ARTICLE ENDOCRANIAL MORPHOLOGY OF THE URSUS DENINGERI VON REICHENAU 1904 FROM THE SIMA DE LOS HUESOS (SIERRA DE ATAPUERCA) MIDDLE PLEISTOCENE SITE NURIA GARCI ´ A, *,1,2 ELENA SANTOS, 3 JUAN LUIS ARSUAGA, 1,2 and JOSE MIGUEL CARRETERO 1–3 1 Dpto de Paleontologı ´a. Facultad de Ciencias Geolo ´ gicas, Universidad Complutense de Madrid, Ciudad Universitaria s/n, 28040, Madrid, Spain, ngarcia@isciii.es, jlarsuaga@isciii.es; 2 Centro Mixto UCM-ISCIII de Evolucio ´ n y Comportamiento Humanos, C/ Sinesio Delgado 4, Pab 14, 28029, Madrid, Spain; 3 Laboratorio de Evolucio ´ n Humana, Dpto de Ciencias Histo ´ricas y Geografı´a, Edificio I+D+i, Universidad de Burgos, Plza Misael Bañuelos s/n, 09001, Burgos, Spain, esantos@beca.ubu.es, jmcarre@ubu.es ABSTRACT—A complete skull of Ursus deningeri recovered from the Middle Pleistocene site of the Sima de los Huesos in the Sierra de Atapuerca (Spain) was studied with computed tomography. This technique makes it possible to study new and relevant information in the endocranial structures and avoids any damage to the specimen. The previously unde- scribed endocranial cavity of U. deningeri is compared with related species such as Ursus spelaeus Rosenmüller 1794 and Ursus arctos Linne ´ 1758. The characters analyzed include the braincase morphology and also quantitative and volumetric measurements, which are only attainable with this technique. In particular, the volume, surface area, and form of the frontal sinuses and brain cavity, the shape of the ethmoid bone, and the basioccipital pneumatization were studied. The resulting analysis supports an ancestor-descendant relationship for the two ursids included in the cave bear phylogenetic lineage (U. deningeri and U. spelaeus). In contrast, U. arctos exhibits a different endocranial morphology, one that seems to be mainly plesiomorphous. Finally, this study has identified several new endocranial characters which are useful in reconstructing ursid phylogeny. INTRODUCTION In the mid-1970s and 1980s, new methods previously restricted to the fields of engineering and medicine were applied for the first time in palaeontology (Lovejoy and Burnstein, 1977; Martin and Atkinson, 1977; Jungers and Minns, 1979; Ruff and Jones, 1981; Conroy and Vannier, 1984; Wind, 1984; Kennedy, 1985; Ruff and Leo, 1986). The use of high-resolution computed to- mography (CT) allows for the non-destructive analysis and mea- surement of internal structures and cavities in fossil remains (Wind, 1984; Seidler et al., 1997; Joeckel, 1998; Thompson and Illerhaus, 1998; Brochu, 2000; Zollikofer and Ponce de Leo ´ n, 2000; Marino et al., 2001; Domı´nguez et al., 2002, 2004; Rae et al., 2002; Weaver, 2002; Bruner et al., 2003; Clack et al., 2003; Milner et al., 2003). The use of CT has become increasingly feasible and refined as an analytical tool for examining endocra- nial anatomy in fossil vertebrates. Recent research has been car- ried out on the internal morphology and dimensions in extant and fossil anthropoids (Tate and Cann, 1982; Conroy and Van- nier, 1984, 1986; Conroy et al., 1990; Spoor et al., 1993, 1994; Spoor and Zonneveld, 1995; Hublin et al., 1996; Seidler et al., 1997) and other vertebrate groups (Rowe et al., 1995, 1997; Joeckel, 1998), and this tool is particularly useful for reconstruct- ing the virtual endocast in well preserved fossil skulls (Conroy et al., 1990; Rowe et al., 1995; Zollikofer et al., 1995, 1998; Seidler et al., 1997; Thompson and Illerhaus, 1998; Brochu, 2000; Weaver, 2002; Bruner et al., 2003; Bruner, 2004; Franzosa and Rowe, 2005). The Sima de los Huesos (SH) site is located deep within the Cueva Mayor–Cueva del Silo cave system (Sierra de Atapuerca, Spain), far removed from any modern day surface entrance. The site has yielded the well-preserved skeletal remains of dozens of human individuals (Arsuaga et al., 1993, 1997 and references therein) lying stratigraphically below a jumble of bones of the Middle Pleistocene cave bear Ursus deningeri (Garcı´a et al., 1997). No herbivores are present among the faunal assemblage from the site, contrasting with their common occurrence at homi- nid occupation sites. Bear and human bones were directly dated by the combined ESR and U-series methods (Bischoff et al., 1997), and yielded ages that ranged between 200 ± 4 ka and 320 ± 4 ka. Samples from a newly exposed in situ speleothem at the site were analyzed by the high resolution U-series technique, and yielded a minimum age of 550 ka (Bischoff et al., 2007). Imme- diately below this speleothem, bones belonging to bears and human remains were also recovered. Thus, the radiometric dates confirm a mid-Middle Pleistocene age for the SH fossils. Taxo- nomic assignment of the SH ursids is based on abundant post- cranial and dental remains recovered during several campaigns (Torres, 1977). The evolutionary stage of different elements (mandible, dentition and metapodials) of U. deningeri from the SH site, does not correspond to transitional forms of U. spelaeus (Bonifay and Bussière, 1989), but, rather, resembles more closely the morphology seen in mid-Middle Pleistocene representatives of this species of ursid (Garcı´a et al., 1997; Garcı´a, 2003). Due to their fragile nature, complete cranial remains are very rare, which makes the recovery of a very well-preserved skull from an ursid at SH an extraordinary discovery and a valuable contribution for research on the evolution of the cave bear lin- eage. A study of the endocranial cavities of this specimen was undertaken through CT scanning and virtual reconstruction. The SH skull was compared with different extant and fossil ursid * Corresponding author: N. Garcı´a, Centro Mixto UCM-ISCIII de Evolucio ´ n y Comportamiento Humanos, C/ Sinesio Delgado 4, Pab 14, 28029, Madrid, Spain, ngarcia@isciii.es Journal of Vertebrate Paleontology 27(4):1007–1017, December 2007 © 2007 by the Society of Vertebrate Paleontology 1007