Ontogeny, Diversity, and Systematics of Pachycephalosaur Dinosaurs from the Belly River Group of Alberta

Transcription

1 Ontogeny, Diversity, and Systematics of Pachycephalosaur Dinosaurs from the Belly River Group of Alberta by Ryan K. Schott A thesis submitted in conformity with the requirements for the degree of Master of Science Department of Ecology and Evolutionary Biology University of Toronto Copyright by Ryan K. Schott 2011

2 Ontogeny, Diversity, and Systematics of Pachycephalosaur Dinosaurs from the Belly River Group of Alberta Ryan K. Schott Master of Science Department of Ecology and Evolutionary Biology University of Toronto 2011 Abstract Pachycephalosaur diversity and systematics is poorly resolved, mainly due to the incomplete nature of their remains. The largest sample of pachycephalosaurs is from the Belly River Group of Alberta, but this sample is also the most problematic in terms of alpha taxonomic diversity. Material from this assemblage has been proposed to represent from a single species to up to four distinct genera. Each of these views depends on different interpretations of morphology, ornamentation, ontogeny, and sexual dimorphism, but none have been adequately tested. Here I analyze the diversity of pachycephalosaurs with particular emphasis on Stegoceras validum and 'Prenocephale' brevis. The ontogeny of these species is described using comparative morphology, histology, and morphometrics. New specimens are described and the first detailed pachycephalosaur growth series supported by multiple, independent lines of evidence is presented. Phylogenetic analysis utilizing results from this study provides a new hypothesis for the evolutionary relationships of pachycephalosaurs. ii

3 Acknowledgments I am indebted to all the individuals that have given me the supervision, guidance, and assistance that have made this Master's thesis possible. I am especially thankful to my supervr David Evans, for first taking me in as an undergraduate student and providing me with the excellent projects that have culminated in this thesis. Without his support and guidance this would not have been possible. I am also grateful to my supervry committee, Dan Brooks (cosupervr), Tim Dickinson, and, especially, Mark Goodwin for their insight into my various projects. I am thankful to my external examiners, Deborah McLennan and Jason Head, for their comments and feedback on my work. Special thanks to the members of the Evans Lab, especially Nicolas Campione and Caleb Brown, for discussions and assistance with many aspects of this thesis. I thank Warren Fitch (UCMZ), Brandon Strilisky (TMP), Phil Currie (UALVP), and Kieran Shepherd (CMN), for access to collections and loans of material. Thanks to Caleb Brown for discovering UCMZ (VP) , Miriam Reichel for discovering and preparing UALVP 49531, and Phil Bell for discovering UALVP Special thanks to the artists who provided the excellent scientific illustrations contained in this thesis, Caleb Brown, Johnathan Ho, Kurt Latanville, and Amy Janzen. Thanks to Ian Morrision for preparing UCMZ(VP) and numerous other specimens. I thank Nathan Scaiff for discussions and assistance regarding CT methodologies, Angel Ai for discussion regarding histology, and Bob Sullivan and Tom Williamson for discussions regarding pachycephalosaur morphology and systematics. Thanks to the University of Texas High-Resolution X-ray CT Facility for producing the HRCT images, and Arius 3D for the 3D surface scans and reconstruction of Foraminacephale brevis. Finally I would like to deeply thank my girlfriend, Alicia Dobbs; my parents, Daniel and Patricia Schott; my sisters Nicole and Natalie Schott; my grandparents, Ellen Duffy and iii

4 Karl and Françoise Schott; and my great uncle, Robert Reid; for their love and support throughout my academic pursuits. iv

5 Table of Contents Abstract...ii Acknowledgments...iii Table of Contents...v List of Tables...vi List of Figures...vii List of Appendices...ix Chapter 1: Introduction...1 References...4 Chapter 2: Cranial Ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria) and the evolution of cranial doming and ornamentation in Pachycephalosauria Author Contributions Introduction Systematic Paleontology Description Taxonomic Assessment Frontoparietal Allometry Frontoparietal Bone Histology Squamosal Variation and Ontogeny Discussion...60 References...69 Appendices...77 Chapter 3: Systematics and Ontogeny of Foraminacephale brevis gen. nov. (Ornithischia: Pachycephalosauria) Introduction Taxonomic and Systematic History of Foraminacephale brevis Systematic Paleontology Description Frontoparietal Variation and Ontogeny Frontoparietal Histology Phylogenetic Systematics Discussion References Appendices v

6 List of Tables Chapter 1: Introduction...1 Table 1.1 A comparn of Belly River Group pachycephalosaur taxonomy...3 Chapter 2: Cranial Ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria) and the evolution of cranial doming and ornamentation in Pachycephalosauria...6 Table 2.1 Reference and taxonomic assignment of Ornatotholus browni...11 Table 6.1 Allometric regression of Stegoceras validum frontoparietal ontogeny...31 Table 6.2 Allometric regression of Stegoceras validum frontoparietal ontogeny excluding all specimens less domed than TMP Table 6.3 Allometric regression of Stegoceras validum frontoparietal ontogeny excluding all specimens less domed than TMP , except for three partially domed frontals...35 Table 8.1 Allometric Regression of Stegoceras validum squamosal ontogeny...57 Table 9.1 Ontogenetic changes in the morphology of Stegoceras validum...68 Chapter 3: Systematics and Ontogeny of Foraminacephale brevis gen. nov. (Ornithischia: Pachycephalosauria)...81 Table 5.1 Allometric regression of Foraminacephale brevis frontoparietal ontogeny Table 5.2 Allometric regressions comparing the frontoparietal ontogeny of the Belly River Group taxa Table 5.3 Importance of the first four components for the principal components analysis of the Belly River Group taxa with the covariance matrix Table 5.4 Importance of the first four components for the principal components analysis of the Belly River Group taxa with the correlation matrix Table 5.5 Loadings of variables for the first four components for the principal components analysis of the Belly River Group taxa with the covariance matrix Table 5.6 Loadings of variables for the first four components for the principal components analysis of the Belly River Group taxa with the correlation matrix vi

7 List of Figures Chapter 1: Introduction...1 Chapter 2: Cranial Ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria) and the evolution of cranial doming and ornamentation in Pachycephalosauria...6 Figure 4.1 UCMZ(VP) Figure 4.2 UALVP Figure 6.1 Illustration of the 21 linear measurements taken between 27 homologous landmarks...36 Figure 6.2 Growth series of Stegoceras validum...37 Figure 6.3 Bivariate logarithmic plots with RMA regression lines...38 Figure 6.4 Bivariate logarithmic plots with RMA regression lines for frontoparietal thickness vs. width...39 Figure 7.1 High-Resolution CT Images of AMNH Figure 7.2 High-Resolution CT Images of TMP Figure 7.3 High-Resolution CT Images of ROM Figure 7.4 Outline of methodology for calculation of relative void-space for ANMH 5450 (left), TMP (middle), and ROM (right)...50 Figure 7.5 Bivariate logarithmic plots with RMA regression lines for frontoparietal thickness vs width and frontoparietal thickness vs frontal length and the CT scans for ANMH 5450, TMP , and ROM 5355 and their relative void-spaces...51 Figure 8.1 Hypothetical growth series of Stegoceras validum squamosals...58 Figure 8.2 Bivariate logarithmic plots with RMA regression line for node height vs squamosal width...59 Chapter 3: Systematics and Ontogeny of Foraminacephale brevis gen. nov. (Ornithischia: Pachycephalosauria)...81 Figure 4.1 Squamosals of Foraminacephale brevis, Sphaerotholus buchholtzae, and Stegoceras validum...98 Figure 4.2 Postorbitals of Foraminacephale brevis, Sphaerotholus buchholtzae, and Stegoceras validum...99 Figure 4.3 3D reconstruction of Foraminacephale brevis Figure 4.4 Juvenile parietal of Foraminacephale brevis, TMP Figure 4.5 Juvenile parietal of Foraminacephale brevis, TMP Figure 4.6 Juvenile frontoparietal of Foraminacephale brevis, TMP Figure 4.7 Juvenile frontoparietal of Foraminacephale brevis, TMP Figure 4.8 Juvenile frontoparietal of Foraminacephale brevis, TMP Figure 4.9 Juvenile frontoparietal of Foraminacephale brevis, TMP Figure 4.10 Subadult frontoparietal of Foraminacephale brevis, TMP Figure 4.11 Subadult frontoparietal of Foraminacephale brevis, TMP Figure 4.12 Adult frontoparietal of Foraminacephale brevis, TMP Figure 4.13 Adult frontoparietal of Foraminacephale brevis, TMP Figure 5.1 Growth series of Foraminacephale brevis Figure 5.2 Bivariate logarithmic plots with RMA regression lines for selected variables of Foraminacephale brevis Figure 5.3 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa vii

8 Figure 5.4 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa Figure 5.5 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa Figure 5.6 Principal component analysis of the Belly River Group taxa Figure 6.1 Calculation of relative void-space for the Foraminacephale brevis growth series Figure 6.2 Calculation of relative void-space for the Foraminacephale brevis growth series Figure 6.3 Bivariate logarithmic plots with RMA regression lines for Foraminacephale brevis with the frontal CT slices from the F. brevis growth series and their relative void-spaces Figure 6.4 Bivariate logarithmic plots with RMA regression lines for Foraminacephale brevis with the parietal CT slices from the F. brevis growth series and their relative void-spaces Figure 7.1 Strict consensus of 54 most parsimonious trees viii

9 List of Appendices Chapter 1: Introduction...1 Chapter 2: Cranial Ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria) and the evolution of cranial doming and ornamentation in Pachycephalosauria...6 Appendix 1 Description of measurements...77 Appendix 2 Specimens of Stegoceras validum used in the allometric analyses and their measurements...78 Appendix 3 HRCT Methods...79 Chapter 3: Systematics and Ontogeny of Foraminacephale brevis gen. nov. (Ornithischia: Pachycephalosauria)...81 Appendix 1 Description and illustration of measurements Appendix 2 Specimens used in the allometric analyses and their measurements Appendix 3 Specimens of used in the principal components analysis and their measurements Appendix 4 HRCT Methods Appendix 5 List of characters used in the phylogeneic analysis ix

10 1 Chapter 1 Introduction Pachycephalosaurs are a group of small to medium-sized herbivorous dinosaurs from the Late Cretaceous of North America and Asia (85 65 Ma). While complete specimens are extremely rare, the abundance of lated cranial remains indicates that they were an important part of dinosaur faunal assemblages. Unfortunately, the lack of complete material has resulted in a poor understanding of the diversity and systematics of the group. Since pachycephalosaurs were first discovered and described their evolutionary relationships have been poorly resolved and their taxonomy in constant flux (Nopcsa, 1904; Lambe, 1918; Gilmore, 1924; Romer, 1927; Sternberg, 1945; Rozhdestvensky, 1964; Maryańska and Osmólska, 1974; Dong, 1978; Perle et al., 1982; Sereno, 1984, 1986, 2000; Sues and Galton, 1987; Williamson and Carr, 2002; Sullivan, 2003, 2006; Maryańska et al., 2004; Schott et al., 2009). Resolution of pachycephalosaur diversity and systematics is required for examination of biodiversity changes and patterns of evolution in this clade. The Belly River Group assemblage exemplifies the issue of pachycephalosaur diversity. Material from this assemblage has been proposed to represent from a single species (Stegoceras validum) to at least four distinct genera (Stegoceras, Prenocephale, Hanssuesia, and Colepiocephale) with potentially two additional unnamed species and several intermediate views of diversity (Williamson and Carr, 2002; Sullivan, 2003, 2006; Maryańska et al., 2004; Ryan and Evans, 2005; Schott et al., 2009). As a result, the validity of even the named species remains in question (see Table 1.1). Such discrepancies can have a large effect on the analysis large scale patterns and thus resolution of pachycephalosaur diversity will contribute significantly to our

11 2 understanding biodiversity changes and patterns of evolution in terrestrial vertebrates leading into the transition from the Mesozoic to the Cenozoic world. Most of the discrepancy in pachycephalosaur taxonomy and systematics is due to the lack of specimens preserved with diagnostic peripheral elements, most notably the squamosal. Currently only one of the possible Belly River Group species (Stegoceras validum) is known from more than lated frontoparietal domes. In order to begin to resolve the issue of pachycephalosaur diversity within the Belly River group, I examine in detail individual, ontogenetic, and specific variation in Belly River Group taxa using multiple methodologies, including comparative morphology, allometric growth curves, and quantitative histology. Knowledge of variation in these taxa combined with an extensive survey of pachycephalosaur morphology is used to create a new character matrix, which is turn used in a phylogenetic analysis of Pachycephalosauria.

12 3 Table 1.1 A comparn of Belly River Group pachycephalosaur taxonomy. Williamson and Carr (2002) Sullivan (2003, 2006) Maryańska et al. (2004) Stegoceras validum Stegoceras validum Stegoceras validum Stegoceras lambei Colepiocephale lambei Stegoceras validum Stegoceras sternbergi Hanssuesia sternbergi Stegoceras validum Stegoceras breve Prenocephale brevis Stegoceras validum

13 4 References Dong, Z A pachycephalosaur from the Wang-shih Formation in Lai-yang (Country), Shantung (Province). Vertebrata PalAsiatica 16: Gilmore, C. W On Troodon validus, an orthopodus dinosaur from the Belly River Cretaceous of Alberta, Canada. Bulletin of the Department of Geology, University of Alberta, Canada. Bulletin of the Department of Geology, University of Alberta 1:1 43. Lambe, L. M The Cretaceous genus Stegoceras, typifying a new family referred provisionally to the Stegosauria. Transactions of the Royal Society of Canada 12(4): Maryañska, T and H. Osmólska Pachycephalosauria, a new suborder of ornithischian dinosaurs. Paleontologica Polonica 26: Maryañska, T., R. E. Chapman, and D. B. Weishampel Pachycephalosauria; pp in D. B. Weishampel, P. Dodson, and H. Osmólska (eds.), The Dinosauria (2nd ed.). University of Cambridge Press, Berkeley. Nopcsa, F Dinosaurierreste aus Siebenürgen. Denksch. Kais Akad. Wiss. 74: Perle, A., T. Maryañska, and H. Osmólska Goyocephale lattimorei gen. et sp. n., a new flat-headed pachycephalosaur (Ornithischia, Dinosauria) from the Upper Cretaceous of Mongolia. Acta Palaeontologica Polonica 27: Rozhdestvensky, A. K Pachycephalosauridae. Osnovy Paleontologii 12: Romer, A. S The pelvic musculature of ornithischian dinosaurs. Acta Zoologica 8: Ryan, M. J. and D. C. Evans Ornithischian dinosaurs; pp in P. J. Currie and E. B. Kroppelhus (eds.), Dinosaur Provincial Park. Indiana University Press, Bloomington and Indianapolis.

14 5 Schott, R. K., D. C. Evans, T. E. Williamson, T. D. Carr, and M. B. Goodwin The anatomy and systematics of Colepiocephale lambei (Dinosauria: Pachycephalosauridae). Journal of Vertebrate Paleontology 29: Sereno, P. C The phylogeny of Ornithischia: a reappraisal; pp in W. Reif and F. Westphal (eds.), Third Symposium on Mesozoic Terrestrial Ecosystems, Short Papers. Tübingen: Attempto Verlag. Sereno, P. C Phylogeny of the bird-hipped dinosaurs (Order Ornithiscia). National Geographic Research 2: Sereno, P. C The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia; pp in M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin (eds.), The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge. Sternberg, C. M Pachycephalosauridae proposed for dome-headed dinosaurs, Stegoceras lambei, n. sp., described. Journal of Paleontology 19: Sues, H.-D. and P. M. Galton Anatomy and classification of the North American Pachycephalosauria (Dinosauria: Ornithschia). Paleontographica Abteilung A 198:1 40. Sullivan, R. M Revision of the dinosaur Colepiocephale lambei (Ornithischia, Pachycephalosauridae). Journal of Vertabrate Paleontology 23: Sullivan, R. M A taxonomic review of the Pachycephalosauidae (Dinosauria: Ornithischia) New Mexico Museum of Natural History and Science Bulletin 35: Williamson T. E. and T. D. Carr A new genus of derived pachycephalosaurian from western North America. Journal of Vertebrate Paleontology 22:

15 6 Chapter 2 Cranial Ontogeny in Stegoceras validum (Dinosauria: Pachycephalosauria) and the evolution of cranial doming and ornamentation in Pachycephalosauria 1 Author Contributions The following paper is from a mutli-authored manuscript. The authorship is as follows: Ryan K. Schott, David C. Evans, Mark B. Goodwin, John R. Horner, Caleb Marshall Brown, and Nicholas R. Longrich. The project and experiments were conceived by myself and David Evans. The experiments and analyses of the data were performed by myself. The paper was written by primarily myself, with contributions from all the authors, especially David Evans. Section Morphological Comparns was written by Mark Goodwin and myself. Section 9.2 HRCT and Presence of Collagen Fibers along Cranial Sutures was written by Mark Goodwin. Material and data were contributed by David Evans, Mark Goodwin, John Horner, Caleb Brown, and Nicholas Longrich. Scientific illustrations were done by Caleb Brown. 2 Introduction Pachycephalosaurs are a clade of small- to medium-sized herbivorous dinosaurs that inhabited North America and Asia during the Late Cretaceous (Maryańska et al., 2004). Most pachycephalosaur species are known primarily from cranial material, and most specimens consist only of the thickened frontoparietal that is characteristic of the group. Traditionally, two types of pachycephalosaurs have been recognized: those with thickened, but relatively flat frontoparietals and those with frontoparietals that are thickened to form a dome. These two morphological types

16 7 have been recognized as separate clades, the flat-headed Homalocephalidae and the domed Pachycephalosauridae (Sternberg, 1945; Dong, 1978; Sues and Galton, 1987). Other studies, including the most recent phylogenetic analyses, do not recognize Homalocephalidae and instead the flat-headed taxa are found to form successive sister taxa to Pachycephalosauridae, which remain a monophyletic group (Sereno, 1986, 1999, 2000; Williamson and Carr, 2002; Sullivan, 2003; Maryańska et al., 2004; Schott et al., 2009). Despite this, separation of flat-headed and domed specimens into separate taxa is not always accurate. The first North American pachycephalosaurid specimen found resembling the flat-headed condition, AMNH 5450, consists of an incipiently domed frontoparietal. A summary of the taxonomic history, diagnostic characters, and morphological interpretations of AMNH 5450 is provided in Table 2.1. This specimen, characterized by the separation of low frontal and parietal domes by a shallow transverse depression and the large size of the supratemporal fenestrae, was originally referred to Stegoceras validum by Galton (1971). Following Brown and Schlaikjer (1943), Galton (1971) suggested that there was sexual variation in the degree of doming and hypothesized AMNH 5450, with its relatively low dome, represented a female morph of S. validum. Subsequently, Wall and Galton (1979) transferred AMNH 5450 to a new species, S. browni, based on a lack of overlap with S. validum in three indices (frontoparietal: width vs. length, height vs. length, and height vs. width) described by Brown and Schlaikjer (1943). Later, Galton and Sues (1983) erected a new genus, Ornatotholus, for S. browni. This was based largely on the lack of a dome in AMNH 5450, which was similar in size to a domed specimen of S. validum (CMN 138), and on the relatively larger diameter of the open supratemporal fenestrae of AMNH 5450 compared to S. validum. Galton and Sues (1983) described additional lated flat frontals and parietals from the Campanian of Alberta that they referred to Ornatotholus in further support of the taxonomic distinction of O. browni from Stegoceras. Galton and Sues

17 8 (1983) did not believe that such pronounced differences in doming could be due to individual or sexual variation alone, and considered the many small tubercles that cover the dorsal surface of O. browni to be unique and therefore diagnostic. Goodwin et al. (1998) suggested that inflation of the frontoparietal dome was an ontogenetic character based on the histological features of a specimen referred to Stegoceras validum (MOR 295), which were interpreted as evidence of highly vascular, fast-growing bone consistent with an increase in doming with ontogenetic age. The putatively diagnostic features of O. browni identified by previous authors (Galton, 1971; Wall and Galton, 1979; Galton and Sues, 1983), such as the shallow transverse depression and large supratemporal fenestrae, were hypothesized to be ontogenetically variable within Stegoceras, and it was argued that O. browni might represent a juvenile ontogenetic stage of Stegoceras (Goodwin et al., 1998). In accord with Goodwin et al. (1998), Williamson and Carr (2002) constructed a hypothetical growth series for Stegoceras validum in which they posited that O. browni might represent a juvenile of S. validum. However, they admitted that they cannot demonstrate synonymy because Ornatotholus browni lacks any characters that would allow it to be referred to S. validum and the characters of the frontoparietal dome that diagnose S. validum appear only to develop at later ontogenetic stages (Williamson and Carr, 2002). Consequently, Williamson and Carr (2002) considered O. browni to be Pachycephalosauridae incertae sedis. Sullivan (2003) formally synonymized O. browni and S. validum, but did not address the issues raised by Williamson and Car (2002), and his reasons for synonymy are unclear. In their reviews of Pachycephalosauria, Maryańska et al. (2004) and Sereno (2000) continued to recognize Ornatotholus browni as a distinct taxon, but provided no further justification or comments on its ontogenetic status.

18 9 As recognized by Williamson and Carr (2002), pachycephalosaurs with flat-headed frontoparietals that have been attributed to Ornatotholus browni, and subsequently considered juvenile specimens of Stegoceras validum, have presented a taxonomic problem because they maintain a plesiomorphic condition seen in putatively distinct flat-headed taxa (Wannanosaurus, Goyocephale, Homalocephale) and lack diagnostic characters that would link them definitively to adult representatives. Furthermore, recent authors have begun to question the validity of these putatively primitive flat-headed taxa suggesting that they may represent juveniles of previously known taxa (Horner and Goodwin, 2009) or even paedomorphic adult morphologies (Sullivan, 2007). Here we test the hypothesis that Stegoceras validum developed ontogenetically from a flat-headed to a domed morphology using multiple independent methodologies. These include traditional comparative morphology, including the description of the first definitively flat-headed S. validum specimens; constrained geography and stratigraphy; quantitative measures of allometric growth; and comparns, both qualitative and quantitative, of bone histology using high-resolution X-ray computed tomography (HRCT). We use these methodologies to link flatheaded and domed frontoparietal morphologies in a continuous ontogenetic series supported by multiple, independent lines of evidence, and provide the first detailed analysis of frontoparietal ontogeny in a pachycephalosaur. 2.1 Institutional Abbreviations AMNH, American Museum of Natural History, New York, USA; CMN, Canadian Museum of Nature, Ottawa, Canada; MOR, Museum of the Rockies, Bozeman, USA; ROM, Royal Ontario Museum, Toronto, Canada; TMP, Royal Tyrrell Museum of Palaeontology, Drumheller, Canada; UALVP, University of Alberta Laboratory of Vertebrate Paleontology,

19 10 Edmonton, Canada; UCMP, University of California Museum of Paleontology, Berkeley, USA; UCMZ(VP), University of Calgary Museum of Zoology (Vertebrate Paleontology Collection), Calgary, Canada; Z. PAL, Palaeozoological Institute, Warsaw, Poland.

20 11 Table 2.1 Reference and taxonomic assignment of AMNH 5450, the holotype of Ornatotholus browni with accompanying diagnostic characters and additional comments from this study. Reference Diagnosis and/or Description Comments (this study) Galton (1971) Stegoceras validum Female dimorph and most primitive pachycephalosaurid Gender assignment untestable and primitive condition is unsupported morphologically Wall and Stegoceras browni Galton (1979) Skull roof thick, low fp dome, and a Diagnosis based on shallow depression separates frontal from ontogenetic features not slightly lower dome on parietal present in subadult adult S. validum Galton and Sues (1983) Taxonomic Assignment Ornatotholus browni Low fp dome, frontal and parietal separated by a shallow transverse depression, parietal slightly lower than frontal, and stf much larger than in S. validum Sues and Ornatotholus browni Low fp dome, in holotype frontal and Galton (1987) parietal dome divided by a shallow transverse depression; frontal and parietal covered by prominent tubercles dorsally Diagnosis for genus is defined by transitional ontogenetic characters absent in adult Stegoceras sp. Revised diagnosis defined by transitional ontogenetic characters absent in adult Stegoceras sp. Goodwin et al. (1998) Stegoceras validum (juvenile) Reinterpret O. browni as a juvenile S. First reappraisal of O. browni validum based on its flat, uninflated as a juvenile S. validum frontalparietal, open midline frontal suture and frontal-parietal suture dorsally, open stf, holotype is 30% smaller than domed S. validum specimens Williamson and Carr (2002) Ornatotholus browni O. browni represents a possible juvenile S. Confirmed in this study (nomen dubium) validum Sullivan (2003) Ornatotholus browni O. browni is a synonym of S. validum (=S. validum) Confirmed in this study Maryańska et Ornatotholus browni O. browni included in the al. (2004) Homalocephaloidea and considered a junior synonym of S. validum This Study Stegoceras validum (juvenile) Computer tomography confirms juvenile AMNH 5450 is an early status of the holotype Computer ontogenetic stage of S. tomography confirms juvenile status of the validum holotype O. browni with S. validum (Sullivan, 2003)

21 12 3 Systematic Paleontology Dinosauria Owen 1842 Ornithischia Seeley 1887 Pachycephalosauria Maryańska and Osmólska 1974 Pachycephalosauridae Sternberg 1945 Stegoceras Lambe 1902 Stegoceras validum Lambe 1902 Lectotype CMN 515, nearly complete frontoparietal dome. Type Locality and Horizon. East side of the Red Deer River below the mouth of Berry Creek, Alberta; Dinosaur Park Formation (Campanian). Referred Specimens UCMZ(VP) , partial right frontal, fused parietals, right squamosal, and right postorbital. Additional referred specimens: UCMZ , lated flat parietal; UALVP 49531, partial/fragmentary squamosals, parietals, frontals, postorbitals, supraorbitals. Locality and Horizon UCMZ(VP) : Steveville Railway Grade, Alberta; Dinosaur Park Formation. Detailed locality data on file at the University of Calgary; UALVP 49531: Steveville area, Dinosaur Park Formation, approximately 40 m above the contact with the Oldman Formation, Dinosaur Provincial Park, Alberta. Detailed locality data on file at the University of Alberta. Emended Diagnosis A domed pachycephalosaur that differs from all other pachycephalosaurs in having a pronounced parietosquamosal shelf with ornamentation consisting of numerous minute tubercles on lateral and posterior sides of squamosals with a

22 13 prominent dorsal row of five to eight dorsally projecting primary nodes on each squamosal and a row of small, keel-shaped nodes on the lateral margin of the squamosal. Doming is never developed to the extent found in Prenocephale and Pachycephalosaurus (Based, in part, on Sullivan, 2003). 4 Description Here we describe two new flat-headed specimens of Stegoceras validum UCMZ(VP) , which has an articulated squamosal, and UALVP49531, which has partial associated squamosals. 4.1 Description of UCMZ(VP) UCMZ(VP) (Fig. 4.1) consists of a complete parietal, right squamosal, right postorbital, and incomplete right frontal preserved in articulation. The dorsal surfaces of the bones form a completely flat surface except for the small tubercle-like projections, which form the nodular surface texture present in many pachycephalosaurids. This surface texture is the same as that found covering the uninflated portions of the cranial domes of Stegoceras (Williamson and Carr, 2002) and is consistent with the surface texture thought to typify Ornatotholus browni (Galton and Sues, 1983) Frontal Only the posterior portion of the right frontal of UCMZ(VP) is preserved; the sutural surfaces for the nasal, prefrontal, anterior supraorbital, and a portion of the posterior supraorbital bones are missing. The preserved width of the right frontal is 20 mm.

23 14 The medial wall of the frontal is relatively straight and bears the striated sutural surface for the left frontal. The anterior portion of the lateral wall is also straight and preserves a small portion of the sutural surface for the posterior supraorbital. Posterior to this, much of the lateral margin is damaged. However, there is a portion of the ventral surface of the frontal that articulates clearly with the postorbital. The posterior region of the frontal contacts the parietal in a strong, slightly interdigitated butt joint. The ventral surface of the frontal preserves the anterior portion of the cerebral fossa medially and a small portion of the roof of the right orbital cavity laterally. The rugose sutural surface for the orbitosphenoid separates the cerebral and orbital fossae. Both the cerebral and orbital fossae surfaces are smooth and shallowly concave. The orbital fossa is pierced by several prominent foramina Parietal The parietal is nearly complete, with only a small band of bone missing (mainly dorsally) between the anterior margins of the supratemporal fenestrae. The parietal is 33.6 mm wide at the frontoparietal suture and is 37.4 mm in length along the midline. The open supratemporal fenestrae are prominent. The maximum diameter of the supratemporal fenestra is 11.6 mm, and the minimum distance between the supratemporal fenestrae is 17.3 mm. The parietal contacts the frontals anteriorly in a relatively straight transverse joint. Medially, the parietal has a distinctly pointed interfrontal process that projects between the posteromedial ends of the frontals. Most of the ventral portion of the sutural surface is not preserved, but the anteroventral region of the bone deepens laterally such that the height of the frontoparietal suture (and the participating bones) increases laterally. The anterolateral wall of the parietal has a long parasagittal contact with the postorbital. The lateral wall forms the anterior and medial margins of the large circular supratemporal fenestrae. On the right side, the finished

24 15 surface of the margin of the supratemporal fenestrae is broken off; however the concavity of the surface is evident. On the left side, the anterior margin of the supratemporal fenestrae is not preserved. Posterior to the supratemporal fenestrae, the posteromedial process of the parietal is triangular, in dorsal view, and tapers to a small slip between the squamosals. The squamosal is articulated with the parietal on the right side, whereas the complete sutural surface is exposed on the left side where the left squamosal is missing. The posterior-most section of the parietal is damaged, but it is clear that the parietal would have be been very narrow between the squamosals in posterior view and may have been excluded entirely from the posterior surface of the parietosquamosal bar. The posterior exposure of the posteromedial extension of the parietal is highly variable in Stegoceras validum and may be widely exposed as in UALVP 2, or nearly excluded posteriorly as in the lectotype CMN 515 (see also Sullivan, 2003). The ventral surface of the parietal is formed of the posterior half of the cerebral fossa anteromedially, whereas the posterior region that supports the supraoccipital consists mainly of broken bone surface. The lateral portion of the ventral surface of the parietal contains portions of the temporal chambers. The medial wall of the temporal chamber is highly concave, and the relatively flat dorsal surface is pierced by several foramina Squamosal Only the right squamosal is preserved, which is complete except for its ventral projections. The overall morphology of the squamosal is typical of Stegoceras validum (e.g., CMN 138, CMN 8816, TMP , UALVP 2), where it forms the posterolateral margin of the supratemporal fenestra anteromedially, and contributes to the ornamented parietosquamosal

25 16 and squamosopostorbital bars posterolaterally. The squamosal is 32.3 mm in mediolateral width and 17.8 mm in anteroposterior length. Anteriorly, the articulated squamosal contacts the right postorbital. The sutural surface is sinuous in dorsal view, a condition also present in other specimens of Stegoceras validum (e.g., CMN 138). The lateral surface of the squamosal has an irregular, nodal surface texture that continues anteriorly onto the postorbital. The dorsolateral edge has two thin, low, anteroposteriorly long nodes that contribute to a row of five similarly shaped nodes that make up the lateral (squamosopostorbital) node row (excluding the large pyramidal vertex node at the junction between the posterior and lateral node rows). These nodes are morphologically distinct from those of the posterior (parietosquamosal) node row. The medial wall of the squamosal forms the lateral and posterior margins of the supratemporal fenestra. The squamosal contributes to nearly half the margin of the fenestra. The squamosal contacts the parietal to form the parietosquamosal bar along the caudal margin of the skull. In posterior view, the squamosal almost reaches the midline posterior to the parietal, but it is unclear whether it contacts the opposite squamosal. The posterior bar of the squamosal increases in dorsoventral height laterally, as in other specimens of S. validum (Williamson and Carr, 2002). The posterodorsal margin of the squamosal has eight dome-shaped nodes that form the posterior (parietosquamosal) node row. This is slightly greater than expected for Stegoceras validum, but the number of squamosal nodes forming the posterior node row is variable in this taxon, ranging from five to at least seven (see Sullivan, 2003). In UCMZ(VP) , the restriction of the parietal between the squamosal may account for the increased number of nodes. The nodes are variable in width, with the most medial being the largest, but are nearly all the same height. Many small, irregularly-sized nodes (minute nodes) that do not form distinguishable rows or clusters, a condition typical of S. validum (e.g., CMN 138, CMN 8816,

26 17 TMP , UALVP 2), are present below the posterior node row. At the corner of the squamosal, where the lateral and posterior borders meet, there is a large pyramidal node (the vertex node) that is larger than the others and connects the posterior and lateral node rows. The ventral surface of the squamosal is smooth (anodal) where preserved. Broken regions demarcate the bases of the occipital plate and the quadrate processes, which are not preserved. Medially, the postorbital process of the squamosal extends anteriorly and underlaps the ventral surface of the postorbital in a tongue-like projection, a condition typical in S. validum (e.g., CMN 138) and pachycephalosaurids in general Postorbital The right postorbital is nearly complete, with only a small portion on the anterolateral surface missing. The rugose anterior surface marks the contact of the posterior supraorbital. The postorbital contacts the parietal medially and the squamosal posteriorly. The bone is 24.8 mm long anteroposteriorly and 18.2 mm wide mediolaterally. In dorsal view, the suture for the posterior supraorbital is angled posterolaterally, and extends further posteriorly immediately before its lateral margin. This pattern is observed in other specimens of Stegoceras validum (e.g., CMN 138). The medial wall of the postorbital is oriented posteromedially. The anterior-most portion of the medial wall would have contacted the frontal, but is not preserved. Posterior to this, the postorbital contacts the parietal. The posteriormost part of the medial wall forms a small section of the anterolateral border of the supratemporal fenestrae. In dorsal view, the lateral wall of the postorbital is relatively straight. The dorsolateral edge contains four low nodes identical to, and continuous with, those on the dorsolateral margin of the squamosal. In lateral view, the postorbital wall is relatively flat with a

27 18 subtle surface texture that is much less distinct than on the dorsal surface. Posteriorly, the postorbital contacts the squamosal in a complex, interlapping joint. 4.2 Description of UALVP UALVP consists of a fragmented frontoparietal and incomplete peripheral skull elements including the frontals, parietal, squamosals, postorbitals, posterior supraorbitals, and a complete posterior supraorbital (Fig. 4.2). Although many of the bones are incomplete, the proportions of the skull roof can be accurately estimated due to the preserved articulations. The dorsal surface texture of the bones is similar to that seen in UCMZ(VP) , as is the shape of the flat frontals and parietal. The preserved portions of the squamosals and posteromedial extension of the parietal indicate open supratemporal fenestrae that are proportionately larger than in UCMZ(VP) Both squamosals are incomplete, but a portion of the primary posterior node row is preserved. The posterior node row is incomplete on both sides, so the total number of nodes cannot be determined. There are four nodes preserved on the partial left squamosal and we estimate that at least one to two additional nodes are missing. The primary posterior squamosal nodes of UALVP show a marked difference in morphology from UCMZ(VP) The preserved nodes in UALVP are larger, subtriangular, and more widely spaced than in the latter specimen, but they resemble the nodes of UALVP 2 closely. Both domed and triangular node morphologies are present and within the range of variation seen in Stegoceras validum specimens from Dinosaur Provincial Park (e.g., UALVP 2, CMN 138, TMP ). The morphology and sutural contracts of the postorbital are identical to UCMZ(VP) The partial left squamosal and partial left postorbital preserve the lateral node row with a large vertex node that would have connected the lateral and posterior node rows. The five preserved lateral

28 19 nodes are similar in shape, although they are slightly larger and more distinct than those in UCMZ(VP) The left posterior supraorbital is nearly complete, missing only a small portion medially where it articulates with the postorbital. The dorsal surface texture of the posterior supraorbital is tuberuclate as in the other bones. The anterior sutural surface for articulation with the anterior supraorbital is relatively straight except where it extends ventrally and slightly posteriorly at the lateral edge. The lateral edge forms the postorbital-supraorbital bar, which is devoid of any nodes and slightly convex anteroposteriorly. Ventral to the bar, the lateral wall is slightly concave mediolaterally, but otherwise flat with a surface texture similar to the minute nodes present on the posterior wall of the squamosal, although somewhat less distinct. The posterior sutural surface for articulation with the postorbital is narrow and angles anteriorly. Medially the supraorbital contacts the frontal along a straight suture until it is excluded from contact by the postorbital. At this point the lateral portion of the posterior supraorbital continues posteriorly along the anterolateral edge of the postorbital. In ventral view, the medial portion of the posterior supraorbital forms the posterolateral portion of the slightly concave roof of the orbital cavity. In this same view the lateral portion forms a convex ridge that extends ventrally and would have formed the dorsal margin of the orbit.

29 20 Figure 4.1 UCMZ(VP) , a flat-headed juvenile Stegoceras validum in dorsal, ventral, posterior, and lateral views. Abbreviations: cf, cerebellar fossa; f, frontal; ln, lateral nodes; o, orbit; p, parietal; pn, posterior nodes; po, postorbital; s, squamosal; ssf, sutural surface for frontal; ssp, suture for parietal; stf, super temporal fenestra; sss, suture for squamosal; t, tubercles; tlp, tongue like process of squamosal; vn, vertex node.

30 21 Figure 4.2 UALVP 49531, a flat-headed juvenile Stegoceras validum dorsal, ventral, posterior, and lateral views. Abbreviations as in Figure 4.1.

31 22 5 Taxonomic Assessment The specimens UCMZ(VP) and UALVP consist of flat frontals, parietals, and the diagnostic squamosal and postorbital bones. The shape and ornamentation of these bones is within the range of variation of Stegoceras validum (e.g., UALVP 2, CMN 138, TMP ). The squamosal of UCMZ(VP) has one more node (eight) on its posterior edge than is typical of S. validum, but the number of nodes is highly variable in this taxon (e.g., CMN 138, TMP , UALVP 2). The shape and distribution of these nodes, including the distribution of minute nodes on the posterior edge of the squamosal, is distinctive of S. validum. The squamosal of UALVP 49531, with an estimated six to seven nodes in the primary node row, is within the range observed for S. validum (e.g., UALVP 2, CMN 138, TMP ). Additionally, the shape and surface texture of the squamosal and postorbital in both specimens is identical to small specimens of S. validum (e.g., CMN 138, TMP ). The flattened frontals and parietals described here closely resemble those previously referred to Ornatotholus browni (e.g., TMP ; Galton and Sues, 1983; UCMP ; Goodwin, 1990), which may suggest they could be referred to this taxon. However, the holotype of Ornatotholus browni was diagnosed on three characters of the frontoparietal which are addressed in the following section. The tuberculate surface texture identified as diagnostic of Ornatotholus browni by Galton and Sues (1983) covers the dorsal surface of the frontal, parietal, supraorbitals, postorbital, and squamosal of UCMZ(VP) , but a similar texture also covers the uninflated portions of the dome of Stegoceras (Williamson and Carr, 2002) and other pachycephalosaurs (e.g., Homalocephale, Goyocephale, Wannanosaurus, Dracorex, Stygimoloch). Furthermore, the squamosal and postorbital of domed S. validum specimens (e.g., TMP ) have identical surface texture to that seen in flat-headed specimens (UCMZ(VP) and UALVP 49531).

32 23 Thus, tuberculate surface texture appears to be a, presumably plesiomorphic, feature of known pachycephalosaurs that is associated with uninflated portions of the skull (Williamson and Carr, 2002; Goodwin and Horner, 2006). Comparing AMNH 5450 with the small Stegoceras specimen CMN 138, Galton and Sues (1983) diagnosed Ornatotholus browni on the basis of having supratemporal fenestrae that were larger than Stegoceras validum. While the conclusion of Williamson and Carr (2002) that the supratemporal fenestrae close with ontogeny may be generally correct, there is a high amount of variation in supratemporal fenestrae size. UCMZ(VP) has a maximum supratemporal diameter of 11.6 mm, whereas TMP , a domed specimen of S. validum has an average maximum supratemporal fenestra diameter of 16.5 mm. This is much larger than would be expected if the supratemporal fenestrae simply closed with ontogeny. CMN 138, another domed specimen of S. validum, and about the same size as TMP , has an average maximum supratemporal fenestra diameter of 9.1 mm. In addition to variation in relative size between specimens, the supratemporal fenestrae also show marked asymmetry within specimens, exemplified by CMN 8816 where the left fenestra appears to be closed, but the right is open. Clearly variation in the size of the supratemporal fenestrae is more complex than presently understood and further study will be needed to clarify the development of this character. The final character used to diagnose O. browni is the presence of a low dome divided by a shallow transverse depression (Galton and Sues, 1983). A shallow transverse depression is not found in other specimens referred to the genus (e.g., TMP , UCMP ), or in the other known flat frontals or parietals (e.g., UCMZ(VP) , UCMZ , TMP ), nor is it found in any other pachycephalosaur specimen known. Rather than a diagnostic character, we concur with Williamson and Carr (2002) and Goodwin et al. (1998) that this depression is most likely a transitory ontogenetic feature related to the initiation of dome

33 24 growth. Early in dome ontogeny, doming may begin separately in the frontals and parietal, producing a shallow transverse depression. This depression likely reflects the growth plate in the dermal bone that is obliterated from external view as the frontal and parietal domes combine to form a single dome. Evidence, for this feature is also found in a partially inflated frontal (TMP ) where the dome curves ventrally just anterior to the sutural surface for the parietal and thus would have formed a similar transverse depression along the frontoparietal suture. This character is not found in either flat or more fully domed specimens and thus is considered a transitory ontogenetic feature rather than a diagnostic character. Due to the diagnostic nature of the squamosal of Stegoceras validum (e.g., Sullivan 2003), we refer UCMZ(VP) and UALVP to S. validum. At the same time, the distinct nature of the pachycephalosaur squamosal and its associated ornamentation precludes us from referring UCMZ(VP) and UALVP to any other species. Given the diagnostic nature of the squamosal morphology and that these specimens occur in the same formation and geographic area as known S. validum specimens, we conclude that UCMZ(VP) and UALVP represent juvenile individuals of S. validum. Thus we corroborate the hypotheses of ontogenetic changes in the skull from flat-headed to domed in this taxon proposed by Goodwin et al. (1998), Williamson and Carr (2002), and Sullivan (2003), and for the first time it is possible to definitively place flat-headed and domed pachycephalosaurids into an assemblage-based ontogenetic series. The flat frontals and parietals from the Dinosaur Park Formation previously referred to Ornatotholus browni (e.g., TMP ) are suggestive of S. validum based on their similarity with UCMZ(VP) and are referred to this taxon. Additional analyses will allow us to further test these hypotheses and quantitatively describe ontogenetic change.

34 25 6 Frontoparietal Allometry Chapman et al. (1981) were the first to attempt to quantify dome allometry in pachycephalosaurs. They performed a reduced major axis (RMA) regression of dome length and dome thickness and concluded that the dome thickened metrically with increased length (Chapman et al., 1981). Their study was based on a sample of frontopratietal domes and incomplete skulls from Alberta that they identified as Stegoceras validum (except for S. edmontonense from the Horseshoe Canyon Formation and the poorly preserved holotyoe frontoparietal of Gravitholus albertae). On the basis of their morphometric analysis, they interpreted this sample as a single, sexually dimorphic species (excluding S. edmontonense and G. albertae). The original 'Stegoceras validum' sample set of Chapman et al. (1981) is now known to include specimens from at least three different formations (Foremost, Oldman, and Dinosaur Park) spanning nearly five million years of time (Eberth, 2005), and that are now referred to four different genera: Stegoceras, Prenocephale, Colepiocephale, and Hanssuesia (Sullivan, 2003). Thus the Chapman et al. (1981) study more likely described interspecific differences than the intraspecific allometry of a single taxon throughout its ontogenetic range. In addition, there were several problems with the analytical approach, notably the measurements and indices used by Chapman et al. (1981) were not based on homologous landmarks (Goodwin, 1990). Evidence of the flat-headed condition of juveniles in Stegoceras validum allows a detailed, morphometric analysis of frontoparietal ontogeny in a single pachycephalosaur taxon, S. validum, for the first time. The specimens referred to S. validum represent the most complete ontogenetic series known for any pachycephalosaur, and thus are ideally suited to serve as a model for frontoparietal dome growth (Williamson and Carr, 2002). Here we perform the first analysis of frontoparietal allometry of S. validum that includes specimens from the complete

35 26 known size range and which utilizes measurements based on homologous landmarks. All of the specimens included in our analysis are from the upper Belly River Group (mostly Dinosaur Park Formation) of southern Alberta, and are referable to Stegoceras validum (Sullivan, 2003; Ryan and Evans, 2005) with the arguable exception of the Ornatotholus browni holotype (AMNH 5450). 6.1 Methods Dome growth was analyzed using frontoparietal heights, lengths, and widths for a total of 21 measurements (Fig. 6.1). All measurements were taken between homologous morphological landmarks (based, in part, on those identified by Goodwin, 1990) recorded at sutural contacts. A description of the individual measurements taken is found in Appendix 1. Measurements were taken, where preserved, from 40 specimens of Stegoceras validum (including Ornatotholus browni, ANMH 5450) representative of the hypothesized growth series (Fig. 6.2, Appendix 2). These specimens are all from the Late Campanian of Alberta and do not include specimens reported from Montana or New Mexico (Goodwin, 1990; Sullivan and Lucas, 2006) to ensure the sample represents an assemblage of a single, tightly constrained taxon. Specimens and their measurements are listed in Appendix 2. Measurements were log-transformed to fit the linear allometric growth function. This has the additional benefit of normalizing the distribution of the data. Regressions were calculated using reduced major axis regression (RMA, also known as standard major axis regression) implemented in the 'lmodel2' software package (Legendre, 2008; see also Legendre and Legendre, 2008) for the statistics program R. RMA is the most appropriate method for tests of allometry (Smith, 2009; Warton et al., 2006). Statistical significance of the allometric pattern was determined based on the confidence intervals (e.g., if the confidence interval encompasses one the slope is statistically metric).

36 27 Due to the varying completeness of specimens, up to four different standard (x) variables were used: frontoparietal length, frontoparietal width, parietal length, and frontal length. Frontoparietal length is perhaps the most intuitive index of size, however only 10 specimens have complete frontoparietals. We also used parietal length and frontoparietal width as standard variables for comparns in order to include the flat-headed specimens and maximize sample size (and thus statistical power). In these cases, is was possible to include lated frontals and/or parietals in the bivariate analyses. Dodson (1975) found that intraspecific allometries in a growth series of extant Alligator mississippiensis showed very high correlation coefficients and thus we consider a high correlation coefficient as support for the hypothesis that the specimens of Stegoceras validum represent an ontogenetic series of a single species. In order to further test the hypothesis that flat-headed specimens (including the Ornatotholus browni holotype, AMNH 5450) are juveniles of domed specimens, as well as assess the effect of the flat-headed specimens on the regressions, we reconstructed each plot with the flat-headed taxa excluded and compared the results. This was performed in two stages. First, all specimens that were less thick than TMP , the smallest unequivocal domed Stegoceras validum (based on the presence of an articulated squamosal), were excluded. This included all flat-headed specimens, some partially domed frontals, and the holotype of Onatotholus browni (AMNH 5450). At the second stage, we added back in the three partially domed frontals (TMP , , and ) and reran the regressions. 6.2 Results Results of the bivariate comparns are listed in Tables , and graphs of select indices are presented in Figures 6.3 and 6.4. Regressions using parietal width as the standard variable have the highest sample size and thus the greatest statistical power. Comparns using

37 28 frontoparietal length and parietal length have lower sample sizes that may hinder the ability to detect positive and negative allometry statistically. Comparns using frontal length as the standard variable, while having relatively high sample sizes, tended to have much lower correlation coefficients ( ). This suggests that frontal length is not ideal as a standard variable, most likely due to a high amount of individual variation in frontal length. The correlation coefficients are generally fairly high (most are >0.7, see Table 6.1), they are not as high as Dodson (1975). Dodson (1975) noted that the high correlation coefficients of his results were partially due to the large size range of his Alligator sample. Our size range of Stegoceras is not nearly as large, and thus we would expect to find somewhat lower correlation coefficients even in an intraspecific ontogenetic series Heights The height of the dome generally exhibits positive allometric growth (Table 6.1, Fig. 6.3A). The allometry is strongest when compared to frontoparietal width and frontal length, where the slopes are all significantly positive. In all cases, the thickness of the frontoparietal shows strong positive allometry. When compared to frontoparietal and parietal length the slopes are generally smaller, although in most cases the difference is not significant. The slopes for the heights are considerably higher when frontal length is used as the standard variable. This is likely due to the negative allometry of frontal length when compared to the other standard variables (see below), but also may be affected by the lower correlation coefficients Widths The growth of the width of the dome does not show a single definitive pattern (Table 6.1, Figure 6.3B). When compared to frontoparietal width, some of the dome widths exhibit metry

38 29 (W:n/prf and W:po/stf/sq) and others show significant negative allometry (W:prf/aso, W:aso/pso, W:pso/po). All of the widths show metry when frontoparietal length is used as the standard variable. The same is true when parietal length is used as the standard variable except that the width across the frontoparietal is positively allometric Lengths The lengths of the frontoparietal sutures all exhibit statistically metric growth and in most cases the slopes are near one (Table 6.1, Fig. 6.3C). Both the frontal and parietal lengths show negative allometry when compared to frontoparietal width (Table 6.1, Fig. 6.3D). While the length of the frontoparietal is statistically metric the slope is rather low (0.67) and thus detection is likely hindered by the small sample size (n = 10). Additionally, frontal length is negatively metric when compared to both frontoparietal and parietal lengths Flat-headed Taxa Excluded When flat-headed specimens, the holotype of Ornatotholus browni, and partially domed frontals are excluded from the bivariate analysis, allometric regression of frontoparietal heights generally resulted in statistically lower slopes (Table 6.2, Fig. 6.4). The trends, however, were similar in all comparns to those using the total sample, although with the smaller sample size and corresponding loss of statistical power most of the results do not differ significantly from metry. The exception to this is comparns based on frontal length, which showed positive allometry, but the low correlations suggest the results are not meaningful. We did not recover positive allometry for frontoparietal thickness compared to frontoparietal width or length, and while this is certainly complicated by the loss of sample size, it is likely that the small specimens anchor the regression line. It may also suggest that the slope decreases for larger frontoparietals,

39 30 however the small sample size of large sized domes prevents us from further addressing this possibility. When the three partially domed frontals (TMP , , and ) were added back into the data set, we again recovered only statistically positive allometric relationships when frontoparietal heights are regressed against width (Table 6.3, Fig. 6.4). In most cases, the slopes are not significantly different than when all specimens are included in the analysis, with the exception of frontoparietal thickness, which was significantly lower.

40 31 Table 6.1 Allometric regression of Stegoceras validum frontoparietal ontogeny. Abbreviations: CI, 95% confidence interval;, metry; +, positive allometry; o, negative allometry; H:n/n, height of the sutural surface at the contact of the nasals; H:n/prf, height of the sutural surface at the contact of the nasal and prefrontal; H:prf/aso, height of the sutural surface at the contact of the prefrontal and anterior supraorbital; H:aso/pso, height of the sutural surface at the contact of the anterior supraorbital and posterior supraorbital; H: pso/po, height of the sutural surface at the contact of the posterior supraorbital and postorbital; T:fp, thickness of the frontoparietal; L:aso, length of the supraorbital suture; L:pso, length of the posterior supraorbital suture; L:po, length of the postorbital suture; L:f, length of the frontal; L:p, length of the parietal; L:fp, length of the frontoparietal; W:n/prf, width between the nasal/prefrontal sutural contacts; W:prf/aso, width between the aso/pso supraorbital sutural contacts; W: aso/pso, width between anterior and posterior supraorbital sutural contacts; W:pso/po, width between the posterior supraorbital and postorbital sutural contacts; W:po/stf/sq, width between the contacts of postorbital suture and the supratemporal fenestrae or the squamosal suture if fenestrae are closed. Heights Frontoparietal Width N H:n/n 17 H:n/prf 16 H:prf/aso 26 H:aso/pso 22 H:pso/po 24 T:fp 39 R Slope CI Intercept CI Allometry Frontoparietal Length N H:n/n 9 H:n/prf 7 H:prf/aso 10 H:aso/pso 8 H:pso/po 10 T:fp 10 R Slope CI Intercept CI Allometry R Slope CI Intercept CI Allometry + + Parietal Length H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp N

41 32 Frontal Length R Slope CI Intercept CI Allometry Frontoparietal Width N W:n/prf 21 W:prf/aso 25 W:aso/pso 24 W:pso/po 24 W:po/stf/sq 15 R Slope CI Intercept CI Allometry Frontoparietal Length N W:n/prf 9 W:prf/aso 10 W:aso/pso 10 W:pso/po 9 W:fp 10 W:po/stf/sq 8 R Slope CI Intercept CI Allometry R Slope CI Intercept CI Allometry + H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp N Widths Parietal Length W:n/prf W:prf/aso W:aso/pso W:pso/po W:fp W:po/stf/sq N

42 33 Lengths Frontoparietal Width N L:aso 26 L:pso 26 L:po 18 L:f 21 L:p 19 L:fp 10 R Slope CI Intercept CI Allometry Frontoparietal Length N L:aso 10 L:pso 10 L:po 9 L:f 10 L:p 10 R Slope CI Intercept CI Allometry R Slope CI Intercept CI Allometry Parietal Length L:aso L:pso L:po L:f N

43 34 Table 6.2 Allometric regression of Stegoceras validum frontoparietal ontogeny excluding all specimens less domed (thick) than TMP Abbreviations as in Table 6.1. Heights Frontoparietal Width N H:n/n 14 H:n/prf 12 H:prf/aso 18 H:aso/pso 16 H:pso/po 17 T:fp 24 R Slope CI Intercept CI Allometry + + Frontoparietal Length N H:n/n 8 H:n/prf 6 H:prf/aso 9 H:aso/pso 7 H:pso/po 9 T:fp 9 R Slope CI Intercept CI Allometry + N R Slope CI Intercept CI Allometry N R Slope CI Intercept CI Allometry + Parietal Length H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp Frontal Length H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp

44 35 Table 6.3 Allometric regression of Stegoceras validum frontoparietal ontogeny excluding all specimens less domed (thick) than TMP , except for three partially domed frontals (TMP , , and ). Abbreviations as in Table 6.1. Heights Frontoparietal Width N H:n/n 15 H:n/prf 14 H:prf/aso 21 H:aso/pso 19 H:pso/po 20 T:fp 27 R Slope CI Intercept CI Allometry R Slope CI Intercept CI Allometry Frontal Length H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp N

45 36 Figure 6.1 Illustration of the 21 linear measurements taken between 27 homologous landmarks. Abbreviations: aso, anterior supraorbital; f, frontal; fnb, frontonasal boss; n, nasal; p, parietal; prf, prefrontal; po, postorbital; pso, posterior supraorbital; sq, squamosal; ss, sutural surface for; stf, supratemporal fenestrae.

46 37 Figure 6.2 Growth series of Stegoceras validum in dorsal (top) and lateral (bottom) views depicting the transition from a flat-headed to domed frontoparietal morphology that occurred through ontogeny in this taxon. A, UCMZ(VP) ; B, UCMZ(VP) ; C, AMNH 5450 (holotype of Ornatotholus browni); D, CMN 138; E, CMN 515 (holotype of Stegoceras validum); F, UALVP 2.

47 38 Figure 6.3 Bivariate logarithmic plots with RMA regression lines for selected variables. A, heights of frontoparietal vs width across frontoparietal suture. Blue, height of nasal; red, height of frontoparietal at the contact of the anterior and posterior supraorbitals; green, thickness of frontoparietal. B, widths of frontoparietal vs width across frontoparietal suture. Blue, width of frontonasal boss; red, width across the supraorbital lobes; green width across the frontoparietal at the contact of the posterior supraorbital and postorbital sutures. C, lengths of frontoparietal vs width across frontoparietal suture. Blue, length of frontal; red, length of parietal; green length of frontoparietal. D, lengths of frontoparietal vs length of parietal. Blue, length of anterior supraorbital suture; red, length of posterior supraorbital suture; green, length of postorbital suture; orange, length of frontal. Measurements are log (mm).

48 39 Figure 6.4 Bivariate logarithmic plots with RMA regression lines for frontoparietal thickness vs. width with only specimens more domed (thicker) than TMP (red), with the addition of three partially domed frontals (blue), and with all specimens included (grey). Measurements are log (mm).

49 40 7 Frontoparietal Bone Histology Osteohistology, the study of bone microstructure, has been used to study the ontogeny of both extant and extinct animals (e.g., Chinsamy, 1995; Reid, 1996; Stark and Chinsamy, 2002; Botha and Chinsamy, 2000, 2004, 2005; Ray et al., 2004; Erickson, 2005; Tumarkin-Deratzian et al., 2006, 2007; de Boef and Larsson, 2007). However, these studies, and the methodologies used therein, have largely been confined to the study limb bones, and analysis of limb bone histology is not feasible for pachycephalosaurs due to the lack of postcranial remains. Osteohistology of pachycephalosaurs has been studied exclusively in cranial bone, which is dermal in origin and formed by intramembranous ossification, whereas limb bones are formed by endochondral ossification (Ham and Leeson, 1961). Despite differences in the formation process, the relative growth differences in dermal bones at various ontogenetic stages should produce differences similar to those characterized in long bones (Ham and Leeson, 1971). Dermal bones and long bones contain the same materials and both bone types undergo similar environmental variability (Ham and Leeson, 1971). As a result, differences in tissue types and structures observable between juvenile and adult limb bones should also be observable between juvenile and adult cranial bone (Tumarkin-Deratzian, 2007). Cranial histology of pachycephalosaurs has previously been examined by Brown and Schlaikjer (1943), Goodwin et al. (1998), Goodwin and Horner (2004), and Horner and Goodwin (2009). The first two studies described the cranial histology of a single specimen. The study of Goodwin and Horner (2004), however used histological sections of frontoparietal domes from a multi-taxic size series to test functional hypotheses of the dome. Horner and Goodwing (2009) used histology of frontoparietals and cranial ornamentation as support for the hypothesized synonymy of Dracorex, Stygimoloch, and Pachycephalosaurus, but were unable to analyze the histology of all three taxa. While these studies were purely qualitative, the results suggested that

50 41 the vascularity of the dome decreased with ontogenetic age (Goodwin and Horner, 2004). Here we test this hypothesis quantitatively using CT-data from three specimens from our growth series of Stegoceras validum in addition to providing qualitative observations on bone histology in these specimens. 7.1 Methods High Resolution Computed Tomography (HRCT) scans were chosen as a non-destructive alternative to thin-sectioning for examining bone histology. While the benefits of a nondestructive alternative are obvious, the appropriateness of HRCT scans as a replacement for thinsectioning has not been tested and thus will need to be further addressed in a future study. HRCT scans were taken from specimens at three stages in the ontogenetic growth series presented in this paper (ANMH 5450, TMP84.5.1, ROM 53555). Morphological comparns were made between the scans of the three specimens including an analysis of the zones of Goodwin and Horner (2004). Scans were performed by the University of Texas High-Resolution X-ray CT Facility (UTCT). Analytical details of each scan are provided in Appendix Relative Vascularity Relative vascularity could not be measured directly due to the nature of both CT data and fossil specimens. Void space in the CT scans was chosen as a reasonable proxy for canal space, as any empty space in the specimens should be the result of canals through the bone. We cannot know what proportion of the canals were actually occupied by vessels, but it follows they should be highly correlated. Additionally, because we are measuring only relative changes the exact method used will have less of an impact as long as the correlation is consistent between the specimens.

51 42 Void space was calculated using transverse sections from the CT data. Sections were chosen at the boundary of the posterior supraorbital and postorbital sutures on the frontal from three specimens of Stegoceras validum (the holotype of Ornatotholus browni, AMNH 5450; TMP ; and ROM 53555). This ensures that the sections will be homologous between the three specimens and thus directly comparable. Each slice was thresholded with the Huang method (Huang and Wang, 1995; Wang et al., 2002), which minimizes fuzziness in the 2-D grayscale histogram, implemented in ImageJ (Rasband, ; see also Abramoff et al., 2004). There are numerous different methods of thresholding and while some have been found to perform better than others, the best method is still highly dependent on the images being analyzed (Sezgin and Sankur, 2004). Therefore, we chose the Huang method (Huang and Wang, 1995; Wang et al., 2002) based on a survey of different thresholding techniques as the method that best approximated what we interpret as the canal space in the scan. To properly compare void space, a region of interest was specified by a rectangle of the same relative proportions drawn onto each thresholded image at the base of the braincase immediately to the left of the interfrontal suture and extended so that the lop left corner met the edge of the frontal. This technique ensured that homologous areas were compared despite differences in the shapes of the frontals. This step was also used to eliminate potential artifacts, such as the effects of beam hardening which can result in the edge of the specimen being recognized as void space. Thresholded images were cropped to this region of interest and relative void space was calculated using the voxel counter plug-in for ImageJ (Rasband, ; see also Abramoff et al., 2004).

52 Results Morphological Comparns Three specimens were HRCT scanned for this study, AMNH 5450, the holotype of Ornatotholus browni and what is here and elsewhere (Goodwin et al., 1998; Williamson and Carr, 2002; Sullivan, 2003) described as a juvenile Stegoceras validum; TMP , a relatively small specimen of S. validum and presumed juvenile or subadult; and ROM one of the largest known specimen of S. validum and a presumed adult. Reconstructions and slices from the HRCT scans of AMNH 5450 are shown in Figure 7.1. Plaster-filled breaks along the frontoparietal suture and anterior frontal, nearly indistinguishable on the original, are clearly visible by darker brown color. In dorsal view (Fig. 7.1A), the low-relief tubercular ornamentation is evident as are the interfrontal and frontoparietal sutures, which are filled with plaster. In sagittal section (Fig. 7.1B), the contrasting vascularity within Zones I III is visible. The plaster-filled frontoparietal suture is traceable from the roof of the braincase to the dorsal surface of the frontoparietal. The slightly 'wavy' brighter layer in the upper region of Zone II is likely an artifact produced by mineralized collagen fibers present across and bordering the interfrontal suture. A horizontal slice from above the roof of the braincase through the highly vascular Zone II reveals the interfrontal and plaster-filled frontoparietal suture (Fig. 7.1A). Anteriorly the prefrontal-frontal suture is partially obscured by plaster. In anterior view, the interfrontal suture is visible along the midline of the frontal and is tightly interdigitated anteriorly (Fig. 7.1C). The transverse slice through the frontals (Fig. 7.1C) shows the open interfrontal suture, which extends from the roof of the braincase to the dorsal surface of the thickened but undomed frontal. The internal microstructure is zonal and a slightly truncated 'm' shaped layer in Zone II appears as a brighter colored region with relatively less

53 44 vascularity. This appears to be the same HRCT artifact produced by the presence of mineralized connective tissue or bundles of collagen fibers present along patent cranial sutures. Reconstructions and corresponding slices of TMP are shown in Figure 7.2. The reconstruction of the left lateral view (Fig. 7.2B) confirms the inflated frontal and anterior parietal contribute to the cranial dome at this ontogenetic stage. The dorsal surface is covered by tubercular ornamentation and the posterior parietal is thickened but undomed. Sutures between the lateral cranial elements are visible. A trace of the frontoparietal suture on the dorsal surface occurs as a 'blurry' region on the dorsal surface, but is effectively indistinguishable in the fossil. In sagittal section (Fig. 7.2B), the highly vascular tissue of the interior skull is apparent, along with the braincase and patent cranial sutures. Zone II appears to either abruptly end in the middle of the skull (Fig. 7.2B middle) or exhibit a visible texture change where it contacts a layer of what looks to be relatively denser, radiating tissue (Fig. 7.2B right). We interpret this change in texture as an artifact of HRCT produced by the increase in concentration or change in orientation of mineralized collagen fibers along the interfrontal and frontal-parietal sutures. Higher resolution histology thin-section slides unmistakably show the continuation of Zone II nearly to the edge of the frontal dome in pachycephalosaurs of this ontogenetic stage (see Goodwin and Horner, 2004; Fig. 4B, Fig. 5C). The frontal-parietal suture is visible as a sinuous darker vertical line bordered by lighter-colored tissue denoting this intermembranous bone growth site in the skull (Fig. 7.2B right). The interfrontal and frontoparietal sutures are not visible on the dorsal surface of the skill because they are tightly closed, obscuring the relatively early developmental stage of the frontoparietal dome (Fig. 7.2A). Sutures between the supraorbital, postorbital, and squamosal are patent and along with the modest frontoparietal dome and open supratemporal fenestrae, support the juvenile ontogenetic age assignment of this skull. A horizontal slice above the roof of the braincase reveals the increased vascularity of Zone II within the frontoparietal

54 45 dome and patent cranial sutures throughout the skull (Fig. 7.2A). The lighter colored zones bordering the patent interfrontal, the interdigitated frontoparietal, and associated cranial sutures are likely an HRCT artifact produced by the abundant mineralized collagen (Sharpey s) fibers present along these contacts. The anterior transverse section of TMP (Fig. 7.2B) shows the interfrontal suture dividing Zone II nearly in half along the midline where it contacts the dorsal roof of the braincase. Differences in the relative vascularity between Zones I, II, and III are visible, but the boundaries between these zones are not as clearly defined using HRCT as they are in histological slides at this stage of ontogeny (see Goodwin and Horner 2004 Fig 5C). Zone II extends nearly to the dorsal surface of the highly expanding, fast growing frontoparietal dome in this stage of ontogeny. This is obscured by the HRCT artifact giving this bone a denser appearance or texture compared with the surrounding more vascular tissue. The more posterior coronal section shows the HRCT produced artifact of asymmetry in Zone II that is caused by the presence of mineralized collagen (Sharpey s) fibers present in concentrated along the frontalparietal and interfrontal sutures. Frontal, sagittal, and transverse sections from CT scans of ROM are shown in Figure 7.3. The frontal section (Fig. 7.3A) reveals that both the inferfrontal and frontoparietal sutures remain open internally even at the largest known size of Stegoceras validum. While the frontoparietal suture is still bordered by the lighter colored zone at this stage, the inter-frontal suture no longer is. In both the sagittal and the transverse sections (Fig 7.3B,C) the three histological zones are clearly distinguishable, but the lighter-colored 'wavy' band that was found through zone II appears to have expanded and now separates Zones I and II. This lighter-colored band still extends upwards into Zone II in the transverse section (Fig. 7.3C, arrow). Additionally, in the transverse section the open interfrontal suture is visible. Overall, these sections show less

55 46 vascular canals than the smaller specimens, however, Zone I and II are still distinguishable as being less dense than Zone III due to their darker colour in the scan Relative Vascularity The specimens and slices used to calculate relative vascularity are illustrated in Figure 7.4. It was found that the proxy for relative vascularity (relative void space) decreased with each successively larger specimen with 20% in AMNH 5450, to 17% in TMP , to 7% in ROM This is further illustrated in Figure 7.5 where relative void space is plotted with the regressions on frontoparietal thickness on frontoparietal width and frontal length, respectively. From these graphs it can be seen that the decrease in our proxy for relative vascularity corresponds well with the development of the frontoparietal dome. This corresponds with the qualitative observations of Goodwin and Horner (2006) for the multitaxic size sample.

56 47 Figure 7.1 High-Resolution CT Images of AMNH 5450 in A, dorsal view (left) and frontal section (right); B, lateral view (left) and sagittal section (right); and C, anterior view (left) and transverse section through the frontal (right). The histological zones of Goodwin and Horner (2004) are denoted. The arrow identifies an artifact likely produced by mineralized collagen fibres present across and along the interfrontal suture. Abbreviations: f, frontal; f-f, inter-frontal suture; f-p, frontoparietal suture; p, parietal; po, postorbital; pso, posterior supraorbital, sq, squamosal; Z-1 to Z-3, histological zones 1 to 3.

57 48 Figure 7.2 High-Resolution CT Images of TMP in A, dorsal view (left) and frontal section (right); B, lateral view (left), median sagittal section (middle), and lateral sagittal section (right); and C, anterior view (left), anterior transverse section through the frontal (middle), and posterior transverse section through the parietal (right). The histological zones of Goodwin and Horner (2004) are denoted. The arrow identifies an artifact likely produced by mineralized collagen fibres present across and along the interfrontal and frontoparietal sutures. Abbreviations: f, frontal; f-f, inter-frontal suture; f-p, frontoparietal suture; p, parietal; po, postorbital; pso, posterior supraorbital, sq, squamosal; Z-1 to Z-3, histological zones 1 to 3.

58 49 Figure 7.3 High-Resolution CT Images of ROM in A, frontal section; B, median sagittal section; and C, transverse section through the posterior portion of the frontal. The histological zones of Goodwin and Horner (2004) are denoted. The arrow identifies an artifact likely produced by mineralized collagen fibres. Abbreviations: f, frontal; f-f, inter-frontal suture; f-p, frontoparietal suture; p, parietal; Z-1 to Z-3, histological Zones 1 to 3.

59 50 Figure 7.4 Outline of methodology for calculation of relative void-space for ANMH 5450 (left), TMP (middle), and ROM (right). A transverse section was taken from CT scan of the frontal at the contact of the posterior supraorbital and postorbital sutures. Scans were thresholded using the Huang method (Huang and Wang 1995). Void-space was counted from an area of interest (red square).

60 51 Figure 7.5 Bivariate logarithmic plots with RMA regression lines for frontoparietal thickness vs width (A) and frontoparietal thickness vs frontal length (B) and the CT scans for ANMH 5450, TMP , and ROM 5355 and their relative void-spaces (a proxy for vascularity). Note the substantial decrease of vascularity with increased dome development (thickness, A) and size (frontal length, B). Measurements are log (mm).

61 52 8 Squamosal Variation and Ontogeny The squamosal is one the most diagnostic pachycephalosaur bones and is often used in species diagnoses and phylogenetic characters (Williamson and Carr, 2002; Sullivan, 2003). Horner and Goodwin (2009) recently postulated ontogenetic changes in node pattern, number, and size in the hypothesized ontogenetic transition from Dracorex to Pachycephalosaurus. Despite these points, very little is known about squamosal variation and ontogeny in pachycephalosaurs. Stegoceras validum is the pachycephalosaur taxa with highest number of identified articulated and lated squamosals. These squamosals are referred to S. validum based on their general morphology and nodal ornamentation, as well as their relative size and stratigraphic occurrence. Due to the large number of referred squamosals, S. validum is an ideal candidate to examine squamosal variation and ontogenetic trends. 8.1 Methods Both articulated and lated squamosals referred to Stegoceras validum were examined and compared morphologically. A size rage of these squamosals were placed into a hypothetical growth series. The series was used to examine changes in morphology as well as the allometric growth of the nodes. Squamosal width was measured ventral to the node row and the height of the second posterior node was measured from the base of the node to the apex. The second node was chosen because the first (medial-most) node was often enlarged relative to the other nodes and sometimes straddled the parietosquamosal suture and we wanted to avoid any possible confounding factors caused by this. Measurements were taken from eight squamosals (both lated and articulated). Methodology for the regression is same as that done for frontoparietal allometry.

62 Results Comparative Morphology A hypothetical growth series of squamosals referable to S. validum is presented in Figure 8.1. This series contains the complete known size range of S. validum squamosals and thus is indicative of the variation and ontogeny of the S. validum squamosal. The smallest squamosal referable to S. validum (TMP ) is an lated squamosal with a preserved width of 19.5 mm. While the squamosal is not fully complete, its width is nearly complete with only the medial sutural contact for the parietal showing some abrasion. Six posterior nodes, including the large vertex node, and one lateral node are preserved. The height of the node medial to the vertex node is 3 mm. Overall the node morphology is similar to that of UCMZ(VP) The squamosal bar of TMP , however, differs in that it is almost non-existent. In most larger S. validum squamosals, including UCMZ(VP) , the posterior edge of the squamosal below the node row forms a flat squamosal bar. The squamosal bar is often highly textured and covered with minute nodes. In TMP this bar is not present in posterior view as the squamosal is very dorsoventrally flat. Thus, the relative width of the squamosal bar may be ontogenetically variable, but some fairly small squamosals do have tall squamosal bars (e.g., TMP ), while other larger squamosals have less distinct bars (e.g., TMP ) suggesting that individual variation may obscure the ontogenetic pattern. Node morphology and number is highly variable among the known size series. The highest number of posterior squamosal nodes (including the vertex node) is nine (UCMZ(VP) ). There are some lated squamosals that have eight posterior nodes (e.g., TMP and ). A number of specimens have seven nodes (e.g., TMP , , UALVP 49437), while most specimens have six (e.g., CMN 138, TMP , , , , UALVP 2). This large range in node number raises the question of

63 54 whether node number changes through ontogeny. The size series shown here does not suggest either a strong increase or decrease in node number, but there is some evidence of a weak trend towards a decreased node number with increasing size. Additionally, close inspection of nodes shows that some nodes are connected more than others and look as if they may have split apart or joined together ontogenetically. This is shown clearly in UCMZ(VP) where the nodes just left of the vertex node nearly form a single node. Some specimens also have nodes that are much smaller than all the other nodes on the squamosal. In the lated squamosal TMP ,there is a small, slender node between two much larger nodes. One of the nodes on the lated squamosal TMP has a secondary protrusion on its left side that may have originally been a distinct node (which would have brought the node count to 9, the same as UCMZ(VP) ). TMP is an interesting lated squamosal with at least seven, but probably eight posterior nodes (The missing posteromedial-most section is about the size of a node, although the anterior portion containing most of the sutural surface for the parietal is preserved). The nodes of this specimen are peculiar in that they are flattened and joined dorsally forming a fairly continuous ridge. This morphology may be homologous to the lateral ridges found on Stegoceras validum that appear to be formed from the confluence of lateral nodes (e.g., in UALVP 2; see Fig. 6.2). The same morphology is also seen in Prenocephale prenes (Z. Pal. MgD-I/104). A lateral ridge is preserved on the postorbital of the relatively smaller TMP UCMZ(VP) has a row of anteroposteriorly elongate lateral nodes that are continuous along the squamosal and postorbital, but that do not form a ridge. Interestingly, the lateral ridge does not extend onto the squamosal of TMP (which has distinct and elongate lateral nodes), but does on the larger UALVP 2. This suggests the possibility of the ontogenetic formation of the lateral ridge from distinct nodes that begins anteriorly and progresses posteriorly. The posterior nodes may also be affected by this process, as well as coalescing

64 55 through ontogeny (thus reducing node number), and eventually may form into a ridge continuous with the lateral ridge, as appears to be the case in TMP It is important to note, however, that TMP , while possessing this partial posterior ridge is smaller than the squamosal of UALVP 2 and has a higher number of nodes that are also smaller. The posterior nodes of Stegoceras validum squamosals all have a basic pyramidal shape with the apex pointed dorsally. This differs from the conical, posteriorly pointed nodes of Prenocephale prenes (Z. Pal. MgD-I/104) and the tongue-like nodes of Sphaerotholus buchholtzae (TMP ). The nodes of S. validum vary in that the apexes can either be rounded or pointed. There does not appear to be any consistent pattern in the shape and this may simply reflect individual variation or preservation. Several other changes to squamosal morphology appear to correlate with ontogeny. The thickness of the parietosquamosal bar generally increases, although as discussed above this is individually quite variable. The dorsal surface of the squamosal goes from flat (e.g., UCMZ(VP) ) to inflated and somewhat domed (e.g., TMP , UALVP 2) as the squamosal and other peripheral elements are incorporated into the dome. Additionally, the squamosal bar progresses from being flat in posterior view to convex, although this is also individually variable (e.g., TMP is flat and CMN 138 is curved, but both are roughly the same size). This trait is likely also connected to the incorporation of the peripheral elements into the dome. This is corroborated by CMN 138 which is more ontogenetically advanced than TMP (thicker dome, smaller fenestrae, peripheral elements further incorporated into the dome, squamosals ventrally convex).

65 Node Allometry Regression of node height on squamosal width for eight squamosals is graphed in Figure 8.2 with regression statistics available in Table 8.1. The slope of the regression is fairly close to one (1.26) and is statistically metric. The correlation between the variables is fairly high (r = 0.78), which suggests that despite the small sample size the results are meaningful.

66 57 Table 8.1 Allometric Regression of Stegoceras validum squamosal ontogeny. Abbreviations: CI, 95% confidence interval;, metry; +, positive allometry; o, negative allometry; H:node, height of the 2nd squamosal node. Squamosal Width N H:node 8 R 0.78 Slope 1.26 CI Intercept CI Allometry

67 58 Figure 8.1 Hypothetical growth series of Stegoceras validum squamosals in posterior (top) and dorsal (bottom) views. A, TMP ; B, TMP ; C, UCMZ(VP) ; D, TMP ; E, TMP; F, UALVP, 49437; G, TMP ; H, TMP ; and I, UAVLP 2.Note the variation in node number and shape throughout the series that seems to lack an ontogenetic signal. The overall pattern of nodal ornamental, however, remains constant. Scale equals 2 cm.

68 59 Figure 8.2 Bivariate logarithmic plots with RMA regression line for node height vs squamosal width. Measurements are log (mm).

69 60 9 Discussion 9.1 The Ontogeny of Stegoceras validum The recognition of undomed individuals of Stegoceras validum is consistent with the hypothesis that the skull of this pachycephalosaur underwent marked changes in shape during growth, with flat-headed Stegoceras juveniles developing into adults with highly domed skulls. Contrary to the results of Chapman et al. (1981), our bivariate analyses show that frontoparietal thickness exhibits significant positive allometry with respect to frontoparietal width and length, a result consistent with a growth series from flat-headed to fully domed. While the regression slopes decreased when flat-headed specimens were excluded from the analysis, this result is largely explained by the substantial decrease in sample size and range, and the more interesting possibility that growth rates actually decreased in larger specimens, as has been found with the positive allometry of beetle horns (Pomfret and Knell, 2006). Unfortunately, we lack the sample size required to test this further. Several important changes in the shape of the frontoparietal are evident throughout the ontogeny of Stegoceras validum. The width of the frontoparietal is positively allometric with respect to its length, and the width across the frontoparietal suture increased at a faster rate than the width across the supraorbital lobes. Additionally, the parietal grew at a faster rate than the frontal. Combined, these changes would serve to modify the shape of the frontoparietal throughout ontogeny beyond the changes expected from simple thickening. As the frontoparietal increases in size, the dome becomes relatively thicker and wider posteriorly than it does anteriorly, resulting in the distinctive pear-shaped dome that characterizes Stegoceras. A similar growth pattern may also characterize Colepiocephale, given the similar shape of its cranial dome. The differences in the growth rate of the frontals and parietal, in addition to serving to

70 61 modify the shape of the skull, have implications for the use of frontal and parietal lengths in phylogenetic analyses. Our results also differ from Chapman et al. (1981) with respect to the inference of sexual dimorphism in Stegoceras validum. With the strict inclusion of only S. validum material from the Upper Belly River Group of Alberta (Oldman and Dinosaur Park Formations), our analysis did not reveal any compelling evidence of sexual dimorphism in dome shape based on visual inspection of the plots. The available sample of large individuals is small, making the identification of bimodal trends or clusters difficult. Allometric patterns based on our sample suggest a significant amount of variation in the growth trajectory of the dome that appears to be largely size-related with no clear distinctions. Therefore, it is most likely that the results of Chapman et al. (1981) were influenced by the inclusion of specimens that do not pertain to S. validum (e.g., those of 'Prenocephale' brevis and Colepiocephale lambei). The growth series assembled here reveals that other morphological changes accompanied the doming of the skull (Table 9.1) some of which were hypothesized previously by Williamson and Carr (2002). One notable difference between the adult and juvenile is the fusion of the frontals to each other and to the parietals. Fusion appears to occur first in the frontals and then the co-ossified frontals fuse to the parietals; e.g., in CMN 3135 and CMN the inter-frontal suture is no longer seen either dorsally or ventrally, but a well-developed frontoparietal suture is retained. However, this fusion only occurs in the outer layers of bone as both the frontoparietal and interfrontal sutures remain open internally, even in the largest specimens (ROM 53555, Fig. 7.3). Specimens also show a marked decrease in relative void space (a proxy for vascularity) coinciding with the development of the dome. While the functional implications for this have previously been addressed (Goodwin and Horner, 2004), with the new methodology developed

71 62 in this paper for quantifying changes in relative vascularity, it will now be possible to assign relative ontogenetic ages to specimens. Our methodology can be extended to other pachycephalosaur taxa and used to test the various conflicting hypotheses of synonymy and alpha taxonomy (e.g., Williamson and Carr, 2002; Sullivan, 2003; Maryańska et al., 2004). Additionally, comparns of histology with other pachycephalosaur taxa will likely shed additional light on the function of the dome in this group. Ontogenetic modifications to the skull are not confined to the frontals and parietals; the posterior (parietosquamosal) and lateral shelves formed by the parietal, squamosals, postorbital and supraorbitals are shortened, relative to overall size, as they become incorporated into the dome during ontogeny. This is accomplished by inflation and incorporation into the dome of the medial extension of the parietal, the postorbitals, and the posterior supraorbitals. The anterior supraorbital bones and the supraorbital lobe of the frontal do not become overly inflated and thus do not contribute to the dome in adult Stegoceras validum. This results in deep grooves separating the supraorbital lobes from a prominent frontonasal boss that is characteristic of Stegoceras and other closely related taxa. In contrast to changes in dome morphology, variation in the ornamentation of the parietosquamosal shelf does not appear to correlate with size. The parietosquamosal ornamentation of the small, flat-headed specimens from the upper Belly River Group described here closely resemble that in larger, domed specimens of Stegoceras validum from the same stratigraphic interval. The diagnostic nodal ornamentation of the squamosal in S. validum is present even in the smallest specimens and it appears that primary node growth is metric with respect to squamosal width. This result is somewhat expected based on qualitative observation, but its is interesting to note that a transition from Dracorex to Stygimoloch would have required positive allometry of the squamosal horns (nodes). If this proves to be correct it may suggest a

72 63 change in the function of the cranial ornamentation between Stegoceras and Dracorex/Stygimoloch. Also, there is no evidence of the squamosal nodes being resorbed in adult Stegoceras, as has been hypothesized in Pachycephalosaurus (Horner and Goodwin, 2009). Although the nodal ornamentation shows a large degree of intraspecific individual variability both in the number and shape of the posterior nodes, this does not appear to be dependent on ontogenetic stage or linked to sexual maturity or sexual dimorphism. Rather, the pattern of ornamentation appears to be constant, and thus is potentially more taxonomically and phylogenetically informative than ontogenetically variable features of dome shape. This in turn supports the validity of taxa named on the basis of lated squamosals, such as Alaskacephale (Gangloff et al., 2005; Sullivan, 2006). Stygimoloch, a case in point, was originally named on the basis of an lated squamosal (Galton and Sues, 1983), and subsequently has been shown to have other unique characters associated with the dome once more complete specimens were discovered (Goodwin et al., 1998). However, these characters were most recently hypothesized to be a product of growth within an ontogenetic sequence for Dracorex Stygimoloch Pachycephalosaurus (Horner at al., 2007; Horner and Goodwin, 2009). The maintenance of the ornamental pattern in Stegoceras validum throughout ontogeny suggests that pachycephalosaur squamosal ornamentation was potentially important for visual communication and species recognition throughout its entire ontogeny. As a result, it is unlikely that this ornamentation primarily functioned in sexual display. As hypothesized by Horner and Goodwin (2006) for Triceratops, and others concerning additional taxa (see Padian et al., 2010), juvenile morphologies that differ significantly from adult morphologies suggest strong intraspecific social signaling. Clearly it was important for juveniles to recognize adults and vice versa. The pattern seen in pachycephalosaurs may be a useful model for other taxa, but ontogeny clearly evolves in different ways in different taxa. In ceratopsian dinosaurs, for instance, the

73 64 parietal and squamosal ornamentation varies markedly between adults and juveniles (Sampson, 1997; Horner and Goodwin, 2006; Currie et al., 2008). Furthermore, these findings contribute to a growing body of data that supports the hypothesis that similar types of ornamentation in dinosaurs were used for intraspecific visual communication and species recognition even at very early ontogenetic stages, rather than being tightly associated with sexual display at maturity (Goodwin and Horner, 2004; Main et al., 2005; Goodwin et al., 2006; Padian and Horner, 2010). 9.2 HRCT and Presence of Collagen Fibers along Cranial Sutures HRCT reveals morphological and sutural details that are not easily distinguishable on the surface of pachycephalosaur skulls due to relative ontogenetic stage, preservation, or excess plaster and consolidants during preparation. HRCT confirms the interfrontal and frontal-parietal sutures, reconstructed plaster regions, and glued breaks in the holotype of Ornatotholus browni (=Stegoceras validum; AMNH 5450). In the partial skull of a juvenile Stegoceras validum, TMP , vascular zones and cranial sutures are clearly seen, however the boundaries between these zones are less clear than revealed by standard histological slides on account of an HRCT-produced artifact from the presence of mineralized collagen (Sharpey s) fibers along cranial sutures. The arrangement and orientation of these collagen fibers and the corresponding ectocranial suture sinuosity and cross-sectional interdigitization reflect loading and skull deformation in mammals (Herring and Teng, 2000; Jaslow, 1990) and fish (Markey and Marshall, 2007). These fibrous joints or sutural contacts between the bones of the skull are linked together by collagen fibers and connective tissue that bridge the contacting cranial bones (Jaslow, 1990). In mammals, these sutures are the site of intramembranous bone growth in the skull and the major center of bone expansion within the craniofacial vault (Opperman, 2000). New bone is

74 65 produced at the sutural edges of the bone front in response to external stimuli during cranial morphogenesis. There is evidence that the frontoparietal domes in pachycephalosaurs, and cranial ornamentation in most dinosaurs, grow much differently than the cranial bones in mammals and other tetrapods (Horner and Goodwin, 2009; Horner and Lamm, 2009). In the HRCT scans of pachycephalosaurs presented here, these intermembranous bone growth sites appear to be relatively denser (and lighter color) compared to the surrounding tissue in the artificially colored slices from the skull. We hypothesize that this HRCT produced artifact is due to the preservation of dense concentrations of mineralized collagen (Sharpey s) fibers along these cranial sutures visible in histological slides of Stegoceras (Goodwin and Horner, 2004). The boundaries between Zones I III defined by Goodwin and Horner (2004) are also less distinct in the HRCT slices, however Zone I and II are still highly vascular compared to Zone III. 9.3 Ontogenetic Assessment of Flat-headed Pachycephalosaur Taxa The presence of a transitional flat-headed morphology in Stegoceras validum has significant implications for the delineation of pachycephalosaur species. Recent work by a number of authors has called into question the validity of all flat-headed taxa, with the potential that they may be immature specimens, many of which may pertain to previously named taxa (Horner and Goodwin, 2009), or be paedomorphic in nature (Sullivan, 2007). Due to large samples from a relatively restricted stratigraphic interval, the reconstructed ontogenetic series of S. validum serves as the most complete model for testing and confirming ontogenetic variation in pachycephalosaurs. The flat-headed pachycephalosaur taxon Wannanosaurus has been considered a juvenile based on unfused cranial sutures, large supratemporal fenestrae, and nodular surface texture (Sereno, 2000; Butler and Zhao, 2009) all characters that are shared with juvenile Stegoceras that

75 66 are reduced and eventually eliminated during ontogeny. Goyocephale, Homalocephale, and Dracorex, the remaining flat-headed taxa, also appear to share these three characteristics (Maryańska and Osmólska, 1974; Perle et al., 1982; Bakker et. al., 2006; Evans et al., in press), and thus may also represent juveniles. Each of the flat-headed taxa show the three characters identified in this study as being linked to a juvenile, flat-headed condition in Stegoceras. It is unclear whether these characters are linked solely to a flat-headed state or to a juvenile state, the difference being that if these characters are associated with a flat-headed condition, rather than ontogenetic stage, they would also be expected to be present in flat-headed adults. Proper resolution of this problem will require the use of multiple, independent methodologies, as developed in this study, including the use of comparative morphology of stratigraphically and geographically constrained taxa, allometric growth curves, and quantitative histological analysis. 9.4 Implications for the Taxonomy and Systematics of Fossil Species This study demonstrates that as Stegoceras matured, the skull underwent a remarkable metamorphosis. The extensive nature of these changes is such that juveniles and adults differ radically in their general appearance, and this has led previous workers to create separate taxa for adults and juveniles. This practice is not unique within vertebrate paleontology and similar issues have arisen in the taxonomy of other dinosaurs. Large morphological differences between adults and juveniles of lambeosaurine dinosaurs, for instance, led to the creation of the taxon Procheneosaurus to accommodate juveniles of Corythosaurus and Lambeosaurus (Evans et al., 2005; Evans, 2010). Brachyceratops was erected on the basis of a juvenile centrosaurine (Sampson et al., 1997; Brown et al., 2009) and Aublysodon and Nanotyrannus (Carr and Williamson, 2004) were erected on the basis of juvenile tyrannosaurid material. The key to resolving such problems is to establish ontogenetic models which allow one to recognize

76 67 characters associated with early ontogenetic stages, fossils of juveniles, identify features that are stable over ontogeny and therefore useful understanding the relationships of the animals in question. We argue that to create a robust alpha taxonomy or systematic framework it is imperative to have a model of the ontogeny of the study organism. Understanding which changes are likely to be the result of ontogeny, and which are not, can help in the choice of more stable characters, formulating a diagnosis, and undertaking a phylogenetic study. In the case of Stegoceras, the growth series created for this taxon and supported by multiple, independent lines of evidence, also serves as a model for interpreting the variation observed in other pachycephalosaurs and dinosaurs in general. However, it should be kept in mind that patterns of growth, like any other aspect of an organism, can evolve. Even within a single clade, ontogenetic patterns can show marked differences. For instance, the epoccipital ossifications of the ceratopsid frill become greatly elaborated over the course of development Pachyrhinosaurus (Currie et al., 2008), whereas these ossifications become low and difficult to differentiate from the frill as Triceratops matures (Horner and Goodwin, 2006; Scannella and Horner, 2010). While the patterns of ontogenetic change in Stegoceras provide a useful starting point for understanding the evolutionary and developmental relationships among other pachycephalosaurs, they may not characterize all pachycephalosaurs perfectly. Thus, it is crucial that the ontogeny of each taxon be analyzed within an ontogenetic framework supported by multiple, independent lines of evidence.

77 68 Table 9.1 Ontogenetic changes in the morphology of Stegoceras validum. Juvenile Frontals and parietals flat Adult Frontals and parietals thickened and domed Dorsal surface of frontals and parietals covered in numerous tubercles Frontoparietal smooth, tubercles present only on periphery Frontal-frontal suture open Frontal-parietal suture open Frontoparietal highly vascular Frontoparietal rectangular Posterior and lateral shelves flat and prominent Posteromedial extension of parietal long and flat Postorbitals and posterior supraorbitals flat Frontals fused externally Frontals and parietals fused externally Vascularity highly reduced Frontoparietal pear-shaped Posterior and lateral shelves reduced and incorporated into the dome Posteromedial extension of parietal short and mostly incorporated into dome Postorbitals and posterior supraorbitals inflated and contribute to dome Frontonasal boss flat Frontonasal boss high, convex and separated by grooves

78 69 References Abramoff, M. D., P. J. Magelhaes, and S. J. Ram Image Processing with ImageJ. Biophotonics International 11(7): Botha, J., and A. Chinsamy Growth patterns deduced from the bone histology of the cynodonts Diademodon and Cynognathus. Journal of Vertebrate Paleontology 20: Botha, J., and A. Chinsamy Growth and life habits of the Triassic cynodont Trirachodon, inferred from bone histology. Acta Palaeontologica Polonica 49: Botha, J., and A. Chinsamy Growth patterns of Thrinaxodon liorhinus, a non-mammalian cynodont from the Lower Triassic of South Africa. Palaeontology 48: Brown, B. and E. M. Schlaikjer A study of the troödont dinosaurs with the description of a new genus and four new species. Bulletin of the American Museum of Natural History 82: Brown, C. M., A. P. Russell, and M. J. Ryan Pattern and transition of surficial bone texture of the centrosaurine frill and their ontogenetic and taxonomic implications. Journal of Vertebrate Paleontology 29: Butler, R. J. and Q. Zhao The small-bodied ornithischian dinosaurs Micropachycephalosaurus hongtuyanensis and Wannanosaurus yansiensis from the Late Cretaceous of China. Cretaceous Research 30: Carr, T.D. and T.E. Williamson Diversity of Late Maastrichtian Tyrannosauridae (Dinosauria, Theropoda): Zoological Journal of the Linnean Society 142: Chapman, R. E., P. M. Galton, J. J. Sepkoski, Jr., and W. P. Wall A morphometric study of the cranium of the pachycephalosaurid dinosaur Stegoceras. Journal of Paleontology 55:

79 70 Chinsamy, A Ontogenetic changes in the bone histology of the Late Jurassic ornithopod Dryosaurus lettowvorbecki. Journal of Vertebrate Paleontology 15: Currie, P. J., W. Langston, Jr., and D. H. Tanke A new species of Pachyrhinosaurus (Dinosauria, Ceratopsidae) from the Upper Cretaceous of Alberta, Canada. National Research Council Research Press, Ottawa, Canada: de Boef, M., and H. C. E. Larsson Bone microstructure quantifying bone vascular orientation. Canadian Journal of Zoology 85: Dodson, P Functional and ecological significance of relative growth in Alligator. Journal of Zoology 175: Erickson, G. M Assessing dinosaur growth patterns: a microscopic revolution. Trends in Ecology and Evolution 20: Evans, D. C Cranial anatomy and ontogeny of Hypacrosaurus altispinus, and a comparative analysis of skull growth in lambeosaurines (Ornithischia: Hadrosauridae). Zoological Journal of the Linnean Society 159: Evans, D.C., C.A. Forster, and R.R. Reisz The type specimen of Tetragonosaurus erectofrons (Ornithischia: Hadrosauridae) and the identification of juvenile lambeosaurines; pp in P. Currie and E. Koppelhus (eds.), Dinosaur Provincial Park: A Spectacular Ancient Ecosystem Revealed. Indiana University Press, Bloomington. Evans, D. C., C. M. Brown, M. J. Ryan, K. Tsogtbattar. in press. Cranial ornamentation and ontogenetic status of Homalocephalae calathocercos (Ornithischia: Pachycephalosauria) from the Nemegt Formation, Mongolia. Journal of Vertebrate Paleontology.

80 71 Galton, P. M A primitive dome-headed dinosaur (Ornithischia: Pachycephalosauridae) from the Lower Cretaceous of England and the function of the dome of pachycephalosaurids. Journal of Paleontology 45: Galton, P. M. and H.-D. Sues New data on pachycephalosaurid dinosaurs (Reptilia: Ornithischia) from North America. Canadian Journal of Earth Science 20: Goodwin, M. B Morphometric landmarks of pachycephalosaurid cranial material from the Judith River Formation of northcentral Montana; pp in K. Carpenter and P. J. Currie (eds.), Dinosaur Systematics Approaches and Perspectives. Cambridge University Press, Cambridge. Goodwin, M. B., and J. R. Horner Cranial histology of pachycephalosaurs (Ornithischia: Marginocephalia) reveals transitory structures inconsistent with head-butting behavior. Paleobiology 30: Goodwin, M. B., E. A. Buchholtz, and R. E. Johnson Cranial anatomy and diagnosis of Stygimoloch spinifer (Ornithischia: Pachycephalosauria) with comments on cranial display structures in agonistic behavior. Journal of Vertebrate Paleontology 18: Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian The smallest known Triceratops skull: new observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology 26(1): Ham, A. W., and T. S. Leeson Histology. Pitman Medical Publishing Co. Ltd, London, U.K., 942 pp. Herring, S. W. and S. Teng Strain in the braincase and its sutures during function. American Journal of Physical Anthropology 112:

81 72 Horner, J. R. and M. B. Goodwin Major cranial changes in Triceratops. Proceedings of the Royal Society (B): 273: Horner, J. R. and M. B. Goodwin Extreme cranial ontogeny in the Upper Cretaceous dinosaur Pachycephalosaurus. PLoS ONE 4(10):e7626. Horner, J. R., and E. T. Lamm Metaplasia provided dinosaur skulls extreme morphological plasticity during ontogeny. Journal of Vertebrate Paleontology 29(3, Supplement):117A. Horner, J., M. Goodwin, and H. Woodward Synonomy consequences of dinosaur cranial ontogeny. Journal of Vertebrate Paleontology Abstracts Supplement 27(3):92A. Huang, L.-K. And M.-J. J. Wang Image thresholding my minimizing the measues of fuzziness. Pattern Recognition 28: Jaslow, C.R Mechanical properties of cranial sutures. Journal of Biomechanics 23: Lambe, L. M New genera and species from the Belly River series (Mid-Cretaceous). Contributions to Canadian Paleontology, Geological Survey of Canada 3: Legendre, P Model II Regression User`s Guide, R edition. R Vignette. Legendre, P. and L. Legendre Numerical Ecology, 2nd Edn. Elsevier Science, Amsterdam. Main, R. P., A. de Ricqles, J. R. Horner, and K. Padian The evolution and function of thyreophoran dinosaur scutes: implications for plate function in stegosaurs. Paleobiology, 31(2): Markey, M.J. and C.R. Marshall Linking form and function of the fibrous joints in the skull: a new quantification scheme for cranial sutures using the extant fish Polypterus endlicherii. Journal of Morphology 268:

82 73 Maryańska, T and H. Osmólska Pachycephalosauria, a new suborder of ornithischian dinosaurs. Paleontologica Polonica 26: Maryańska, T., R. E. Chapman, and D. B. Weishampel Pachycephalosauria; pp in D. B. Weishampel, P. Dodson, and H. Osmólska (eds.), The Dinosauria (2nd ed.). University of California Press, Berkeley. Opperman, L.A Cranial sutures as intramembranous bone growth sites. Developmental Dynamics 219: Owen, R Report of British fossil reptiles. Pt II. Report of the British Association for the Advancement of Science 11: Padian, K. and J. R. Horner The evolution of bizarre structures in dinosaurs: biomechanics, sexual selection, social selection or species recognition Journal of Zoology in press, doi: /j x, 15pp. Padian, K., J. R. Horner, D. Fowler, and J. Scannella How a synergy of species recognition and social signaling explains cranial anatomy and ontogeny in several groups of dinosaurs. Journal of Vertebrate Paleontology 30 (Supp. 4):143A. Perle, A., T. Maryańska, and H. Osmólska Goyocephale lattimorei gen. et sp. n., a new flat-headed pachycephalosaur (Ornithischia, Dinosauria) from the Upper Cretaceous of Mongolia. Acta Palaeontologica Polonica 27: Pomfret, J. C. and R. J. Knell Sexual selection and horn allometry in the dung beetle Euoniticellus intermedius. Animal Behavior 71: Rasband, W. S ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, Ray, S., J. Botha, and A. Chinsamy Bone histology and growth patterns of some nonmammalian therapsids. Journal of Vertebrate Paleontology 24:

83 74 Reid, R. E. H Bone histology of the Cleveland-Lloyd dinosaurs and of dinosaurs in general, part 1: Introduction to Bone Tissues. Brigham Young University Geology Studies 41: Ryan, M. J., and D. C. Evans Ornithischian Dinosaurs. pp in P. J. Currie and E. Koppelhus (eds.), Dinosaur Provincial Park: A spectacular ancient ecosystem revealed. Indiana University Press, Bloomington, Indiana. Sampson, S. D Dinosaur combat and courtship. pp in J. O. Farlow and M. K. Brett-Surman (eds.), The Complete Dinosaur. Indiana University Press, Wayne, IN. Sampson, S. D., M. J. Ryan, and D. H. Tanke Craniofacial ontogeny in centrosaurine dinosaurs (Ornithischia: Ceratopsidae): Taxonomic and behavioral implications. Zoological Journal of the Linnean Society 221: Scannella, J. B. and J. R. Horner Torosaurus Marsh, 1891, is Triceratops Marsh, 1889 (Ceratopsidae: Chasmosaurinae): synonomy through ontogeny. Journal of Vertebrate Paleontology 30: Seeley, H.G., On the classification of the fossil animals commonly named Dinosauria. Proceedings of the Royal Society of London 43: Sereno, P. C Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). National Geographic Research 2: Sereno, P. C The evolution of dinosaurs. Science 284: Sereno, P. C The fossil record, systematics and evolution of pachycephalosaurs and ceratopsians from Asia; pp in M. J. Benton, M. A. Shishkin, D. M. Unwin, and E. N. Kurochkin (eds.), The Age of Dinosaurs in Russia and Mongolia. Cambridge University Press, Cambridge.

84 75 Sezgin, M. and B. Sankur Survey over image thresholding techniques and quantitative performance evaluation. Journal of Electronic Imaging 13: Smith, R. J Use and Misuse of the Reduced Major Axis for Line-Fitting. American Journal of Physical Anthropology 140: Starck, J. M., and A. Chinsamy Bone microstructure and developmental plasticity in birds and other dinosaurs. Journal of Morphology 254: Sternberg, C. M Pachycephalosauridae proposed for dome-headed dinosaurs, Stegoceras lambei, n. sp., described. Journal of Paleontology 19: Sues, H.-D. and P. M. Galton Anatomy and classification of the North American Pachycephalosauria (Dinosauria: Ornithschia). Paleontographica Abteilung A 198:1 40. Sullivan, R. M Prenocephale edmontonensis (Brown and Schlaikjer) new comb. and P. brevis (Lambe) new comb. (Dinosauria: Ornithischia: Pachycephalosauria) from the Upper Cretaceous of North America; pp in S. G. Lucas and A. B. Heckert (eds.), Dinosaurs of New Mexico. New Mexico Museum of Natural History and Science Bulletin No. 17. Sullivan, R. M Revision of the dinosaur Stegoceras Lambe (Ornithischia, Pachycephalosauridae). Journal of Vertebrate Paleontology 23: Sullivan, R. M A taxonomic review of the Pachycephalosauridae (Dinosauria: Ornithischia) New Mexico Museum of Natural History and Science Bulletin 35: Sullivan, R. M Doming, heterochrony, and paedomorphosis in the Pachycephalosauridae (Ornithischia: Dinosauria): taxonomic and phylogenetic implications. Journal of Vertebrate Paleontology Abstracts Supplement 27(3):154A 155A. Turmarkin-Deratzian, A. R Fibrolamellar bone in wild adult Alligator mississippiensis. Journal of Herpetology 41:

85 76 Tumarkin-Deratzian, A. R., D. R. Vann, and P. Dodson Bone surface texture as an ontogenetic indicator in long bones of the Canada goose Branta Canadensis (Anseriformes: Anatidae). Zoological Journal of the Linnean Society 148: Tumarkin-Deratzian, A. R., D. R. Vann, and P. Dodson Growth and textural ageing in long bones of the American alligator Alligator mississippiensis (Crocodylia: Alligatoridae). Zoological Journal of the Linnean Society 150:1 39. Wall, W. P., and P. M. Galton Notes on pachycephalosaurid dinosaurs (Reptilia: Ornithischia) from North America, with comments on their status as ornithopods. Canadian Journal of Earth Science 16: Wang, Q., Z. Chi, and R. Zhao Image thresholding by maximizing the index of nonfuzziness of the 2-D grayscale histogram. Computer Vision and Image Understanding 85: Warton, D. I., I. J. Wright, D. S. Falster, and M. Westoby Bivariate line-fitting methods for allometry. Biol. Rev. 81: Williamson T. E. and T. D. Carr A new genus of derived pachycephalosaurian from western North America. Journal of Vertebrate Paleontology 22:

86 77 Appendices Appendix 1 Description of measurements. All measurements were taken between homologous morphological landmarks. Most are positioned at the dorsal margin of the contact between sutural surfaces for the peripheral skull bones on the frontoparietal. These landmarks are as follows: n/n, n/prf, prf/aso, aso/pso, pso/po, po/stf/sq. The widths, heights, and length of the sutural surfaces were all taken between the corresponding landmarks. The length of the frontal, parietal and frontoparietal were measured ventrally along the midline between landmarks located at the anterior edge of the frontoparietal at the sutural contact of the nasals, along the midline at the contact between the frontals and parietal in the endocranial fossa, and at the posterior edge of the dome along the midline. Frontoparietal thickness was measured using proportional calipers between landmarks located within the endocranial fossa at the contact of the frontals and parietal and on the dorsal surface of the frontoparietal at the contact of the frontals and parietal. All other measurements were done using standard digital calipers. Locations of each landmark are shown in Figure 6.1. Abbreviations: aso, anterior supraorbital; n, nasal; prf, prefrontal; po, postorbital; pso, posterior supraorbital; sq, squamosal; stf, supratemporal fenestrae.

87 78 Appendix 2 Specimens of Stegoceras validum used in the allometric analyses and their measurements. Abbreviations: H:n/n, height of the sutural surface at the contact of the nasals; H:n/prf, height of the sutural surface at the contact of the nasal and prefrontal; H:prf/aso, height of the sutural surface at the contact of the prefrontal and anterior supraorbital; H:aso/pso, height of the sutural surface at the contact of the anterior supraorbital and posterior supraorbital; H: pso/po, height of the sutural surface at the contact of the posterior supraorbital and postorbital; T:fp, thickness of the frontoparietal; L:aso, length of the supraorbital suture; L:pso, length of the posterior supraorbital suture; L:po, length of the postorbital suture; L:f, length of the frontal; L:p, length of the parietal; L:fp, length of the frontoparietal; W:n/prf, width between the nasal/prefrontal sutural contacts; W:prf/aso, width between the aso/pso supraorbital sutural contacts; W: aso/pso, width between anterior and posterior supraorbital sutural contacts; W:pso/po, width between the posterior supraorbital and postorbital sutural contacts; W:po/stf/sq, width between the contacts of postorbital suture and the supratemporal fenestrae or the squamosal suture if fenestrae are closed. Specimen# AMNH5450 BMNH-R BMNH-R CMN1108A CMN138 CMN1594 CMN2379 CMN38428 CMN515 CMN8816 ROM58311 ROM803 ROMvalidum TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP TMP UALVP2 UALVP49531 UALVP51913 UALVP6 UALVP8502 UAVLP8504 UCMZ(VP) UCMZ(VP) UCMZ(VP)unnumbered H:n/n H:n/prf H:pfr/aso H:aso/pso H:pso/po W:n/pfr W:pfr/aso W:aso/pso W:pso/po W:f/p W:po/stf/sq L:aso L:pso L:po L:f 41.3 L:p L:fp T:f/p

88 79 Appendix 3 HRCT Methods. University of Texas High-Resolution X-ray CT Facility Scans of the frontoparietal of Ornatotholus browni (AMNH 5450; Late Cretaceous Dinosaur Park Fm, Alberta Canada, collected by Barnum Brown, 1913) for Dr. Mark Goodwin of the Museum of Paleontology, The University of California at Berkeley. Specimen scanned by Matthew Colbert 23 October x bit TIFF images. II, 180 kv, ma, no filter, empty container wedge, no offset, slice thickness 2 lines (= mm), S.O.D. 155 mm, 1000 views, 2 samples per view, inter-slice spacing 2 lines (= mm), field of reconstruction 50 mm (maximum field of view mm), reconstruction offset 5000, reconstruction scale 800. Acquired with 15 slices per rotation. Reconstructed with beam-hardening correction: 0.0, 0.8, Rotation correction processing done by Rachel Racicot using IDL routine DoRotationCorrection. Total slices = 824. Scans of an incomplete skull of Stegoceras sp. (TMP ; Judith River Fm, Upper Cretaceous; Steveville Area, AB) for Dr. Mark Goodwin of the Museum of Paleontology, The University of California at Berkeley. Specimen scanned by Matthew Colbert 4 November x bit TIFF images. II, 180 kv, ma, no filter, empty container wedge, no offset, slice thickness 1 line (= mm), S.O.D. 315 mm, 1200 views, 2 samples per view, inter-slice spacing 1 line (= mm), field of reconstruction 101 mm (maximum field of view mm), reconstruction offset 8000, reconstruction scale 650. Acquired with 27 slices per rotation. Reconstructed with beam-hardening correction: 0.0, 0.6, 0.25, 0.1. Rotation correction processing done by Rachel Racicot using IDL routine DoRotationCorrection. Total slices = 945.

89 80 Scans of a partial skull of Stegoceras validum (ROM 53555; Ornithischia, Pachycephalosauridae; Upper Cretaceous, Campanian; Belly River Group, Dinasaur Park Fm; Red Deer River, near Dinosaur Provincial Park) for David Evans of the Royal Ontario Museum. Specimen scanned by Matthew Colbert 13 May x bit TIFF images. P250D, 450 kv, 1.3 ma, small spot size, 1 brass filter, air wedge, 190% offset, integration time 32 ms, slice thickness = 0.25 mm, S.O.D. 585 mm, 1600 views, 1 ray per view, 1 sample per view, inter-slice spacing = 0.23 mm, field of reconstruction mm (maximum field of view mm. Field of view reported on scan form corresponds to a corrected value after applying a correction of %. Field of view used in ACTIS equals mm. Reported maximum field of view reflects uncorrected value. This scan was originally archived with an incorrectly reported field of view, which was corrected in August, 2010.), reconstruction offset 5800, reconstruction scale Streak- and ring-removal processing done by Satomi Ishihama based on correction of raw sinogram data using IDL routines RK_SinoDeStreak with parameter cutoff=0.08 and RK_SinoRingProcSimul with parameters bestof5=11 (applied to some slices) and nuke_inds=[ , ] (applied to slices 208, ). Reconstructed with rotation angle = 3 degrees. Postreconstruction ring correction done by Satomi Ishihama using UTCT ring correction program with the parameters oversample=3.0, binwidth=19, sectors=60 for slices 208, , and with parameters oversample=3.0, binwidth=23, sectors=1 for select slices. Total final slices = 376.

90 81 Chapter 3 Systematics and Ontogeny of Foraminacephale brevis gen. nov. (Ornithischia: Pachycephalosauria) 1 Introduction For nearly a century the pachycephalosaur taxon Prenocephale brevis (originally Stegoceras brevis Lambe, 1918), from the Belly River Group of Alberta, has been known only from lated frontoparietals. Due largely to the lack of a complete parietosquamosal bar and a poor understanding of pachycephalosaur ontogeny, this taxon has been highly problematic in terms of both it taxonomy and systematic placement. It has been synonymized with Stegoceras validum (Brown and Schlaikjer, 1943; Maryańska et al., 2004), but also recognized as a distinct taxon in two different genera, originally Stegoceras (Lambe, 1918; Sternberg, 1945; Williamson and Carr, 2002) and more recently Prenocephale (Sullivan, 2000, 2003, 2006). Most recently this taxon and been referred to as Sphaerotholus brevis by Longrich et al. (2010). Such differences in taxonomy can have a significant effect the interpretation of large-scale diversity changes and patterns of dinosaur evolution. As no consensus has been reached on its taxonomic or systematic placement a revision of this taxon is necessary. Here I identify and describe the first peripheral elements assignable to 'Prenocephale' brevis, including the first squamosal. These elements provide information on key parts of the morphology of this taxon that have previously been lacking. Using detailed laser surface scans of these peripheral bones and complete frontoparietal domes, a 3D reconstruction of the posterior peripheral margin of the skull dome has been assembled that provides the most complete picture of this taxon to date. In addition, I identify and describe the first juvenile and adult specimens of 'P'. brevis and, with the addition of these specimens, produce the first hypothetical growth series

91 82 for the taxon. Quantitative analyses of frontoparietal ontogeny using allometric growth curves and cranial histology are used to test the hypothesis that this growth series represents the ontogeny of a single taxon distinct from the other pachycepalosaur taxa present in the Belly River Group assemblage. The results presented here, in combination with previous work (Sereno, 2000; Williamson and Carr, 2002; Sullivan, 2003; Schott et al., 2009), are used to create a new pachycephalosaur character matrix for phylogenetic analysis of this clade. 1.1 Institutional Abbreviations AMNH, American Museum of Natural History, New York, USA; CMN, Canadian Museum of Nature, Ottawa, Canada; ROM, Royal Ontario Museum, Toronto, Canada; TMP, Royal Tyrrell Museum of Palaeontology, Drumheller, Canada; UALVP, University of Alberta Laboratory of Paleontology, Edmonton, Canada; UCMZ(VP), University of Calgary Museum of Zoology, Vertebrate Palaeontology Collection, Calgary, Canada; Z. PAL., Palaeozoological Institute, Warsaw, Poland. 2 Taxonomic and Systematic History of Foraminacephale brevis 'Prenocephale' brevis was originally named by Lambe (1902) as Stegoceras brevis with the frontoparietal dome CMN 1423 as the holotype. Brown and Schlaikjer (1943) synonymized S. brevis with Troödon (=Stegoceras) validus as they considered the differences between the taxa to be due to sex and age rather than specific. They identified specimens previously referred to S. brevis to be juvenile females of T. validus (Brown and Schlaikjer 1943). Sternberg (1945) resurrected both the genus Stegoceras and the species Stegoceras brevis, which he diagnosed, in part, based on the small, short, and moderately thick dome with a down-turned posterior extension of the parietal and pitted dorsal surface texture. Sternberg (1945) noted distinct

92 83 differences in the morphology of the parietal and timing of the closure of the sutures between S. validum and S. brevis (the parietal was always down-turned in S. brevis and never in S. validum, and the sutures in specimens of S. validum were open, while the same sutures were closed in smaller specimens of S. brevis). Chapman et al. (1981) revived the hypothesis that Stegoceras represented a single species and that the morphological variation was due to age and sex, based on the results of their morphometric study. Their study assumed a priori that the Stegoceras specimens (excluding S. edmontonensis) all represented a single species, apparently based on the taxonomic assignments of Wall and Galton (1979) who only mentioned S. validum and not the other three potential species of this genus (for further critiques of this morphometric study see Chapter 1). Following the results of Chapman et al. (1981), Sues and Galton (1987) synonymized S. brevis (along with S. lambei and S. sternbergi) with S. validum. Sullivan (2000) resurrected the species 'Stegoceras' brevis and transferred it to the genus Prenocephale based on several generic characters, including the lack of a parietosquamosal shelf, closed supratemporal fenestrae, and having the medial-most posterior nodes on the downturned parietal; and several specific characters including the medial-most posterior nodes situated fully on the parietal and a sloped anterior frontal with three distinct lobes. Williamson and Carr (2002) recognized S. brevis as a distinct species, but did not support its assignment to the genus Prenocephale. They presented the first phylogenetic analysis to include this taxon, and recovered a monophyletic Stegoceras with S. brevis as the sister taxon to a polytomous clade consisting of S. validum, S. lambei and S. sternbergi (Williamson and Carr, 2002). Sullivan (2003) presented an alternate phylogeny which, while it did not recover a monophyletic Prenocephale, found 'P'. brevis to be the sister taxon of P. edmontonensis, which in turn were the sister group to P. goodwini. Based on this phylogeny, Sullivan (2003) continued to support his generic assignment of Prenocephale brevis. Additionally, Sullivan (2003) proposed a new

93 84 unnamed species of Prenocephale for the specimen TMP , a specimen that Williamson and Carr (2002) identified as a possible adult S. brevis. Sullivan (2003) rejected this possibility, but provided no supporting argument other than the large size of the specimen. Maryańska et al. (2004) did not recognize any species of Stegoceras other than S. validum in their study and phylogenetic analysis, but no argument was presented to justify the synonymy of the taxa. Ryan and Evans (2005) recognized S. brevis as a distinct species, but chose to follow a more conservative taxonomy and thus did not recognize the generic asssingment of Sullivan (2003). Additionally, they referred to, and figured, TMP as an adult S. brevis, presumably following Williamson and Carr (2002), although no discussion of this specimen was provided. Sullivan (2006) continued to support the generic placement of P. brevis in Prenocephale. While he claimed that there is no demonstrable growth series for P. brevis, he did not further discuss TMP , nor did he provide an updated phylogenetic analysis that included this specimen (Sullivan, 2006). 3 Systematic Paleontology Dinosauria Owen, 1842 Ornithischia Seeley, 1887 Pachycephalosauria Maryańska and Osmólska, 1974 Foraminacephale gen. nov. Etymology Combination of forāmina (Latin: foramina) and cephale (Greek: head), referring to the numerous foramina the cover the dorsal surface of the skull of this taxon. Type Species Foraminacephale brevis Lambe, Diagnosis Same as for species.

94 85 Foraminacephale brevis (Lambe, 1918) Holotype CMN 1423, fairly complete frontoparietal dome missing the anterior portion of the frontonasal boss. Type Locality and Horizon Dinosaur Park Formation, Dinosaur Provincial Park, Alberta. Newly Referred Specimens Squamosal: UALVP 49440; postorbitals: TMP , ; parietals: CMN 12351, TMP ; frontoparietals: TMP Locality and Horizon TMP : Dinosaur Park Formation (DPF), Dinosaur Provincial Park (DPP), Alberta, Canada. TMP : DPF, DPP, Alberta Canada. TMP : DPF, Steveville, Alberta, Canada. TMP : DPF, DPP, Alberta, Canada. UALVP 49440: DPF, Happy Jack's Locality, DPP, Alberta, Canada. Hypodigm See Sullivan (2003) and Appendix 2. Diagnosis Pachycephalosaur dinosaur characterized by the following autapomorphies: tall, smooth squamosal bar, single row of six squamosal nodes with single corner node just ventral to node row, distinct pitted surface texture on dorsal surface of frontoparietal and associated peripheral elements. Differs further from Stegoceras validum, Hanssuesia sternbergi, and Colepiocephale lambei in the following characters: 'down-turned' medial extension of parietal and closure of supratemporal fenestrae very early in ontogeny. Differs further from Sphaerotholus and Prenocephale in the following characters: grooves on frontal and slit-like temporal chamber.

95 86 4 Description Here I describe the first squamosal and postorbitals, as well as the first juvenile and adult material, referable to Foraminacephale brevis. With this new material I present a 3D reconstruction of skull of F. brevis and the first ontogenetic growth series for this taxon. 4.1 Peripheral Elements The squamosal is one of the most diagnostic bones in pachycephalosaurs, with most species identifiable solely on the basis of its morphology and associated ornamentation. Several of the most contentious pachycephalosaur taxa from the Campanian of North America have been erected on the basis of lated frontoparietals, while diagnostic peripheral bones remain unknown. Previously, Foraminacephale brevis was only known from lated frontoparietal domes that have a distinctive pitted surface texture and morphological shape, but here we identify and describe the first peripheral elements assignable to this taxon, including the squamosal and its associated ornamentation. These are identified as pertaining to F. brevis based on the morphology of the sutural contacts with the frontoparietal dome and a shared pitted surface texture that is characteristic of the taxon. This provides a more complete understanding of the morphology of F. brevis, as well as critical information on the ornamentation of this taxon, which will help to resolve the taxonomy and phylogenetic placement of this species Squamosal An lated squamosal (UALVP 49440; Fig. 4.1) is identified here as pertaining to Foraminacephale brevis. The squamosal is nearly complete, missing only parts of the ventral processes. It is most similar to the squamosals of Sphaerotholus buchholtzae (TMP ) and Prenocephale prenes (Z. Pal. MgD-I/104). The squamosal has sutural surfaces for the

96 87 parietal medially and anteromedially and the postorbital anteriorly. The sutural surface is continuous indicating that the supratemporal fenestrae were fully closed. The position and angle of the sutural surface is most similar to P. prenes (Z. Pal. MgD-I/104) and S. buchholtzae (TMP ) and would have articulated with the frontoparietal of F. brevis with the parietosquamosal bar at an angle in posterior view. The dorsal surface of the squamosal is fairly continuous with the dome, except for a dorsal expansion at the anterolateral edge. The squamosal of P. prenes (Z. Pal. MgD-I/104) has a similar expansion, although to a lesser degree, while the dorsal surface of S. buchholtzae (TMP ) lacks this feature. The dorsal surface of UALVP is very smooth, but is pierced by several small foramina that are identical to those found on the peripheries of F. brevis domes. Similar foramina are also present on the dorsal surfaces of the dome and peripheral elements of Sphaerotholus buchholtzae (TMP ), except to a much lesser degree, whereas P. prenes (Z. Pal. MgD-I/104) has a much more rugose surface texture, somewhat similar to the tuberculate texture of Stegoceras (e.g., CMN 138, TMP , UALVP 2), but also pierced by foramina. The dorsal surface of the squamosal is separated from the posterior squamosal bar by a row of six similarly sized nodes, including the vertex node. The medial-most posterior node straddles the parietosquamosal suture. The posterior nodes are somewhat flattened posteriorly and pointed dorsally, similar to S. buchholtzae (TMP ) and unlike the posteriorly projecting conical nodes of P. prenes (Z. Pal. MgD-I/104) or the pyramidal nodes of S. validum (e.g., CMN 138, TMP , UALVP 2). The vertex node separates the posterior node row from the lateral node row that, on the squamosal, consists of a single elongated node. A similar, albeit less pronounced, lateral node is found on the squamosal of S. buchholtzae (TMP ), which also has a second lateral node that is even more elongate and less pronounced. A mediolateral crack, and associated strip of missing bone, lies along the ventral edge of the

97 88 posterior node row and obscures the ventral bases of these nodes. Ventral to the vertex node is a lateroventral corner node, as in Prenocephale (Z. Pal. MgD-I/104) and Sphaerotholus (TMP , NNMH P-27403). This node is slightly smaller than the posterior nodes, but much larger than the tiny corner node of S. buchholtzae (TMP ). In contrast, the corner node of P. prenes (Z. Pal. MgD-I/104) is much larger than the posterior nodes. Medial to the corner node is a dorsally angled row of three small, flat, node-like structures. This row may have continued, but is obscured by the strip of missing bone ventral to the primary node row. Other than the corner node and the flat, node-like structures, the squamosal bar is smooth, devoid of the minute nodes found in Stegoceras (e.g., CMN 138, TMP , UALVP 2), and pierced by many small foramina adjacent to its ventral margin. The squamosal bar is proportionally much taller than that found in any other pachycephalosaur and maintains the same height throughout its length. This is most similar to P. prenes (Z. Pal. MgD-I/104), which has a tall squamosal bar, but proportionally is not as tall as the much smaller squamosal of Foraminacephale brevis, and increases slightly in height laterally. In contrast, the squamosal of S. buchholtzae (TMP ) is very short and further shortens laterally. Ventrally, the squamosal bars of F. brevis and S. buchholtzae (TMP ) are very similar. Both curve dorsally and thin laterally. Most of the ventral surface is missing and the ventral processes of the squamosal are not preserved. The preserved posterior section of the temporal chamber has the slit-like morphology typical of F. brevis and Stegoceras (e.g., CMN 138, CMN 515, TMP , UALVP 2) that is distinct from the open morphology of S. buchholtzae (TMP ) and other pachycephalosaurs.

98 Postorbital Two postorbitals are here referred to Foraminacephale brevis, TMP and TMP (Fig. 4.2). They are both nearly complete missing only the posterior portion of the lateral bar (both) and the anterior portion of the lateral bar (TMP ). Since TMP is more complete the description will focus mainly on this specimen, but significant differences will be noted as appropriate. The medial portion of the postorbital is convex, almost forming a secondary dome. This is similar to the condition found in P. prenes (Z. Pal. MgD-I/104) and unlike the flat dorsoventrally oriented wall formed by the postorbital of S. buchholtzae (TMP ). The dorsal surface is smooth, but pierced by the same small foramina as the squamosal and frontoparietal of Foraminacephale brevis. Again, this differs from the somewhat rugose texture of the postorbital of P. prenes (Z. Pal. MgD-I/104), which still bears small foramina, and the smooth surface of the postorbital in S. buchholtzae (TMP ) that lacks the high quantity and distinctness of the foramina. The sutural surfaces surrounding this portion of the postorbital are complete with contacts for the squamosal posteriorly, frontoparietal medially, and posterior supraorbital anteriorly. The sutural contacts are oriented such that the postorbital would be mostly confluent with the dome, similar to the condition in P. prenes (Z. Pal. MgD-I/104). The two specimens of F. brevis are similar in the length and height of the medial portion, but TMP is much thicker dorsoventrally. In medial view, the dorsal border of TMP is anteroposteriorly convex whereas TMP is much more flat. Lateroventrally to the medial domed portion of the postorbital is the lateral bar. It is separated from the medial portion by a low anteroposterioly oriented ridge. This ridge corresponds to the lateral node row. Although no lateral nodes are present on these postorbitals, both are missing the posterior portion of the bar that contacts the squamosal, which, as described

99 90 above, has a single lateral node. Thus it is possible that one or more nodes would have been present on the missing portion of the postorbital. The ridge in S. buchholtzae (TMP ) is less distinct than in Foraminacephale brevis, but its postorbital also lacks lateral nodes. In P. prenes (Z. Pal. MgD-I/104) the ridge bears four lateral nodes posteriorly. The anterior portion of the ridge in P. prenes (Z. Pal. MgD-I/104) is formed by small tubercles. A second ridge, more prominent in TMP , runs anteroventrally from roughly the midpoint of the main ridge to the anteroventral edge of the postorbital. This ridge is absent in S. buchholtzae (TMP ), but in P. prenes (Z. Pal. MgD-I/104) extends from the anteriormost lateral node and is formed by partially connected nodes or tubercles. The surface texture of the lateral bar is somewhat smooth, but bears small tubercles concentrated around the second ridge. The posteroventral portion of the bar is pierced by several small foramina, but lacks the larger foramina found on the medial portion. Anteriorly, the lateral bar forms the posterior margin of the orbit. This is somewhat concave in TMP , but straight in TMP The posterior margin of the orbit is not preserved in S. buchholtzae (TMP ), but in P. prenes (Z. Pal. MgD-I/104) the morphology is very similar to TMP Posterior to the orbital margin, the lateral bar extends ventrally to contact the jugal and form the postorbital-jugal bar. Posterior to the bar, the postorbital forms the anterodorsal margin of the lateral temporal fenestra. This margin is concave, more so in TMP than TMP In S. buchholtzae (TMP ) the margin is not as strongly concave and in P. prenes (Z. Pal. MgD-I/104) it is only slightly concave. The length of the lateral bar could not be ascertained in either specimen due to breakage, but the heights of the bars could be compared. Whereas the medial portions of the two postorbitals were roughly the same height the lateral bar of TMP is much taller.

100 91 In ventral view, the postorbital contains part of the orbital cavity anteriorly. Medial to the orbital cavity is a notch for the articulation of the jugal. Medially to the jugal notch is the roof of the temporal chamber and at the posterior edge of the postorbital is the notch for the squamosal D Reconstruction Detailed 3D laser surface scans of the lated squamosal (UALVP 49440), one of the postorbitals (TMP ), and two nearly complete frontoparietal domes (CMN 121 and TMP ) were taken by Arius 3D, scaled to the same size, and articulated in three dimensions based on the morphology of their sutural surfaces. This produced a 3D reconstruction of the skull of Foraminacephale brevis which provides the most complete picture of the taxon to date (Fig. 4.3). The squamosal has a primary row of six nodes on its posterior border with the medialmost node straddling the parietosquamosal suture. The posterior region of the squamosal is devoid of the minute nodes found in S. validum (e.g., CMN 138, TMP , UALVP 2), but does have a prominent lateroventral corner node, as in Prenocephale (Z. Pal. MgD-I/104) and Sphaerotholus (TMP , NNMH P-27403). The squamosal bar ventral to the node row is much taller than that found in any other pachycephalosaur and maintains the same height throughout its length. In posterior view, the squamosal is distinctly sloped ventrolaterally, as in Sphaerotholus (TMP , NNMH P-27403) and Prenocephale (Z. Pal. MgD-I/104), but maintains the same depth throughout its width. The postorbital is narrow and tall, as in Sphaerotholus buchholtzae (TMP ), but more convex dorsally, as in Prenocephale prenes (Z. Pal. MgD-I/104). The postorbital bar is smooth and devoid of nodes, as in Sphaerotholus (TMP , NNMH P-27403). The overall morphology is very similar to Prenocephale-Sphaerotholus grade taxa (e.g., smooth squamosal bar, corner node, downturned

101 92 parietal), but also contains several characters shared with Stegoceras (e.g., more than five primary nodes, groves in frontal, slit-like temporal chamber). 4.2 Juvenile Material Several new small specimens are here referred to Foraminacephale brevis, including two small parietals and a frontoparietal dome that each show the diagnostic characters of F. brevis, but exhibit distinct juvenile morphologies Parietals Two small, lated parietals (CMN and TMP ; Figs. 4.4 and 4.5) are here referred to Foraminacephale brevis and represent the smallest known specimens of this taxon. They are both nearly complete, missing only portions of the ventral processes. In TMP , the ventral processes are broken at their bases, but in CMN most of the processes are preserved, particularly on the left side where it is nearly complete. TMP is well preserved, but CMN is water worn. The parietals are thickened, but only slightly domed with both mediolateral and anteroposterior convexity. The dorsal surface texture of TMP is somewhat more rugose than in larger specimens of Foraminacephale brevis (e.g., TMP ) whereas CMN is smooth due to the water worn nature of the specimen. Both are covered with small foramina, which although smaller, are equivalent to those in larger specimens. Anteriorly, the parietal bears the sutural surface for articulation with the frontals. The sutural surface for the postorbital is preserved at the anterolateral edge of the parietal. This surface is extremely small, but proportionally not very different than in larger specimens. Posterior to the postorbital sutural surface is finished bone surface that corresponds to the anteromedial margin of the supratemporal fenestra. In all larger specimens (except for CMN 12351) the supratemporal

102 93 fenestra are closed and this area of the parietal instead bears sutural surface for the squamosal. The margin of the supratemporal fenestra angles strongly medially until it straightens out posteriorly where it meets the medial extension of the parietal. The size of the open supratemporal fenestra would have been proportionally large as the margin encompasses more than half the total length of the parietal. The lateral edges of the medial extension of the parietal are straight and formed at the anterior end by the margin of the supratemporal fenestra and posteriorly by the sutural surface for the squamosal. The sutural surface for the squamosal is missing in TMP , but is discernible in CMN The medial extension of the parietal is not as downturned as in larger specimens of Foraminacephale brevis (e.g., TMP ), but rather continues the slight curvature of the parietal ventrally. Despite this, the morphology of the supratemporal border and medial extension of the parietal are clearly distinguishable from the flat parietals of S. validum (e.g., UCMZ(VP) , UCMZ(VP) ; see Chapter 1). These parietals all have triangular medial extensions and the supratemporal fenestrae are much smaller and circular. In lateral view, CMN reveals the low, slit-like temporal chamber characteristic of F. brevis and Stegoceras (e.g., CMN 138, CMN 515, UALVP 2). TMP has a similar morphology although the ventral process of the parietal is broken at its base. In CMN 12351, the ventral process of the parietal expands ventrally and laterally to form part of the ventral surface of the temporal chamber. On the left side of the parietal the process is nearly complete and a portion of the sutural surface for the squamosal, which would have contacted with the lateral edge of the ventral process, is discernible, albeit heavily water worn.

103 Frontoparietals Two nearly complete frontoparietals, TMP (Figs. 4.6 and 4.7) and TMP (Figs. 4.8 and 4.9), are identified here as juvenile Foraminacephale brevis due to their small size and distinct morphologies. These specimens were previously referred to F. brevis by Sullivan (2003), but their juvenile morphology was not noted or described. The specimens are larger than the small lated parietals described above, but smaller than other specimens of F. brevis. These frontoparietals have domed parietals with closed supratemporal fenestrae, but anteriorly uninflated frontals. Of the two specimens described below, TMP is slightly smaller (frontoparietal length 57.1 mm; compared to 61.1 mm for TMP ), but thicker (frontoparietal thickness 27 mm, compared to 24 mm for TMP ). TMP has thinner frontals that are almost flat (height of nasal suture 2.1 mm compared to 4.3 mm in TMP ). In dorsal view, the specimens are easily identifiable as Foraminacephale brevis due to the general morphology of the frontoparietal, the presence of foramina typical of F. brevis, and the downturned medial extension of the parietal with slit-like temporal chambers. The dorsal surface texture of the frontopariel is somewhat more rugose than in larger specimens (e.g., TMP ), especially in TMP This is partially due to the many furrows that run along the surface of the parietal that appear to be pathological. The frontals, which lack these canals, still show some external rugosity. In dorsal view, the frontonasal boss is identifiable through anterior advancement beyond the supraorbital lobes that flank it on either side. In TMP , the frontonasal boss is almost completely flat except that it is angled dorsally and therefore thickens posteriorly. The frontonasal boss and the supraorbital lobes have a minor amount of mediolateral convexity resulting in 'incipient grooves' that make distinction between the two structures possible. In TMP , the frontonasal boss is less flat with a stronger

104 95 dorsal angle and more mediolateral convexity such that the frontal grooves that separate the boss from the supraorbital lobes are discernible. In both specimens, the supraorbital lobes are largely flat and separate from the posteromedial doming of the frontal, which thickens posteriorly before it contacts the well-domed parietal. The interfrontal suture is visible both dorsally and ventrally, whereas the frontoparietal suture is visible ventral and laterally, but indiscernible dorsally. The morphology of the parietal in TMP is the same as that in larger specimens of Foraminacephale brevis. It has the same sloping parietal with a downturned posteromedial extension bearing two partial nodes (although the nodes are mainly restricted to the squamosal in this specimen). The sutural surfaced for the postorbital and squamosal are the same as is the slitlike temporal chamber. TMP conforms to these features except that the sutural surface for the squamosal is extremely short. It seems likely that this is due the recent closure of the supratemporal fenestra in this specimen and that the suture would further increase in height through ontogeny with inflation of the dome. Evidence to support this may be found on the right side of the parietal where a portion of the supratemporal fenestra may still be open. However, due to breakage in this area, it is difficult to be certain. Beyond the short squamosal suture, TMP has the same downturned parietal bearing two small nodes and slit-like temporal openings that support its referral to F. brevis. 4.3 Adult Material The majority of Foraminacephale brevis specimens appear to be subadult and exhibit the morphology that has typically been attributed to this taxon (e.g., TMP , Figs and 4.11; Williamson and Carr 2002; Sullivan 2000, 2003). One specimen, TMP (Figs and 4.13), is identified here as an adult F. brevis. Previously, Williamson and Carr (2002) suggested TMP might be an adult representative of F. brevis, but noted that the

105 96 specimen was rather large, with a high frontonasal boss, and lacked on nodes on the medial extension of the parietal. Sullivan (2003) assigned this specimen to the genus Prenocephale, correctly identified the presence of partial nodes on the posteromedial extension of the parietal, and noted several differences between this specimen and Sphaerotholus buchholtzae (TMP ). He did not support the referral of this specimen to F. brevis, but this was apparently solely on the basis of size, as no characters were described to distinguish TMP from F. brevis. Ryan and Evans (2005) figured this specimen as an adult Stegoceras brevis, but gave no comment on its morphology. TMP consists of a nearly complete frontoparietal dome missing a section of the dorsal surface near the centre of the dome, the ventral processes, and portions of the lateral sutural surfaces. The specimen is much larger than other specimens referred to this species (frontoparietal length, 95.1 mm; compared to 73.1 mm for UALVP 8508, the next largest specimen). In general, the morphology of this specimen is consistent with Foraminacephale brevis. It has the same dorsal outline and all the characters used to diagnose this species including the down-turned parietal with slit-like temporal chamber and smooth dorsal surface texture punctured by many small foramina. However, there are several morphological differences that support the identification of TMP as an adult F. brevis. The foramina in TMP are less numerous than in smaller specimens of Foraminacephale brevis (e.g., TMP , TMP ) and are further restricted to the periphery. The frontonasal boss is more inflated and convex than in smaller specimens, and the supraorbital lobes are further incorporated into the dome, but still maintain mediolateral convexity. Distinct grooves separate the supraorbital lobes from the frontonasal boss as in all other specimens of F. brevis. Posteriorly, the dome is further inflated, which results in a stronger slope of the parietal. The medial extension of the parietal maintains the same down-turned

106 97 morphology of all other F. brevis specimens. The nodes on the posteromedial extension of the parietal would have been almost fully on the squamosal, as only slight depressions indicating the bases of the nodes are visible on the lateral edge of the extension. In lateral view, the morphology of the sutural surfaces is largely consistent with other specimens of F. brevis (e.g., TMP , TMP ), except for the increased inflation of the supraorbital lobes and frontonasal boss. Additionally, the temporal chamber maintains a slit-like appearance in lateral view.

107 98 Figure 4.1 Squamosals of Foraminacephale brevis (UALVP 49440; top), Sphaerotholus buchholtzae (TMP ; middle), Stegoceras validum (TMP ; bottom) in posterior view.

108 99 Figure 4.2 Postorbitals of Foraminacephale brevis (TMP ; top), Sphaerotholus buchholtzae (TMP ; middle), Stegoceras validum (TMP ; bottom) in left lateral view.

109 100 Figure 4.3 3D reconstruction of Foraminacephale brevis based on the lated squamosal (UALVP 49440), one of the postorbitals (TMP ), and two nearly complete frontoparietal domes (CMN 121 and TMP ) in: A, dorsal; B, ventral; C, lateral; D,

110 101 posterior; E, posterolateral; and F, anteroventral views. Reconstructions were created using detailed laser surface scans of the bones that were then scaled to the same size. This reconstruction provides the most complete picture of this taxon to date.

111 102 Figure 4.4 Juvenile parietal of Foraminacephale brevis, TMP , in dorsal (top) and ventral (bottom) views. super temporal fenestra. Abbreviations: mep, medial extension of parietal; p, parietal; po, postorbital; sq, squamosal; ss, sutural surface for; stf, supratemporal fenestra.

112 103 Figure 4.5 Juvenile parietal of Foraminacephale brevis, TMP , in lateral (top) and posterior (bottom) views. Abbreviations: f, frontal; mep, medial extension of parietal; p, parietal; po, postorbital; ss, sutural surface for; stf, supratemporal fenestra.

113 104 Figure 4.6 Juvenile frontoparietal of Foraminacephale brevis, TMP , in dorsal (top) and ventral (bottom) views. Abbreviations: f, frontal; fp, frontoparietal; if, interfrontal; mep, medial extension of parietal; p, parietal; po, postorbital; sq, squamosal; ss, sutural surface for.

114 105 Figure 4.7 Juvenile frontoparietal of Foraminacephale brevis, TMP , in lateral (top) and posterior (bottom) views. Abbreviations as in Figure 4.6.

115 106 Figure 4.8 Juvenile frontoparietal of Foraminacephale brevis, TMP , in dorsal (top) and ventral (bottom) views. Abbreviations as in Figure 4.6.

116 107 Figure 4.9 Juvenile frontoparietal of Foraminacephale brevis, TMP , in lateral (top) and posterior (bottom) views. Abbreviations as in Figure 4.6.

117 108 Figure 4.10 Subadult frontoparietal of Foraminacephale brevis, TMP , in dorsal (top) and ventral (bottom) views. Abbreviations as in Figure 4.6.

118 109 Figure 4.11 Subadult frontoparietal of Foraminacephale brevis, TMP , in lateral (top) and posterior (bottom) views. Abbreviations as in Figure 4.6.

119 110 Figure 4.12 Adult frontoparietal of Foraminacephale brevis, TMP , in dorsal (top) and ventral (bottom) views. Abbreviations as in Figure 4.6.

120 111 Figure 4.13 Adult frontoparietal of Foraminacephale brevis, TMP , in lateral (top) and posterior (bottom) views. Abbreviations as in Figure 4.6.

121 112 5 Frontoparietal Variation and Ontogeny Sullivan (2006) claimed that there is no known growth series for Foraminacephale brevis, however, I have identified specimens with both juvenile and adult morphologies that differ from the previously described morphology for F. brevis (Sullivan, 2000, 2003, 2006), but which are consistent within an ontogenetic growth series for this taxon (Fig. 5.1). In order to test the hypothesized growth series of F. brevis, growth of the frontoparietal is analyzed allometrically in F. brevis, and compared to the allometric growth in the other three Belly River Group taxa. Additionally, variation in frontoparietal morphology is analyzed quantitatively using multivariate morphometrics. This allows me to quantitatively describe the ontogeny of F. brevis and distinguish it from the ontogenetic trajectories in the other Belly River group taxa. I also note quantitative differences in the morphologies of these taxa, which support F. brevis, and the other three Belly River group taxa as distinct. 5.1 Frontoparietal Allometry In the previous chapter the frontoparietal ontogeny of Stegoceras validum was analyzed and it was found, contra Chapman et al. (1981), that the frontoparietal dome thickened dorsoventrally with positive allometry. Additional ontogenetic changes were identified, such as the frontoparietal widening posteriorly at a faster rate than anteriorly, which results in a triangular appearance of the dome in large specimens, and the negative allometric growth of the frontal with respect to the parietal. Here I perform an allometric analysis of the frontoparietals of Foraminacephale brevis, which allows me to test the hypothesis that F. brevis represents a distinct species within the Belly River Group assemblage. In addition, I test that hypotheses that the small and large specimens of F. brevis described above (TMP , TMP ) are juveniles and adults

122 113 of this species, respectively. Finally, we analyze and compare the growth of the frontoparietal in the Belly River Group taxa (Stegoceras validum, F. brevis, Colepiocephale lambei, and Hanssuesia sternbergi) in order to further test the hypothesis that they are distinct species Methods Linear measurements were taken from the majority of pachyceohalosaur specimens from the Belly River Group of Alberta. The specimens include both fossils and casts (when fossils were not available). A total of 21 linear measurements were identified between 27 homologous landmarks on the frontoparietal, based partly on those identified by Goodwin (1990; see Appendix 1). Landmarks were primarily defined by the junction of cranial sutures visible on the frontoparietal. Due to large amounts of missing data not all measurements were available on all specimens. A complete list, explanation, and illustration of the measurements can be found in Appendix 1. Specimens used and their measurements are shown in Appendix 2. Measurements were transformed logarithmically so that they fit the linear allometric growth function, which has the additional benefit of normalizing the distribution of the data. Regressions were calculated using reduced major axis regression (RMA, also know as standard major axis regression) implemented in the lmodel2 software package (Legendre, 2008; see also Legendre and Legendre, 2008) for the statistics program R. RMA has been considered an appropriate regression model for tests of allometry (Smith, 2009; Warton et al., 2006). Statistical significance of the allometric pattern was determined based on the confidence intervals (i.e., if the confidence interval encompasses a slope of one, it is statistically metric). Due to the varying completeness of specimens, we used four different standard (x) variables: frontoparietal length, frontoparietal width, parietal length, and frontal length. Frontoparietal length is perhaps the most intuitive index of size, however very few specimens

123 114 have complete frontoparietals. In order to include juvenile specimens and maximize sample size (and thus statistical power), we used frontal and parietal lengths and frontoparietal width as standard variables for comparns. In these cases, even lated frontals and/or parietals could be incorporated into the bivariate analyses. A detailed analysis of frontoparietal growth was performed for Foraminacephale brevis using the majority of variables available. Dodson (1975a) found that intraspecific allometries in a growth series of extant Alligator mississipiensis showed very high correlation coefficients and thus we consider a high correlation coefficient as support for the hypothesis that the specimens of F. brevis represent an ontogenetic series of a single species. Following this, presence of the proposed juvenile (TMP ) and adult (TMP ) specimens of F. brevis within the confidence limits for the regression lines would support the hypothesis that they represent juveniles and adults of F. brevis. A second analysis was performed comparing frontoparietal growth in the four potential Belly River group taxa, Foraminacephale brevis, Stegoceras validum, Hanssuesia sternbergi, and Colepiocephale lambei. Dodson (1975b) found that the growth of two sympatric species of Sceloporus was very similar, but that differences were detectable in characters that showed strong positive or negative allometry. Thus, we might expect similar results in the analysis of the Belly River Group assemblage, with major differences between species evident in measurements that are either strongly negatively or positively allometric Results: Foraminacephale brevis Results of the bivariate regressions for Foraminacephale brevis are given in Table 5.1 and graphs of select indices are presented in Figure 5.2. Regressions using parietal width as the standard variable have the highest sample size and thus the most statistical power. Comparns

124 115 using frontoparietal length and parietal length have lower sample sizes that may hinder the ability to detect positive and negative allometry statistically. The correlation coefficients are generally high (most are >0.9, see Table 5.1) which supports the hypothesis that the specimens of F. brevis are an ontogenetic series of a single taxon. Additionally, the proposed juvenile (TMP ) and adult (TMP ) specimens of F. brevis fall within the confidence intervals for the regression lines, and in most cases fall on the regression line (see the smallest and largest specimens in Fig. 5.2), which supports the hypothesis that they represent juveniles and adults of F. brevis. The height of the dome generally exhibits positive allometric growth (Table 5.1, Fig. 5.2A). The positive allometry is weakest when compared to frontoparietal width. The height of the frontonasal boss tends to show the strongest positive allometry, although the slope is not generally significantly greater than the other slopes. There is a general trend for the slopes of the heights to decrease posteriorly, but the differences in the individual slopes are generally not significant. Following this trend, the slope for thickness of the frontoparietal is generally the lowest. The growth of the width of the frontoparietal does not show a single definitive pattern (Table 5.1, Fig 5.2B). When compared to frontoparietal width some of the dome widths show metry (W:n/prf, W:prf/aso, W:pso/po), whereas the width across the frontoparietal at the contact between the anterior and posterior supraorbital suture shows negative metry, and the width across the frontoparietal at the contact between postorbital and squamosal sutures show positive allometry. When frontoparietal length is used as the standard variable the two anteriormost measurements show metry (W:n/prf and W:prf/aso), while the posterior measurements all show positive allometry. If parietal length is used as the standard variable, only the two posterior-most measurements are positively allometric.

125 116 The lengths of the frontoparietal sutures all exhibit statistically metric growth and in most cases the slopes are near one (Table 5.1, Fig. 5.2C&D). The length of the frontoparietal, and the frontals and parietal individually, show negative allometry when compared to frontoparietal width. The frontal and parietal show metric growth with respect to each other and to frontoparietal length Results: Belly River Group Comparns Results of the bivariate regressions comparing the four Belly River Group taxa are given in Table 5.2 and graphs of select indices are presented in Figures Comparns between the Belly River Group taxa were almost exclusively made using the width of the frontoparietal as the standard variable due to the small sample sizes for both Colepiocephale and Hanssuesia, which also lack specimens with complete frontals and/or parietals. Foraminacephale brevis, Stegoceras validum, and Colepiocephale lambei showed relatively high correlation coefficients (r > 0.8; see Table 5.2) for most variables, which support the hypothesis that they are distinct species. The correlation coefficients for Hanssuesia sternbergi are generally much lower (with many below 0.5, see Table 5.2). This is largely due to the very small size range for this taxon (14 mm range in frontoparietal width compared to 67 mm for S. validum), but also suggests the possibility that the Hanssuesia sample represents more than one species, of which some specimens may be adult S. validum. Unfortunately, due to the small sample size, this possibility cannot be properly tested at this time. The height of the frontonasal boss (Table 5.2, Fig. 5.3A) showed similar positive allometry in all four taxa. Hanssuesia has a greater slope (5.48) than the other three taxa, but not significantly so, and with a very low correlation coefficient (r = 0.32). The height of the supraorbital lobes (Table 5.2, Fig. 5.3B) shows similar positive allometry in both

126 117 Foraminacephale brevis and Stegoceras validum (1.84 and 1.9, respectively). Hanssuesia has a higher slope (3.6) than the other taxa, but the difference is not significant and the correlation coefficient is very low (r = 0.03). Colepiocephale shows metric growth of the supraorbital lobe, but the slope (1.18) is not significantly lower than in the other taxa. Both F. brevis and S. validum have positive allometry for thickness of the frontoparietal compared to width of the frontoparietal (Table 5.2, Fig. 5.3C), but the slope in S. validum (2.72) is significantly greater than in F. brevis (1.84). Both Colepiocephale and Hanssuesia show metry. The results are the same if frontal length is used as the standard variable instead of parietal width (Fig. 5.3D). The allometry of the lengths of the sutures is similar between all taxa with all showing metry (Table 5.2, Fig. 5.4A C). The length of frontal is negatively allometric in Foraminacephale brevis and Stegoceras validum, but metric in Colepiocephale and Hanssuesia (Fig. 5.4D). The width of the frontonasal boss is metric in all taxa (Table 5.2, Fig. 5.5A). Stegoceras validum and Foraminacephale brevis show negative allometry for the width across the supraorbital lobes (Fig. 5.5B). Stegoceras validum also shows negative allometry of the width anterior to the frontoparietal suture (W:pso/po; Fig. 5.5C). Both F. brevis and Hanssuesia show positive allometry for the width posterior to the frontoparietal suture (W:po/stf/sq) with the slope in Hanssuesia (4.82) being significantly greater than that in F. brevis (1.16; Fig. 5.5D). Overall, many of the allometric trends were similar between the four taxa and this is to be expected when comparing closely related species (Dodson 1975b). However, differences in the allometric coefficients (slopes) are expected for strongly allometric variables (Dodson 1975b). The thickness of the frontoparietal, a trait which shows strong positive allometry, was significantly different between Foraminacephale brevis and Stegoceras validum and S. validum and Colepiocephale lambei. Additionally, the slopes for the width posterior to the frontoparietal

127 118 suture were significantly different between F. brevis and Hanssuesia sternbergi. These differences further support the hypothesis that these samples represent distinct taxa within the Belly River Group assemblage. It is likely that increased sample sizes and ranges will allow additional differences in the allometries of these species to be detected. 5.2 Frontoparietal Variation in the Belly River Group Taxa Analysis of frontoparietal variation was done using a principal component analysis of the linear measurements of Foraminacephale brevis, Stegoceras validum, Hanssuesia sternbergi, and Colepiocephale lambei. This allows us to identify quantitative differences in the morphology of the four Belly River Group taxa and statistically test the hypothesis that they are distinct taxa Methods All analyses were performed on log-transformed measurements, as log-transformed morphometric data fit a linear function (the allometric growth function) and so will not violate the assumption of linearity inherent in a principal components analysis (PCA). Because standard multivariate techniques cannot accommodate missing data, a subset of the available measurements and specimens were used that balanced sample size and the amount of morphological variation captured by the measurements. A list of measurements and specimens used can be found in Appendix 3. This data was analyzed using a PCA on both covariance and correlation matrices implemented and the statistical program R. In order to statistically test the morphological distinctness of each taxa, a multivariate analysis of variance (MANOVA) was performed in the statistical program PAST (Hammer et al. 2001).

128 Results The results of the PCA on the Belly River Group taxa are shown in Figure 5.6. The first principal component explains 82.1% of the variation in the covariance matrix with components 2 and 3 explaining 6.7% and 4.2% of the variation, respectively (Table 5.3). In the correlation matrix the first principal component explains 72.4% of the variation in the covariance matrix with components 2 and 3 explaining 8.0% and 6.8% of the variation, respectively (Table 5.4). All variables show negative loadings on the first component for the covariance and correlation matrices (Table 5.5 and 5.6), which suggests a strong size component. However, the height of the anterior portion of the frontoparietal loads most strongly on this component for the covariance matrix (Table 5.5), whereas in the correlation matrix the loadings are all very similar (Table 5.6). For the covariance matrix component 2 is most strongly influenced by the height of the posterior supraorbital postorbital sutural contact, followed by the height of the nasal suture and the length of the anterior supraorbital. Component 3 is most strongly influenced by the thickness of the frontoparietal followed by the height of the nasal suture, the prefrontal/anterior supraorbital sutural contact, and the width of the frontoparietal. For the correlation matrix components 2 is most strongly influenced by the width of the endocranial fossa followed by the length of the postorbital and then the length of the anterior supraorbital. The width of the endocranial fossa also loads most strongly on component 3, but follwed by the height of the posterior supraorbital postorbital sutural contact. The covariance matrix generally shows better separation of the taxa than the correlation matrix. Good separation of most taxa was obtained with the covariance plot of the first two principal components. However, there is no separation between Stegoceras validum and Colepiocephale lambei. The plot of the second and third principal components also shows

129 120 separation of the taxa with some overlap between Foraminacephale brevis and Hanssuesia and S. validum and Colepiocephale. The MANOVA found significant differences among the four taxa (Wilks λ = < 0.001, F = 25.4, p < 0.001) which supports the hypothesis that the four four taxa are distinct species. Unfortunately, the sample sizes were too small for pairwise comparns between the species.

130 121 Table 5.1 Allometric regression of Foraminacephale brevis frontoparietal ontogeny. Abbreviations: CI, 95% confidence interval;, metry; +, positive allometry; o, negative allometry; H:n/n, height of the sutural surface at the contact of the nasals; H:n/prf, height of the sutural surface at the contact of the nasal and prefrontal; H:prf/aso, height of the sutural surface at the contact of the prefrontal and anterior supraorbital; H:aso/pso, height of the sutural surface at the contact of the anterior supraorbital and posterior supraorbital; H: pso/po, height of the sutural surface at the contact of the posterior supraorbital and postorbital; T:fp, thickness of the frontoparietal; L:aso, length of the supraorbital suture; L:pso, length of the posterior supraorbital suture; L:po, length of the postorbital suture; L:f, length of the frontal; L:p, length of the parietal; L:fp, length of the frontoparietal; W:n/prf, width between the nasal/prefrontal sutural contacts; W:prf/aso, width between the aso/pso supraorbital sutural contacts; W: aso/pso, width between anterior and posterior supraorbital sutural contacts; W:pso/po, width between the posterior supraorbital and postorbital sutural contacts; W:po/stf/sq, width between the contacts of postorbital suture and the supratemporal fenestrae or the squamosal suture if fenestrae are closed. Heights Frontoparietal Width N H:n/n 8 H:n/prf 8 H:prf/aso 14 H:aso/pso 15 H:pso/po 15 T:fp 15 R Slope CI Intercept CI Allometry Frontoparietal Length N R H:n/n H:n/prf H:prf/aso H:aso/pso 5 1 H:pso/po T:fp Slope CI Intercept CI Allometry Slope CI Intercept CI Allometry Parietal Length H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp N R

131 122 Frontal Length R Slope CI Intercept CI Allometry R Slope CI Intercept CI Allometry Frontoparietal Length N R L:aso L:pso L:po L:f L:p 5 1 Slope CI Intercept CI Allometry Slope CI Intercept CI Allometry H:n/n H:n/prf H:prf/aso H:aso/pso H:pso/po T:fp N Lengths Frontoparietal Width N L:aso 14 L:pso 15 L:po 16 L:f 9 L:p 10 L:fp 5 Parietal Length L:aso L:pso L:po L:f N R

132 123 Widths Frontoparietal Width N W:n/prf 9 W:prf/aso 11 W:aso/pso 12 W:pso/po 16 W:po/stf/sq 16 R Slope CI Intercept CI Allometry + Frontoparietal Length N R W:n/prf W:prf/aso W:aso/pso W:pso/po 5 1 W:fp 5 1 W:po/stf/sq Slope CI Intercept CI Allometry Slope CI Intercept CI Allometry + + Parietal Length W:n/prf W:prf/aso W:aso/pso W:pso/po W:fp W:po/stf/sq N R

133 124 Table 5.2 Allometric regressions comparing the frontoparietal ontogeny of the Belly River Group taxa. Abbreviations as in Table 5.1. Height of Frontonasal Boss N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry Height of Supraorbital Lobe N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry Thickness of Frontoparietal N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry + + Thickness of Frontoparietal vs Length of Frontal N R Slope CI P. brevis Stegoceras Colepiocephale Hanssuesia Intercept CI Allometry + + Length of Anterior Supraorbital N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry Length of Posterior Supraorbital N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry

134 125 Length of Postorbital N R Slope CI Intercept CI Allometry N R Slope CI Intercept CI Allometry Width of Frontonasal Boss N R P. brevis Stegoceras Colepiocephale Hanssuesia Slope CI Intercept CI Allometry Width Across Supraorbital Lobes N R Slope P. brevis Stegoceras Colepiocephale Hanssuesia CI Intercept CI Allometry Width Anterior to Frontoparietal Suture N R Slope P. brevis Stegoceras Colepiocephale Hanssuesia CI Intercept CI Allometry Width Posterior to Frontoparietal Suture N R Slope P. brevis Stegoceras Colepiocephale Hanssuesia CI Intercept CI Allometry + + P. brevis Stegoceras Colepiocephale Hanssuesia Length of Frontal P. brevis Stegoceras Colepiocephale Hanssuesia

135 126 Table 5.3 Importance of the first four components for the principal components analysis of the Belly River Group taxa with the covariance matrix. Comp. 1 Comp. 2 Comp. 3 Comp. 4 Standard Deviation Proportion of Variance Cumulative Proportion

136 127 Table 5.4 Importance of the first four components for the principal components analysis of the Belly River Group taxa with the correlation matrix. Comp. 1 Comp. 2 Comp. 3 Comp. 4 Standard Deviation Proportion of Variance Cumulative Proportion

137 128 Table 5.5 Loadings of variables for the first four components for the principal components analysis of the Belly River Group taxa with the covariance matrix. Variable Comp. 1 Comp. 2 Comp. 3 Comp. 4 H:n/n H:pfr/aso H:aso/pso H:pso/po W:n/pfr W:pfr/aso W:aso/pso W:pso/po W:f/p L:aso L:pso L:po L:ecf W:ecf L:f T:f/p

138 129 Table 5.6 Loadings of variables for the first four components for the principal components analysis of the Belly River Group taxa with the correlation matrix. Variable Comp. 1 Comp. 2 Comp. 3 Comp. 4 H:n/n H:pfr/aso H:aso/pso H:pso/po W:n/pfr W:pfr/aso W:aso/pso W:pso/po W:f/p L:aso L:pso L:po L:ecf W:ecf L:f T:f/p

139 130 Figure 5.1 Growth series of Foraminacephale brevis in dorsal (top) and lateral (bottom) views depicting the ontogeny in this taxon. A, TMP ; B, TMP ; C, TMP ; D, TMP ; and E, TMP

140 131 Figure 5.2 Bivariate logarithmic plots with RMA regression lines for selected variables of Foraminacephale brevis. A, heights of frontoparietal vs width across frontoparietal suture. Blue, height of nasal; red, height of frontoparietal at the contact of the anterior and posterior supraorbitals; green, thickness of frontoparietal. B, widths of frontoparietal vs width across frontoparietal suture. Blue, width of frontonasal boss; red, width across the supraorbital lobes; green width across the frontoparietal at the contact of the posterior supraorbital and postorbital sutures. C, lengths of frontoparietal vs width across frontoparietal suture. Blue, length of frontal; red, length of parietal; green length of frontoparietal. D, lengths of frontoparietal vs length of parietal. Blue, length of anterior supraorbital suture; red, length of posterior supraorbital suture; green, length of postorbital suture; orange, length of frontal. Measurements are log (mm).

141 132 Figure 5.3 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa (Foraminacephale brevis, red; Stegoceras validum, blue; Hanssuesia sternbergi, green; Colepiocephale lambei, yellow). A, height of nasal vs width of frontoparietal. B, height of supraorbital lobe (at contact between anterior and posterior supraorbital sutures) vs width of frontoparietal. C, thickness of frontoparietal vs width of frontoparietal. D, thickness of frontoparietal vs length of frontal. Measurements are log (mm).

142 133 Figure 5.4 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa (Foraminacephale brevis, red; Stegoceras validum, blue; Hanssuesia sternbergi, green; Colepiocephale lambei, yellow). A, length of anterior supraorbital vs width of frontoparietal. B, length of posterior supraorbital vs width of frontoparietal. C, length of postorbital vs width of frontoparietal. D, length of frontal vs width of frontoparietal. Measurements are log (mm).

143 134 Figure 5.5 Bivariate logarithmic plots with RMA regression lines comparing dome growth between the four Belly River Group taxa (Foraminacephale brevis, red; Stegoceras validum, blue; Hanssuesia sternbergi, green; Colepiocephale lambei, yellow). A, width of frontonasal boss (at contact between nasal and prefrontal sutures). B, width across supraorbital lobe (at contact between anterior and posterior supraorbital sutures). C, width anterior to frontoparietal suture (at contact between posterior supraorbital and postorbital sutures). D, width anterior to frontoparietal suture (at contact between postorbital suture and supratemporal fenestrae or squamosal suture if fenestrae are closed). Measurements are log (mm).

144 135 Figure 5.6 Principal component analysis of the Belly River Group pachycephalosaurs (Foraminacephale brevis, red; Stegoceras validum, blue; Hanssuesia sternbergi, green; Colepiocephale lambei, yellow).

145 136 6 Frontoparietal Histology Cranial histology has previously been examined in pachycephalosaurs by Brown and Schlaikjer (1943), Goodwin et al. (1998), Goodwin and Horner (2004), and Horner and Goodwin (2009). The first two studies described the cranial histology of a single specimen. The study of Goodwin and Horner (2004), however, used histological sections of frontoparietal domes from a multi-taxic size series to test functional hypotheses of the dome. Horner and Goodwin (2009) used histology of frontoparietals and cranial ornamentation as support for the hypothesized synonymy of Dracorex, Stygimoloch, and Pachycephalosaurus, but were unable to analyze the histology of all three taxa. While these studies were purely qualitative, the results suggested that the vascularity of the dome decreased with ontogenetic age (Goodwin and Horner 2004). In the previous chapter it was found, using High Resolution Computed Tomography (HRCT) scans, that a proxy for relative vascularity (relative void space) did in fact decrease with increasing ontogenetic development in Stegoceras validum. Here I use the same methodology to examine changes in relative vascularity in specimens of Foraminacephale brevis in order to test the hypotheses that the sample represents a growth series and, through comparns with the results for Stegoceras validum, a distinct species. 6.1 Methods High Resolution Computed Tomography (HRCT) scans were chosen as a non-destructive alternative to thin-sectioning for examining bone histology. While the benefits of a nondestructive alternative are obvious, the appropriateness of HRCT scans in replacing thinsectioning has not been tested and a detailed comparn of HRCT data to histological thinsections is still needed. HRCT scans were taken from five specimens at various stages in the hypothesize growth series presented in this paper (TMP , TMP , TMP

146 , TMP , TMP ). Scans were performed by the University of Texas High-Resolution X-ray CT Facility. Details of each scan are provided in Appendix 4. Due to the nature of both fossil specimens and CT data, vascularity could not be measured directly. Void space in the CT scans was chosen as a reasonable proxy for canal space, as any empty space in the specimens should be the result of canals through the bone. It cannot be known what proportion of the canals was actually occupied by vessels, but it is assumed that the two areas should be highly correlated. Additionally, because we are measuring relative changes the exact method used will have less of an impact as long as the correlation is consistent between the specimens. Relative void space was calculated using transverse sections from the HRCT data. Sections were chosen at the boundary of the posterior supraorbital and postorbital sutures through the frontal and just posterior to the frontoparietal suture through the parietal for five specimens of Foraminacephale brevis representative of each of the stages we have described for the growth series (TMP , parietal only; TMP ; TMP ; TMP ; TMP ). Consistent positioning of the slices ensures that the sections will be homologous between the five specimens and thus directly comparable. For TMP , the transverse section had to be created through resampling of the original scan data in the 3D imaging program Avizo because the specimen had to be CT scanned dorsoventrally due to its large size. In order to calculate relative void space, each slice was thresholded with the Huang method (Huang and Wang, 1995; Wang et al., 2002), which minimizes fuzziness in the 2-D grayscale histogram, implemented in ImageJ (Rasband, ; see also Abramoff et al., 2004). There are numerous different methods of thresholding and while some have been found to perform better than others, the best method is still highly dependent on the image being thresholded (Sezgin and Sankur, 2004). Thus, we chose the Huang method based on a survey of

147 138 different thresholding techniques as the method that best approximated what we interpret as the vascular (canal) space in the scan. In order to properly compare relative void space between specimens, a region of interest needed to be specified. This was determined using a rectangle of the same relative proportions drawn onto each threshold image at the base of the braincase immediately to the left of the interfrontal suture and extended so that the lop left corner met the dorsal edge of the frontal. This technique ensured that, despite differences in the shapes of the frontals, homologous areas were compared. This step was also used to eliminate potential artifacts, such as the effects of beam hardening. Thresholded images were cropped to this region and relative void space was calculated using the voxel counter plug-in for ImageJ (Rasband, ; see also Abramoff et al., 2004). TMP shows a number of prominent cracks in the CT scans that showed as void space in the thresholded image. Since these cracks are not indicative of vascular space, we added an additional step for this specimen where the cracks were removed and the relative void space recalculated. Both the uncorrected and corrected void space values are reported. 6.2 Results Methods and results for the calculation of relative void space are shown in Figures 6.1 and 6.2 and graphs of the growth of the frontoparietal showing the changes in relative void space through ontogeny are shown in Figures 6.3 and 6.4. Foraminacephale brevis showed very low percentages of void space, especially compared to what was found in the previous chapter for Stegoceras validum. The smallest specimen, TMP had the highest levels of void space (1.67%) and the largest specimen had the lowest (0.25%), although the range between these values was not considerable. The other three specimens did not show a consistent decrease in relative void space. Although they are fairly similar in size and it is expected that there be some

148 139 amount of individual variation. Slices through the frontal generally yielded higher relative void spaces than slices through the parietal, except for TMP , which had less relative void space in the frontal section. The trend toward decreased relative void space (vascularity) with increased size supports the hypothesis that the sample represents an ontogenetic series, while the considerably lower relative void spaces in Foraminacephale brevis at all sizes when compared to relative void spaces in Stegoceras validum, supports the hypotheses that the two growth series are representative of separate species.

149 140 Figure 6.1 Calculation of relative void-space for the Foraminacephale brevis growth series. A transverse section was taken from CT scan of the frontal at the contact of the posterior supraorbital and postorbital sutures. Scans were thresholded using the Huang method (Huang and Wang 1995). Void-space was counted from an area of interest (red square). A, TMP ; B, TMP ; C, TMP ; D, TMP

150 141 Figure 6.2 Calculation of relative void-space for the Foraminacephale brevis growth series. A transverse section was taken from a CT scan of the parietal just posterior to the frontoparietal suture. Scans were thresholded using the Huang method (Huang and Wang 1995). Void-space was counted from an area of interest (red square). A, TMP ; B, TMP ; C, TMP ; D, TMP ; E, TMP The value in parentheses for TMP represent the void space when the cracks above the brain case are remove and these are not indicative of vascular space.