exclusion of Sinoconodontidae andMorganucodontidae, monophyly of remaining triconodonts, ‘‘amphiles-tids’’ and Triconodontidae, remains uncertain (Kermack et al., 1973;Rowe, 1988, 1993; Rougier et al., 1996; Ji et al., 1999). Fortunately,the Triconodontidae, to which the new specimen is referred, appear tobe well-supported by synapomorphy (Jenkins and Crompton, 1979; Ci-felli et al., 1998). Triconodontidae, which range from Late (or perhaps?Early–Middle) Jurassic through Late Cretaceous, are the most diversetriconodonts, with some eight described genera (Table 1).SYSTEMATIC PALEONTOLOGYT