842 Journal of Vertebrate Paleontology 23(4):842–856, December 2003 q 2003 by the Society of Vertebrate Paleontology FIRST DEFINITIVE RECORD OF MESOZOIC LIZARDS FROM MADAGASCAR DAVID W. KRAUSE 1 , SUSAN E. EVANS 2 , and KE-QIN GAO 3 1 Department of Anatomical Sciences, Stony Brook University, Stony Brook, New York 11794-8081, U.S.A.; 2 Department of Anatomy and Developmental Biology, University College London, Rockefeller Building, University Street, London WC1E 6JJ, United Kingdom; 3 Division of Paleontology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024-5192, U.S.A. ABSTRACT—We report here the first unequivocal record of a pre-Late Pleistocene lizard from the island of Mada- gascar, based on a nearly complete lower jaw, elements of both the pectoral and pelvic girdles, several vertebrae and ribs, and numerous osteoderms of what is presumed to be a single individual. The specimen, recovered from the Upper Cretaceous (Maastrichtian) Anembalemba Member, Maevarano Formation, Mahajanga Basin, northwestern Madagascar, is identified as a scincoid scincomorph and, more specifically, a new genus and species of ?Cordylidae (Cordyliformes), based on a combination of gnathic, dental, and postcranial characters. The new taxon is the first identifiable lizard from the Late Cretaceous of the African continent (sensu lato). If the new taxon is correctly attributed to the Cordylidae, it constitutes a significant temporal and geographic range extension for the clade since cordylids have no definite representatives in the fossil record and extant forms are restricted to the sub-Saharan portion of mainland Africa. This new record also indicates that cordylids, after their occurrence in the Maastrichtian, became extinct on Madagascar, leaving only zonosaurine Gerrhosauridae as extant representatives of Cordyliformes on the island. Owing to limited knowledge concerning the time of divergence for cordylids and gerrhosaurids relative to the tectonic separation of Africa and Madagascar, and in light of the paucity of Mesozoic lizard fossils in general, and from Gondwana in particular, the discovery of the new taxon in the Late Cretaceous of Madagascar does little to otherwise constrain scenarios concerning the biogeographic history of early cordyliforms. Finally, we document here the observation that lizards appear to have been much less speciose than snakes in Late Cretaceous faunas of Gondwana, whereas the reverse is true in Laurasia. Lizards do not appear to become common in Gondwana until the Early Tertiary. INTRODUCTION Lizards are the most speciose group of terrestrial vertebrates on the island of Madagascar, with approximately 250 species already known and more being discovered and described an- nually; more than 75 have been described since 1960 (e.g., Bauer, 1993; Glaw and Vences, 1994, 2000; Henkel and Schmidt, 2000; Andreone and Greer, 2002). Furthermore, Mad- agascar boasts more species of lizards than any country in mainland Africa, despite the fact that there are at least 19 such countries larger in area, some of them over four times larger (Bauer, 1993). The extant lizard fauna of Madagascar includes chamaeleonids (almost one-half of the world’s species), oplurid iguanians, scincids, gerrhosaurids, and gekkonids; typical main- land African forms (agamids, lacertids, varanids) are absent. Approximately 95% of the species are endemic to the island (Uetz and Etzold, 1996; Glaw and Vences, 2000). The biogeo- graphic origins of the Malagasy lizard fauna and, in particular, the oplurids, the extant relatives of which are otherwise known only from the Americas (as well as the Fiji and Tonga Islands), are shrouded in mystery, in large part owing to the nearly com- plete absence of a lizard fossil record from Madagascar. To our knowledge, the only undisputed records of fossil lizards from the island were noted by Estes (1983a:57–58, 70), who cited the presence of isolated jaw fragments of Chamaeleonidae in- certae sedis from ?Pleistocene and ?Holocene horizons and Geckolepis sp. from the Holocene. Sabatier and Legendre (1985) reported the occurrence of indeterminate squamates in a fauna listed as Plio-Pleistocene, but there is no convincing evidence for an age older than Late Pleistocene (K. Samonds, pers. comm., 2001). Here we report the discovery of the first undoubted pre-Late Pleistocene specimen of a fossil lizard from the island of Madagascar. During the inaugural field season of the Stony Brook Uni- versity/University of Antananarivo Mahajanga Basin Project, on August 15, 1993, University of Antananarivo graduate stu- dent Joseph Augustin Rabarison made an important discovery, a jumbled mass of approximately five partially articulated skel- etons of a new species of the small crocodyliform Araripesu- chus at locality MAD93-33. Locality MAD93-33 occurs in the Berivotra field study area of northwestern Madagascar and, stratigraphically, in the Upper Cretaceous (Maastrichtian) Anembalemba Member of the Maevarano Formation (Rogers et al., 2000). Owing to the distribution of bones, the rock that entombed the skeletons of Araripesuchus was divided in the field and collected in two separate plaster jackets; laboratory preparation of the two blocks has proceeded intermittently ever since. On August 31, 2001, Ms. Virginia Heisey, a technician at Stony Brook University, discovered the lizard dentary de- scribed in this report among the bones of Araripesuchus in one of the blocks. Subsequent preparation and inspection of the block revealed numerous other elements of a lizard, all distrib- uted in a small area (Fig. 1). The relatively small size and distinctive nature of the bones, the absence of duplication of elements, and their discovery in very close proximity to one another strongly suggest that they derive from a single individ- ual. Although the distribution of bones in the block provides no clear indication, it is possible that the lizard specimen com- prised the gut contents of one of the Araripesuchus individuals. The objective of this report is to describe this important new lizard specimen, determine as precisely as possible the phylo- genetic relationships of the taxon it represents, and discuss the implications of its discovery for the paleobiogeography of Gondwanan lizards in general and for the biogeographic origins of the extant Malagasy lizard fauna in particular.