Phylogenetic position of Turanoceratops (Dinosauria: Ceratopsia)

Abstract

Farke et al. (2009) commented on our recent paper describing new material of the Late Cretaceous ceratopsian dinosaur Turanoceratops from Uzbekistan (Sues and Averianov 2009). We provided evidence that Turanoceratops is the first member of the clade Ceratopsidae known outside North America. Farke et al. (2009) coded Turanoceratops using a character-taxon matrix for Ceratopsidae by Dodson et al. (2004) and found its phylogenetic position as a sister-taxon to Ceratopsidae rather than as a member of that clade. Their coding is as follows: 21(1), 22(0), 25(0), 29(0), 57(0), 60(1), 61(1) and 62(0). Our coding of Turanoceratops, based on direct observation of the specimens, is 21(1), 22(0), 23(0), 25(0), 29(1), 30(1), 57(0), 60 (1), and 62(1). We specifically disagree on the coding of four characters: Character 23: postorbital, position of supraorbital ornamentation: centered rostrodorsal or dorsal to orbit (0); centered caudodorsal or caudal to orbit (1). This character was not coded by Farke et al. (2009). The supraorbital horn core is situated rostrodorsal to the orbit on the postorbital ZIN PH 1868/16. Character 29: supracranial cavity complex: absent (0); present, supracranial cavities narrow and shallow, do not underlie supraorbital ornamentation (1); present, supracranial cavities broad, underlie supraorbital ornamentation and may be confluent with extensive cornual sinuses (2). Farke et al. (2009) interpreted Turanoceratops as lacking a supracranial cavity complex. They believed that the feature identified by us as “frontal sinus” likely represents the ventral surface of the postorbital. However, this interpretation is anatomically impossible: ZIN PH 1868/16 preserves the orbital surface on the opposite side and the smooth area identified by us as the frontal sinus is on the dorsal side of the bone. Turanoceratops had a shallow supracranial cavity that does not underlie the supraorbital horns (character 29, state 1). Character 30: frontal, contribution to orbital margin: present (0); absent (1). Farke et al. (2009) did not code this character. Our coding is based on the articular facet for the prefrontal on the postorbital ZIN PH 1868/16, suggesting that the frontal was excluded from the orbital margin. Character 62: tooth ornamentation: subsidiary ridges present, extend from margin to base of tooth (0); subsidiary ridges reduced, present only at margin of teeth (1). Although Turanoceratops has prominent subsidiary ridges on the teeth, none of these ridges reaches the base of the tooth in our sample comprising several dozen teeth. Thus, Turanoceratops should be coded for the derived state (1) of this character. Its dentition is truly transitional between basal neoceratopsians and more derived ceratopsids. We analyzed the modified data matrix using NONA version 2.0 (Goloboff 1999) with the WinClada version 1.00.08 interface (Nixon 1999). One thousand repetitions of the parsimony ratchet (island hopper) algorithm generated Naturwissenschaften (2009) 96:871–872 DOI 10.1007/s00114-009-0552-7