Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
A phylogenetic analysis of oological characters: A case study of saurischian dinosaur relationships and avian evolution
(USC Thesis Other)
A phylogenetic analysis of oological characters: A case study of saurischian dinosaur relationships and avian evolution
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
A PHYLOGENETIC ANALYSIS OF OOLOGICAL CHARACTERS: A CASE STUDY OF SAURISCHIAN DINOSAUR RELATIONSHIPS AND AVIAN EVOLUTION by Gerald Grellet-Tinner A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (GEOLOGICAL SCIENCES) May 2005 Copyright 2005 Gerald Grellet-Tinner Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3180304 INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3180304 Copyright 2005 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS First and foremost, I would like to express my gratitude to all the members of my dissertation committee: D. Bottjer, S. Bottjer, L. Chiappe, F. Corsetti, A. Fischer, and M. Norell. Their guidance on different levels truly contributed to the completion of this dissertation: D. Bottjer, for sharing his knowledge on taphonomic processes; S. Bottjer, for her general comments; L. Chiappe for his guidance on subjects related to vertebrate paleontology, F. Corsetti for his insightful comments and the use of his carbonate laboratory; A. Fisher for sharing his vast knowledge in paleontology and geology; M. Norell for his support and sharing specimens paramount to this research. SEM work was accomplished at the Natural History Museum of Los Angeles, and TLM, PLM, and CL examinations at the Carbonate Laboratory of the Department of Earth Sciences at USC. I would also like to acknowledge many of co-authors and colleagues with whom I have worked during the last four years, particularly among those: P. Makovicky, G. Dyke. Finally, this research would have not been possible without the help and contributions through access to collections and lending specimens of the following individuals: E. Buffetaut, R. Corria, J. Cracraft, P. Currie, G. Dyke, J. Gauthier, J. Homer, C. Magovem, L. Saldago, P. Sweet, M. Vianey-Liaud, and Zhao, Zi-Kui. This research was funded mostly by the Department of Earth Sciences at USC, the Geological Society of America, the AAPS, the Society of Vertebrate Paleontology for work in China, and Luis Chiappe. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What would be the daily life of a PhD candidate without the friendship of his graduate colleagues? So guys, thank you for having been there, especially during hard times! Lastly, I would like to acknowledge Walter Tinner who throughout his life taught me to never give up and always go after what I think is right. Thank you everyone! Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS Page Acknowledgements..................................................................................................... ii List of Tables.................. ix List of Figures...................................................................................................................... x Abstract....................................................................................................... x Chapter I: Introduction: History, Basic Concepts, and Methods....................... 1 The history of fossil eggs........................................................................................ 1 Egg parataxonomy: concepts and caveats............................................................. 2 The nature and formation of reptilian eggs and their phylogenetic inferences... 4 Project description and organization......................................................................6 Material and Methods..........................................................................................................7 Levels of taphonomic and taxonomic confidence............................................... 7 Abbreviations ................................................................................................ 13 Microscopic observations. ....... 14 Macroscopic observations.....................................................................................15 Sample measurements ........................................................ ..15 Phylogenetic analysis............................................................................................ 18 Chapter II: A case study of the preservation of organic membranes of titanosaurid dinosaur eggs from Auca Mahuevo (Argentina): The importance of taphonomy in phylogenetic reconstruction..................................................................... 20 Introduction..................................................... 20 Material ......................... 21 Description... ............... 28 Discussion.......................... 33 Conclusion.......................................... 35 Chapter III: Titanosaurid eggs from Auca Mahuevo (Argentina)-taphonomic and taxonomic levels 1................................................................ 37 Introduction................................................ 37 Material................ ....38 Description... ............... ..38 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion............... 50 Conclusion ............... 58 Chapter IV: Egg, nesting behavior and eggshell structure of the dromaeosaurid Deinonychus antirrhopus-taphonomic level 2 and taxonomic level 1...........................60 Introduction ........................................................... 60 Material.......................................................................... 61 Description.............................................................................................................61 Discussion..............................................................................................................73 Conclusion... ....... ..77 Chapter V: Oviraptorid dinosaurs and the species Citipati osmolka from the Gobi desert-taphonomic and taxonomic levels 1......................................................................78 Introduction............................................................................................................78 Material..................................................................................................................78 Description.............................................................................................................79 Discussion..............................................................................................................85 Conclusion............................................................................................................90 Chapter VI: Re-examination of Troodon formosus, a troodontid dinosaur from Montana-taphonomic and taxonomic levels 1..................................... 91 Introduction................................. ....91 Material.................................................................................................................92 Description ...................................................... 93 Discussion..............................................................................................................99 Conclusion...........................................................................................................104 Chapter VII: An egg clutch of the troodontid Byronosaurus jaffei from the Gobi desert: Novel perspectives on the origin of the avian reproductive physiology- taphonomic level 2 and taxonomic level 1....................................................................105 Introduction... ......... 105 Material .......................... 105 Description....................................... 106 Discussion.................. ......114 Conclusion........... ........... ...117 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter VIII: “Baby Louie”, a putative oviraptorid dinosaur embryo from China- taphonomic level-and comparable from the Gobi desert: Novel perspectives non associated oospecies-taphonomic level 4........................... 118 Introduction............................. 118 Material................................ 119 Description: “Baby Louie” eggs and eggshell structure.................................. 122 F.A.3818.2001-1 egg and eggshell structure...........................127 Discussion ........................................................ 134 Conclusion........................................ 136 Chapter IX: Theropod eggs from Phu Phok (Thailand)-taphonomic level 1, low taxonomic level................................................ 137 Introduction.................................... 137 Material..................................................................... 137 Description...........................................................................................................138 Discussion............................................................................ 139 Conclusion...........................................................................................................144 Chapter X: Possible Enantiomithine eggs from Neuquen (Patagonia, Argentina)- taphonomic level 1, taxonomic level only bracketed to two phylogenetic nodes........................... 123 Introduction.................................... 145 Material................................................................................................................ 145 Description................ 146 Discussion ............................................................................................... 151 Conclusion .......................................................................... ..153 Chapter XI: The fossil bird Lithornis from England and Montana-taphonomic level 3, taxonomic level 1...... 155 Introduction................... 155 Material and Analytical Method. ............................... 156 Description..................................................... 157 Discussion........................................... 162 Conclusion.................. 168 Chapter XII: Paleobiology, oological trends, and Phylogenetic analysis ..........170 Paleobiology and oological trends.................................................. 170 Dinosaur systematic, oological and total evidence phylogenetic analyses. ...185 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conclusion.......................................... ,...............................................................193 References Cited.................................................. 196 Appendix 11.1. Character list. ....... 216 Appendix 11.2. Data matrix................................................... 217 Appendix 12.1. Character list....................................................... 218 Appendix 12.2. Data matrix............................................................................221 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Page Table 1.1. Parataxonomic classification of eggs and eggshells modified from Carpenter, 1999................................................................................................................... 3 Table 1.2. Listed specimens ......... 8 Table 3.1. Titanosaurid egg and eggshell parameters................................... 54 Table 3.2. Comparison of putative eggs and eggshell characters among five South American localities. ................................. 55 Table 4.1. Deinonychus antirrhopus egg and eggshell parameters ............. 54 Table 5.1. Citipati osmolka egg and eggshell parameters............................................. 84 Table 6.1. Troodon formosus egg and eggshell parameters.......................................... 98 Table 7.1. Byronosaurus jaffei egg and eggshell parameters...................................... 108 Table 8.1. Baby Louie and F. A.3818.2001-1 combined egg and eggshell parameters........................................................... 123 Table 9.1. Phu Phok egg and eggshell parameters............................. 140 Table 10.1. Patagonian “enantiomithine” egg and eggshell parameter.................... 148 Table 11.1. Lithomitidae egg and eggshell parameters....................................... 163 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES Page Figure 1.1.............................. .10 Figure 1.2.......................................................................... 17 Figure 2.1....................................................... 23 Figure 2.2............. 25 Figure 2.3.................................. 27 Figure 2.4........................................................................................................................... 31 Figure 3.1........................................................................................................................... 41 Figure 3.2........................................................................................................................... 43 Figure 3.3........................................................................................................................... 47 Figure 3.4.................................................................................................. 49 Figure 4.1............................... 65 Figure 4.2........................................................................................................................... 69 Figure 4.3.................... 72 Figure 5.1..... 82 Figure 5.2................. ...87 Figure 6.1.......... 96 Figure 6.2................................................................ 101 Figure 7.1..... ...110 Figure 7.2.......... ........113 ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.1..........................................................................................................................121 Figure 8.2................ 126 Figure 8.3........ ...130 Figure 8.4...................................................... 133 Figure 9.1.................... 142 Figure 10.1............. .................................150 Figure 11.1.............................. 161 Figure 11.2........................................................................................................................165 Figure 12.1 A............................................ 173 Figure 12.2B.................................................................................................................... 176 Figure 12.2A.................................................................................................................... 186 Figure 12.2B................. 188 Figure 12.3.......................................................................................................................190 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT As a model of archosaurian reproduction the two extant groups, crocodilians and birds, differ in a number of features. Crocodilians produce large clutches of symmetrical eggs that are laid in masse from two functioning ovaries in a simple nest structure in contrast to bird which produce a limited of asymmetrical eggs that are laid from a single ovary in a monoautochronic mode in elaborated nest structures. Crocodilians eggshell is composed of a single structural layer, modem birds display at least three layers. Very little was known about the phylogenetic and paleobiologic values of reproductive and oological features of dinosaurs until the publication of a phylogenetic analysis of paleognath birds followed by a similar but succinct analysis on dinosaurs that supported the theropod-bird origin. The present research based on eight well- identified non-avian and primitive avialans demonstrated that, as the feathers of modem birds, most of the oological characters and reproductive behaviours associated with modem birds are rooted among non-avian theropods. Although the studied taxa in this work represent only still-frames of the entire picture of the saurischian evolution, they clearly reveal reproductive and oological trends common to all these groups; Namely, a reproductive evolutionary cline from crocodilians to modem birds, and a noticeable pattern of coeval development between the acretion of eggshell layers, presence of larger air cells, the reproductive organs, and brooding/incubating behaviours. Most of these pre-adaptations are grouped in two main nodes: One at the xi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. level of Oviraptorosauridae, and the other at Troodontidae. Although that undeniably these two theropods taxa represent important steps in the evolution of avian reproduction, the phylogenetic distance between Oviraptorosauridae and Titanosauria, for instance, cannot be ignored and likely indicates that the reproductive features that appeared in block in oviraptors might have evolved independently across more basal theropods clades. Likewise, although troodontids are in this analysis the obvious precursors to modem avian reproduction, the importance of small-bodied theropods such as those who laid the Phu Phock eggs cannot be dismissed and the eggs of such dinosaurs advocates their closer phylogenetic ties to Aves than troodonts. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I INTRODUCTION: HISTORY, BASIC CONCEPTS, AND METHODS The history of fossil eggs Although the first dinosaur eggs were described as early as 1859 by the French priest and geologist Jean-Jacques Pouech (Buffetaut and Le Loeuff, 1994), traditionally little attention and importance were given to oological data as an alternate or complementary source of characters for phylogenetic inference. For the last one hundred years dinosaur eggs were mostly relinquished to coffee table curiosities, or simply were prized items of individuals assembling huge private collections (as one dinosaur egg discovered at the Flaming Cliff [Mongolia] by Andrews that was sold in 1925 in New York [Gallenkamp, 2001]). More recently, oology (interpretated here as the study of eggs, eggshells, and nests) rebounded not only because of the discovery fossilized oological remains on every continent (except Antartica) but also by the discovery of life assemblages associating egg clutch with a parent dinosaur, or embryo in ovo. At the beginning of this renaissance, eggs of extinct Mesozoic dinosaurs as well as those of their modem descendants were classified into parataxonomic groups (Erben, 1970; Mikhailov, 1987; Hirsch and Quinn, 1990; Vianey-Liaud et al., 1994; Zelenisky et al., 1996; Zao, 1994), a classification initially justified by the lack of fossil eggs associated with parent lineages. However, this parataxonomic classification with its many oospecies offers little in terms of phylogenetic characters (e.g., eggs are classified mostly based on single traits rather than on suites of characters) and cannot be analyzed with modem analytic techniques such as cladistic analysis. Furthermore, little effort was made to try to associate these oospecies with the parent lineages and the few attempts had questionable results (see Chapter 6) as seen in the case of the theropod Troodon formosus (Varricchio et al, 2002). 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Egg parataxonomv: concepts and caveats The existence of the present egg parataxonomic classification is a direct consequence of the paucity of associated skeletal remains with eggs, and the need for erecting a descriptive language that facilitates communication among paleontologists focusing on eggs and their attributes. The development of this classification has gone through many confusing stages during the last decades according to the emphasis that a particular researcher would give to certain characters. Because, the topic of this dissertation is not an historical review of those, I would limit myself to name the principal architects and the classification that is still in use. Challenged by the abundance of dinosaur eggs found in many Chinese Provinces, Zhao (1975, 1979) and Zhao and Ding (1976) devised an organizational system with eggs and their eggshells arranged in oofamilies, oogenera, and oospecies. Subsequently, Hirsch (1989) and Mikhailov (1991) refined this attempt by introducing hierarchical eggshell types (i.e.Testudoid, Crocodiloid, Dinosaurid, Omithoid, and Geckoid) and implemented ten distinct eggshell morphotypes (testutoid, crocodiloid, tubospherulitic, prolatospherulitic, angustispherulitic, filispherulitic, dendrospherulitic, prismatic, geckoid, and ratite). Superimposed on this classification, pore canals were organized as angusticanaliculate, rimocanaliculate, prolatocanaliculate, multicanaliculate, and tubocanaliculate, each of these categories containing many subdivisions. None of their modification of the existing organizational system contributed anything to understanding eggs in a phylogenetic context. The resulting compilation of types, morphotypes, families, genera and species is shown in Table 1.1. Although the intention of these oologists was well founded, their attempts failed to introduce modem analytic techniques such as cladistic analysis in their hierarchical classification (Carpenter, 1999). Furthermore, new oospecies and oofamilies based on either poor samples or limited characters (see Vianey-Liaud et al., [2003] in the 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1.1: Parataxonomic classification of eggs and eggshells modified from Carpenter 1999 Ootypes Oofamilies Oospecies D inosaurid-Spherulitic (eggshell with distinct shell units with a Dendroolithidae 3 radial-tabular ultrastructure) Dictyoolithidae 3 Faveoloolithidae 2 Megaloolithidae 23 ovaloolithidae 6 Spheroolithidae 7 Prismatic (eggshell with two parts. A lower radial- Prismatoolithidae 8 tabular and an upper tabular ultrastructures) Oofamily incertae sedis 1 Ornithoid Elongatoolithidae 15 (eggshell with bird-like shell structure Oblongoolithidae 1 containing a distinct mammillary and continuous layers; no defined shell units; pore are angusticanaliculate. Phaceloolithidae 1 Laevisoolithidae 3 Oofamily incertae sedis 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. case of megaloolithid eggs) were erected and thus created unwarranted oospecies by disregarding for instance the possibility of eggshell morphological variation within a single egg. After the publication of a phylogenetic analysis of paleognath eggs (Grellet-Tinner, 2000), that departed from the traditional parataxonomic egg classification by using cladistic analysis, a few researchers realized that the parataxonomic issues stem from a classification system where categories rest on a single key characteristic (Zelenitsky and Vianey-Liaud, 2003; Zelenitsky, 2003). Nonetheless, these researchers only tried to implement the old system with bits and parts of modem analytic concepts, never departing from the existing classification. The nature and formation of reptilian eggs and their phylogenetic inferrences Prior to the advent of the amniotic egg, metazoan reproduction was limited to the aqueous environment and due to taphonomic biases little of this evolutionary stage was recorded in the fossil record. However, the innovation in the eggshell consistency that evolved in Reptilia not only helped to insulate the growing embryo from outside forces but also played a key role in favoring egg preservation in the fossil record by incorporating large amount of calcium carbonate in the eggshell structure. Reptilian eggshell ranges from a soft leathery shell with inorganic calcium carbonate crystals poorly organized as floating material (Hirsch, 1996) to hard eggshells consisting of interlocking calcium carbonate shell units (Schleich and Kastle, 1988; Hirsch, 1996; Pooley, 1979; Mikhailov, 1992). In extant birds, the other modem reptilian representative, the accretion of a calcium carbonate eggshell to the outer proteineous membranes of amniotic eggs, a biomineralization process, creates an eggshell that is composed of at least three structural eggshell layers (Grellet-Tinner, in press; Grellet- Tinner and Chiappe, 2004). As a model of archosaurian reproduction, the two extant groups, crocodilians and birds, differ in a number of features. Crocodilians produce large clutches of 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. symmetrical eggs that are laid in masse from two functioning ovaries in contrast to birds who produce a limited number of asymmetrical eggs that are laid in a monoautochronic mode by a single ovary. The typical descent of an egg from the left ovary (for most extant birds) to the shell gland (Fig. 1.1 A) follows a complex path comparable in every bird species, though differing in its duration (Parker and Haswell, 1910; Taylor, 1970). Calcitic covering of the ovum and shell pigmentation take place in the shell gland (Makita, 1981) and diagnostic features of eggs are partly controlled by peristaltic movements in the shell gland (Taylor, 1970; Deeming and Ferguson, 1991). The avian reproductive physiology (Romanoff and Romanoff, 1949) is crucial to controlling clutch size, chemical composition (Dauphin, 1990), eggshell morphology, and nest morphology. Therefore, being formed under genetic control, eggs must also retain evolutionary information (Kohring, 1997). This hypothesis was tested by running a cladistic analysis of paleognath eggs (ratites plus tinamous) (Grellet-Tinner, 2000; Grellet-Tinner in press), a clade of birds well constrained in size and which monophyly is widely acknowledged (Lee et al., 1997). The phylogenetic results corroborated the monophyly of Paleognathae previously obtained through molecular and skeletal analyses. However, due to inclusion of fossil taxa in the oological analysis, some internal topologies were incongruent with those from prior analyses based on skeletal or molecular data, but overall the results were congruent (Fig. 1.1B). The discoveries of critical fossils containing embryonic material (Norell et al., 1994, 2001; Homer, 1997; Manning et al., 1997; Mateus et al., 1997; Chiappe et al., 1998, 2001; Swcheitzer et al, 2001) and nest structures (Wenzel, 1992; Clark et al., 1999; Varricchio et al., 1999; Dong and Currie, 1996) played a paramount role by connecting oology with parent lineages. A second but succinct analysis aimed to test the saurischian bird origin hypothesis using a few obvious characters observed in some of these Mesozoic dinosaurs supported this hypothesis and demonstrated that seven 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. additional steps would be required to support the sister-group relationship between Crocodylia and Aves (Grellet-Tinner and Chiappe, 2004) (Fig. 1.1C). Project description and organisation Since the beginning of comparative zoology, the phylogenetic relationship among saurischian dinosaurs rested on osteological characters (Gauthier, 1986) not considering the potential wealth of information hidden in eggs, their eggshells, nests and reproductive and nesting behaviors. One of the questions raised in this research is whether detailed oological observations could bring to the phylogenetic debate a better understanding of the relationship of saurischian dinosaurs and to test whether a particular group of non-avian theropods is closely related to birds. To test this hypothesis, several saurischian taxa were examined on the basis of the degree of eggs and skeletal association. A second question addressed here is whether oological data could, by itself or in complement to established knowledge, bring novel information about the reproductive behaviors of these dinosaurs, possibly their physiology, and perhaps their metabolism. To achieve these objectives, a detailed description of well-identified oological material will be provided in Chapters 3-11. Each chapter is dedicated to a particular species or group of related species and consists, according to the specimen, of a description of the nest, egg, and eggshell structures and morphologies, followed by discussion and conclusion sections. Chapter 2 specially relates to the taphonomic processes exemplified in the particular environment of Auca Mahuevo (Argentina) that lead to the preservation of egg compounds and eggshell structure. The understanding of these processes is paramount to correctly assess the observed oological characters that are the base of this research on the evolution of saurischians. Chapter 3 is dedicated to the titanosaurs from Auca Mahuevo and south America, Chapter 4 to the maniraptors Deinonychus antirrhopus, Chapter 5 to the oviraptor Citipati osmolka, 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 to the troodon Troodon formosus, Chapter 7 to the troodon Byronosaurus jaffei, Chapter 8 to the Macroelongatoolithus xixiaensis oospecies and “Baby Louie”, Chapter 9 to theropod eggs from Phu Phock, Chapter 10 to the possible enantiornithine eggs from Neuquen, Chapter 11 to the neomithines lithornis Vulturinus and L. celetius, and Chapter 12 recapitulates the interpretations made in each chapter and describes the resulting oological trend. Furthermore, it provides a total evidence cladistic analysis including oological characters with already known and published osteological data. Congruence between previous evolutionary hypotheses based on morphological skeletal and these results will be then assessed. Thirdly, this final chapter evaluates the knowledge inferred from oological data and reproductive behaviors in a paleobiological context. MATERIAL AND METHODS Observations of the studied oological material relies on specimens acquired in the field by the author, specimens in various museums, or material kindly donated by institutions and private collectors now curated at the LACM (Table 1.2). The descriptive terminology rests as much as possible on published literature on eggs and eggshells (e.g., Mikhailov, 1991; 1997; Hirsch et al., 1997) in an effort to bridge the parataxonomic literature with this research. However, new terminology is implemented when needed (Grellet-Tinner and Norell, 2002) as explained in the text. In order to observe and describe the eggshell morphology several tools and methods are implemented and complement each other. Levels of taphonomic and taxonomic confidence. Eggs can only be identified taxonomically with certainty either by observing an egg-laying female or by the embryos they contain. In the fossil record, the former would require the unlikely preservation of a female with eggs in its reproductive 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. ITEM MUSEUM eggs & eggshells MCF-PVPH eggs & eggshells MCF-PVPH egg & eggshell MOR 246 eggshell/skull IGM 100/972 eggshell/nest IGM 100/974 egg, & eggshell AM NH3015 eggshell/skull IGM/978/971 eggs, & eggshell IGM 100/979 eggs, & eggshell IGM 100/1127 100/973 eggs, & eggshell L ^M ?7477/149736 eggshell MUCPv-284, -350, eggshells ^ M 16961 eggshell egg clutch Nanyang Museum eggs, & eggshell SKI-1 NAME titanosaurid nests titanosaurid MT Troodon formosus Byronosaurus jajfei Byronosaurus jajfei Deinonychus antirrhopus Citipati osmolka big mama Baby Louie enantiomithine Lithornis celetius Lithornis vulturinus macroelongatoolithid incertae sedis LOCALITY Patagonia, Auca Mahuvo Patagonia, Auca Mahuvo Montana, Egg island and Egg Montain Mongolia, Ukhaa tolgod Mongolia, Ukhaa tolgod Montana. Cashen ranch Mongolia, Ukhaa tolgod Mongolia, Ukhaa tolgod Mongolia, Ukhaa tolgod China, Henan-Xixia (Sanlimiao) Neuquen Montana, Bangtail, Gallatin national Forest, Park County UK, Isle of Sheppey China, Henan-Xixia AGE liu Phock, Nahkon Campanian, Anacleto FM Campanian, Anacleto FM Campanian, Two Medecine FM Campanian-Maastrichtian, Djadokhta FM Campanian-Maastrichtian, KSSW&yFM.V Campanian-Maastrichtian, Djadokhta FM Campanian-Maastrichtian, Djadokhta FM . , . Campanian-Maastrichtian ,Djadokhta FM Maastrichtian, Santonian, Rio Colorado FM Paleocene, Fort Union FM Eocene, London Clay FM Maastrichtian Barremian, Sao Khua FM oo Table 1.2: Listed specimens Figure 1.1 A. Sequential accretion of tissues in the reproductive system of modem birds (modified from Taylor, 1970). B. 14 taxa (eight extant, six fossil taxa) and 15 characters were analyzed with MacClade 3.03 and PAUP 4.0b4a. All the searches were performed with the branch and bound algorithm. Four outgroups, titanosaurids from Patagonia, an Asian troodontid of uncertain specific affinity, and two galliformes were used to determine the polarity of character transformations. Characters are weighed according to the rescaled consistency index resulting in one single tree with 29 steps and a Cl of 0.92 (Grellet-Tinner, submitted). C. Cladistic analysis resulted in 3 most parsimonious cladograms length: 15; Cl: 0.80; RI: 0.90). Strict consensus cladogram (length: 15; Cl: 0.80; RI: 0.90). An additional seven steps would be required to support a sister taxon relationship between crocodiles and birds. Synapomorphies diagnosing nodes: 1[1], calcitic eggshell; 9[1], parental nest attendance; 3[1], presence of eggshell surficial ornamentation; 4[1], two eggshell layers; 5[0], aprismatic delimitation between layers; 6[1], asymmetrical transversal plane symmetry of egg; 10[1], presence of parental brooding behavior; 11[1], monoautochronic ovulation (egg are laid at daily interval); 7[1], open nest morphology; 2[1], blade-shaped shell unit crystallization; 4[2], three or more eggshell layers; 8[1], one functioning ovary; -3[1], absence (reversal ) of eggshell surficial ornamentation; 5 [2], prismatic delimitation between layers (Grellet-Tinner and Chiappe, 2004). 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.1 Crocodylia Cheionia Steps: 15 Cl : .80 Ornithischia Sauropoda Oviraptor Troodon Dromaius Struthio Anser \Rhea Meleagris 5[2] 3[0] 4 P 8[1] 10[1] 1111] 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. system, an event that has not yet been conclusively documented (see Griffiths [2000] and Chen et al. [1998] for a purported occurrence however). Although rare, fossil occurrences of diagnostic embryonic remains in ovo are paramount to identified oological material. Recent discoveries of a variety of fossil eggs containing embryonic material (Currie and Homer, 1988; Homer and Weishampel, 1988; Homer and Currie, 1994; Norell et al., 1994, 2001; Homer, 1997; Manning et al., 1997; Mateus et al., 1997; Chiappe et al., 1998, 2001; Swcheitzer et al, 2001; Grellet-Tinner, 2004, 2005) have provided reliable documentation of the egg morphology of several extinct dinosaur lineages. These discoveries have also made possible the association of eggs of some extinct dinosaurs to specific nest structures (Varricchio et al., 1997; Chiappe and Dingus, 2001; Chiappe et al., 2004) and in some instances, to specific nesting behaviors (Norell et al., 1995; Dong and Currie, 1996; Clark et al., 1999; Varricchio et al., 1999; Chiappe et al, 2000; Grellet-Tinner, 2004, in press). Knowing that preservation of a dinosaur female with eggs in its reproductive system has not yet been conclusively documented, the confidence placed on taxonomic identification of fossil eggs relies mostly on the proximity of skeletal material to a given egg and the researcher’s ability to properly identify this skeletal remain (see Chiappe et al. [2001] in the case of Gobipteryx minuta and Nanantius valifanomi). Hence, there is a need to establish levels of confidence between the association of oological and osteological material and to consider this variable in the phylogenetic results and their interpretations. However, there is a risk that once categories are implemented, fossils would fall in the grey areas between these implemented categories. Furthermore, categories by their nature do not represent a continuum but rigid compartments where items have to be placed in based on limited and arbitrary criteria. Taking in consideration the need of associating identified fossil oological material to their respective parent lineage and the caveats inherent to categories, a two-stage system is here proposed. First, the studied fossils would be given a level 1 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of taphonomic association based solely on their proximity to specific skeletal remains in the field. Then, after extensive observation and perhaps correlation with other described specimens, a level of taxonomic confidence would be assigned to the studied material. According to this classification, taphonomic level 1 is the highest level of taphonomic confidence in the association of fossil eggs with a skeleton. It occurs when diagnostic embryonic remains are found in ovo (Norell et al., 1994, 2001; Chiappe et al., 1998, 2001). This taphonomic level transposes directly to the level 1 of taxonomic confidence provided that the skeletal remains are correctly identified. Should the identification of the skeletal remains be uncertain, controversial (Horner, 1997; Manning et al., 1997; Mateus et al., 1997), rebutted (Chiappe et al, 2001), or being at a low taxonomic hierarchical level not allowing an identification at the species or family level (Swcheitzer et al, 2001), the level of taxonomic confidence might become numerically lower than that of taphonomic association as examplified in chapters 9 and 10 for the Phu Phock and Neuquen eggs. Hence, although a given set of eggs could be irrefutably associated to osteological remains justifying a level 1 of taphonomic association, their taxonomic level might not be equivalent a posteriori because of issues related to the skeletal identification (see Homer and Weishampel [1988] in the case of T . formosus). Taphonomic level 2, still provides a high level of confidence but the association between oological and osteological material is as not as certain as in level 1. More typically it does not involve embryonic remains in ovo, but the egg material is in situ and preserved aposed to or a few millimeters from the skeletal remains. Taphonomic level 3 applies to occurrences where diagnostic skeletons are discovered in the same horizon as eggs that are a few feet away from them (Bonaparte and Vince, 1979). Although proximity would favor the assumption that the eggs could be identified as those belonging to that dinosaur species there is no direct 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. evidence supporting this assumption. Identification resting solely on bone beds or abundance of a particular dinosaur species found in proximity to egg clutches or eggshell fragments has proven dubious at times (see Osborn [1924] for oviraptorid eggs mistakenly assigned to Protoceratops, and Homer and Weishampel [1988] for the misidentification of T . formosus). However, totally rejecting such occurrences as biases could also prove detrimental to the overall progress of oology. Taphonomic level 4 includes any other specimens falling short of the above- mentioned criteria. Although they would be treated as unknown in the taphonomic scheme of level of association, their taxonomic identity could still be recovered by comparison with other identified specimens. The oospecies erected in the existing parataxonomic classification are included in this category. Some of these specimens are associated with nest structures and clutches and have the potential to add valuable characters to phylogenetic analyses. Abbreviations Institutional abbreviations- AMNH-FR, American Museum of Natural History fossil reptiles, New-York; BMNH A, The Natural History Museum (Palaeontology Department), London; GMV, Geological Museum Vertebrate Beijing; IGM, Institute of Geology, Mongolia, Ulaan Baatar; IVPP, Institute of Vertebrate Paleontology and Paleoanthropology, Beijing; LACM, Natural History Museum of Los Angeles County ; MCF-PVPH, Museo Carmen Funes, Plaza Huincul, Argentina; MCZ, Museum of Comparative Zoology, Harvard; MOR, Museum of the Rockies; MUCPV, Museo de Geologia y Paleontologia, Universidad Nacional del Comahue, Neuquen, Argentina; SKI-1, Sahat Sakhan Dinosaur Research Centre; YPM, Yale Peabody Museum of Natural History, New Haven; USC, University of Southern California. Technical abbreviations: CL, cathodoluminescence; DPI, dots per inches; BSEM, backscattered scanning electron microscopy; Fm, formation; MT, Membrana 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Testacea; OUT, operational taxonomic units; PLM, polarized light microscopy; SEM, scanning electron microscope; TLM, transmitted light microscopy. Microscopic observations Scanning electron microscopy (SEM) mostly performed at the AMNH and LACM facilities allows the observation of the samples at variable magnifications. This technique in contrast to TLM or PLM reveals details of the microcrystalline arrangement within shell units in three dimensions (Hirsch and Packard, 1987; Grellet- Tinner and Norell, 2002), the presence and thickness of structural layers, and the distribution of vesicles interpreted as the relics of former protein filaments interwoven in the crystalline matrix of the eggshell (Grellet-Tinner, 2004). However, SEM observations are intimately biased by the quality and the orientation of the fracture of the eggshell specimen. TLM and PLM offer little in terms of three dimensional view but allow better observation of some structures that could be ambiguous in SEM and confirm the presence of various eggshell structural layers, the micro-crystalline arrangement within shell units already observed in SEM views. The “organic lines” that are visible in TLM and analogous to the vesicles and crystallographic cleavages in SEM observations are a perfect example of the way SEM and TLM observations complement each others (Figs. 1.2A, B). As for SEM, thin-section observations are biased by not only the quality of preparation of the thin-sections but also the angle at which they are cut. CL and BSEM analyses were carried out respectively at the USC Department of Earth Sciences and at the LACM facilities to evaluate the alteration and/or replacement of carbonates in fossil eggshells (Amthor, 1993; Korhing and Hirsch, 1996; Barbin, 2000). The main application of CL examinations from thin- sections is to reveal diagenetic textures that are otherwise invisible TLM and SEM, to discriminate those from biological structures. Those fluoresce when bombarded with high-energy electrons, a phenomenon caused by the replacement of calcium or 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. magnesium carbonates (Ca2 + or Mg2 + C 03 ) mostly by manganese in the crystalline structure of the eggshell (Amthor, 1993). However, there is no fluorescence if Ca2 + has been replaced by iron, Fe2 + a fluorescence quencher (it suppresses luminescence). BSEM produces an image of a thin-section where both manganese (Mn2 + ) and iron (Fe2 + ) appear as light zones because of the difference in their atomic number in contrast to that of calcium carbonate (Ca2 + or Mg2 + C 03 ), thus complementing CL observations. To test whether CL alone can discriminate diagenetic from biological structures in eggshell, several eggshells of modem birds were exposed to CL. Regardless on the intensity of the electron beam and the time of exposure no luminescence was noted (Figs. 1.2C-F). Macroscopic observations Whole eggs are rare but not unknown. Taphonomic processes greatly affect egg contour and morphology and their spatial distribution. Whether entire or partially preserved, eggs were photographed with a scale bar from above in a normal direction to avoid any parallax as they were measured. The digital image was imported into a graphic software package and the contour of the specimen was then drawn in order to discern the smallest variation in shape between poles, when entire, or between other eggs that could be of the same oospsecies or family. Variation in the eggshell surficial ornamentation according to its topological position (near the egg equator, poles, or in between) was also reported. When clutches are preserved, number of eggs, their spatial arrangement (i.e. number of egg rows and whether eggs fill up the entire clutch area), and the orientation of their long axis in reference to the ground were noted. Sample measurement Images were directly digitized at high resolution, and saved as Tagged Image Format Files (TIFF). Adobe Photoshop software (version 5.0) is used to enhance 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.2 A. TLM of Deinonychus antirrhopus. Black arrow points to the acicular calcitic crystals of layer 1, white arrows to the organic lines of layer 2. B. SEM of Deinonychus antirrhopus. Black and white arrows point to the same features as those in TLM. Vesicules (in SEM) are aligned in plans following the organic lines (in TLM). C and D. TLM and CL views of the eggshell of an extant emu. No luminescence is noted in this modem eggshell, thus validating the use of CL to distinguish diagenesis in fossil eggshells. E and F. TLM and CL views of an eggshell of an extant Rhea. Same observations as for the extant emu. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.2 100 H rn 15.0 KV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission image contrast and brightness, and for measuring the samples. Egg length and diameter were measured using a Mitutoyo caliper, model # CD-8 CS. When multiple measurements are obtained an average was calculated and recorded. Eggshell thickness was measured from the base of the layer 1 to the outermost inorganic layer. The inner eggshell structural layers were measured between their upper and lower lines of transition with adjacent layers. Phylogenetic analysis Phylogeny is an evolutionary hypothesis that rests upon the quantity and the quality of taxa and characters used to perform a cladistic analysis. There have been alternate views in the systematic literature (DeQueiroz et al., 1995, Sanchez-Villagra and Williams, 1998) regarding the relative merits of phylogenetic analyses based on either separated or combined data sets. Although I favor a total evidence analysis to enhance detection of real phylogenetic groups (DeQueiroz et ah, 1995), I recognize the value of initial data partition to identify similarities and incongruences among the results from these natural subsets of data. Congruent results at high taxonomic levels even with minor internal deviations independently can support a particular evolutionary hypothesis. Although never considered previously by use of numerical cladistic analysis, the phylogenetic approach using oology despite being limited by the available data has great potential to resolve controversial relationships within Saurischia. Aside from characters purely associated with eggs and eggshell strcuture, characters related to nesting behavior, nest structure, egg clutch composition are included on the premise that those of modem birds well document their evolutionary relationships (Winkler and Sheldon, 1993). Two analyses will be thus performed both using heuristic search methods; The first one in the computer program PAUP 4.0b4a (Swofford, 2001) with only reproductive and oological characters for each operational taxonomic unit (OTU) examined in 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. this dissertation to compare and contrast its result with already published cladograms based upon skeletal morphology; A second analysis in the computer program NONA (Goloboff, 1999) under the concept of total evidence (DeQueiroz et al, 1995) incorporating the oological characters for each OTU with the already published skeletal characters in Clark et al., (2002) based on the successive studies of Norell et al. (2001) and Hwang et al. (2004). The character states of the OTUs of which the oological material are not available will be coded with a question mark “?”. One thousand replicates of the three bisection and regraphing algorithm will be implemented in NONA retaining only the ten shortest trees for each replicate. These retained trees will be then subjected to branch swapping extended to cladograms up to 10% longer. A strict consensus search will be performed for the many equally parsimonious cladograms resulted from this analysis. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER II A case study of the preservation of organic membranes of titanosaurid dinosaur eggs from Auca Mahuevo (Argentina): The importance of taphonomy in phylogenetic reconstruction INTRODUCTION Amniotic eggshell consists of an external layer made of laterally juxtaposed calcium carbonate crystals regularly grouped in eggshell units, underlined by a layer of proteinaceous fibers called the membrana testacea (MT) (Fig. 2.1 A). Although the function of the fibrous and proteineous MT is mostly structural, it also acts as a barrier against fungal and bacterial invasion on the egg (Palmer and Guillette, 1995). The MT in extant reptiles is divided into two identifiable layers and is species related (Packard and DeMarco, 1995). Despite some structural misinterpretation inherent to microscope biases, the fine structural composition of the MT is adequately known (Packard and DeMarco, 1995) and enough morphological characters are present to microscopically distinguish it from other organic components found in eggs. Eggshell membranes of modern birds are deposited in the isthmus, the mid-section of the oviduct (Board and Sparks, 1995), and serve as nucleation sites for the growth of the calcium carbonate eggshell units. As a result, protein filaments originating from the upper membrana testacea are embedded at the base of the eggshell once oogenesis is completed (Figs. 2.IB, C). Although fossilization favors the preservation of calcium carbonate in eggs, proteins and other organic compounds that are often destroyed are commonly recovered at Auca Mahuevo in Patagonia, Argentina (Fig. 2. ID), a 79.5-83.5 MA (Chiappe et al., 1999, 2000; Dingus et al., 2000) Campanian locality in the Anacleto Formation. This is the only known site where megaloolithid eggs are preserved with embryonic skin (Chiappe et al., 1998) in addition to skeletal remains in ovo allowing 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a positive titanosaurid identification (Chiappe et al., 2001). Although dinosaur eggs were found in four distinct beds consisting of reddish-brown overbank mudstones (Chiappe et al., 1999, 2000; Dingus et al., 2000), egg-bearing layer 3 has been so far the most productive, and variably preserved titanosaurid eggs are omnipresent. In addition to embryonic skin, the delicate MT is often present, ubiquitously underlying many eggshell fragments, a rare occurrence in the fossil record (Figs. 2.ID, 2.2A, B, E, F, 2.3A-3F, and 2.4D-F). Although a few post-Jurassic instances of MT preservation have been reported in the fossil record (Sochava, 1969; Koslesnikov and Sochava, 1972; Kohring and Sachs, 1997; Kohring, 1999; Grellet-Tinner, 2000), none of these reports matches in quantity or quality the preservation evidenced at Auca Mahuevo. A detailed description of the MT found adhering to the inner eggshell surface of titanosaurid eggs is presented here and compared with the MT observed in eggs of extant reptiles and previously reported fossil examples. Lastly, an interpretation of the taphonomic and diagenetic processes leading to the rare preservation of the MT at this Campanian site is presented. Material Referred samples include MCF-PVPH 250, 264, 441, 442, 444, 445, 446 and specimens for this study were collected during the 1999 joint expedition of the LACM-MCF, from multiple eggs at a quarry site, and from eggs containing embryonic skulls in ovo (MCF-PVPH 147, 250, 262, 263, 264) (Chiappe at al., 2001) all from egg-bearing layer 3 (Chiappe and Dingus, 2001; Chiappe et al 2000). “Stray” eggs and eggshell fragments were compared to those with identifiable skeletal remains to ensure the taxonomic homogeneity of the sampling (since eggshell microstructure is taxonomically differentiated; Grellet-Tinner, 2000). Only the specimens that demonstrate an interesting preservation and diagenetic range were selected for this part of the study. Eggshells used for SEM imaging were treated as described by 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.1 A, cartoon of a modem egg and eggshell including a growing avian embryo. B, SEM of Rynchotus rufescens (AMNH 13376) showing the outermost MT layer in contact with the inorganic calcium carbonate eggshell. C, SEM of Cryptullerus tataupa (AMNH 13374) showing the protein fibers of the MT entering in the inorganic calcium carbonate eggshell. D, thin section of an eggshell fragment from MCF-PVPH 441 displaying on the same specimen preservation of the MT on one side and non-preservation of that membrane on the other. Note the difference between the two distinct layers in the MT. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.1 light albumen m embrana te sta ce a / dense albumen chalaza am niotic sac eggshell y o l k / \a lla n to ic sac n cleation center eggshell unit eggshell unt " .c e £ 'C'’ ce-i.e 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.2 A, eggshells in MCF-PVPH 442 are characteristically stacked with their inner surfaces opposed to the outer surface of another eggshell forming doublets or triplets as shown by arrows 1,2, and 3. B, the original calcium carbonate section of MCF-PVPH 441 adheres to the upper surface of the diagenetically calcified MT. Arrow 1 point to the eggshell, arrow 2 to the space occupied by a freshly pried eggshell exposing the upper surface of the MT, and arrow 3 points to the rest of the exposed MT surface. C and D, respectively MCF-PVPH 250 and 264; these eggshells fragments from eggs containing embryonic titanosaurid skulls as well as others display a tremendous diagenetic alteration on their outer surfaces from highly pitted to pristine. E, SEMs; although mostly preserved as an amorphous calcified layer, as seen in Fig. 2B, the MT of titanosaurid eggs in Auca Mahuevo here in MCF-PVPH 444 is also found with a fibrous consistency reminiscent of the morphology of the original protein material. F, magnification of the preserved MT from specimen MCF-PVPH 444. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.2 _ ------------- SOO^m SOX 500 um 50X ----- ■ ■ » m m m £" : SOOjim SOX 60 um 3 S O X 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.3 A, upper surface of the MT from MCF-PVPH 441. Arrow points to one of the dome-like structures which are casts of eggshell nucleation center also called core by other authors. B, magnification of one of the nucleation centers. Arrow 1 points to the central depression where the tip of the core would fit and arrow 2 to the radiating pattern mirroring the acicular crystallization of each nucleation center. C, inner surface of the original calcium carbonate eggshell of MCF-PVPH 441. Arrow 1 points to the tip of an eggshell nucleation center, and arrow 2 to some calcified MT still adhering to the inner surface of MCF-PVPH 441. D, arrow 1 points to the red discoloration encircling the tip of a core. Arrow 2 points to the blue discoloration surrounding the red one. These concentric color zonings are reminiscent of stratified microbial communities. E, polished cross section of MCF-PVPH 441. Arrow 1 points to the innermost section of the calcified MT, and arrow 2 to the thicker and braided outermost section of the MT, which normally is opposed to the inorganic section of the eggshell. F, voids parallel to the original fibers in the MT of MCF-PVPH 441 viewed with backscattering electron microscopy. 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.3 gg 0.034 mm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Grellet-Tinner (2000) and Makovicky and Grellet-Tinner (2000). Furthermore, the radial section of specimen MCF-PVPH 441 was polished with 0.2 micron silicon carbonide powder for BSEM examination. DESCRIPTION Eggshell fragments were found as such or forming doublets or multilayered aggregates (Fig. 2.2A). Within the eggshell sample, the specimens from egg- bearing layer 3 display tremendous diagenetic variation. This ranges from pristine crystallographic preservation to extremely altered eggshell microstructure (Fig. 2.2C, D) that includes pitting of outer eggshell surfaces and recrystallization of calcite in the pore canals and apertures. This variation is visible in isolated eggshells, eggshell from complete eggs as well as eggs with embryo in ovo and is likely the source of calcium carbonate for a thin white layer that is frequently found underlying the inner surface of these eggshells (Fig. 2.2B). This underlying layer occurs in two states. It commonly appears as a compact layer of white calcite in situ having dome-like structures on its upper surface, and more rarely as preserved calcified strands closely resembling what could have been the original textural organization of the MT. The former is readily observed under a binocular microscope (Figs. 2.3A-D), the latter in SEM (Figs. 2.2E, F) and in thin sections (Figs. 2.4D-F). Koslesnikov and Sochava (1972) reported similar textures and interpreted them as the MT of “multicanalicular” dinosaur eggs of unknown lineage from the Cenomanian site of Ologoy-Ulan-Tsab (Gobi Desert), and Kohring (1999) describe a mesh-like MT in megaloolithid eggs from the Maastrichtian Tremp Basin of Spain. Gross morphological observations and the placement in situ of the two morphs coupled with comparison with the previously described dinosaur MTs (Sochava, 1969; Koslesnikov and Sochava, 1972; Kohring, 1999) confirm the MT identity of these textures preserved in the Auca Mahuevo eggs. SEM observation of the MT layer confirms its identity by the disorganized 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. filament-like structures (Figs. 2.2E, F), a morphology also clearly evidenced in thin section (Figs. 2.4D-F). This layer averages 281 pm in thickness, a value only slightly larger than those reported for a similar structure in megaloolithid eggs from the Tremp Basin (Kohring, 1999). Two distinct zones are visible within this layer. The one proximal to the eggshell is more compact than its distal counterpart. As visible in TLM (Figs. 2.4D-F), the entire thickness of the MT is composed of parallel strands that range from 13.5 to 18.7 pm in diameter. However, as yet, it is unclear whether each of these strands represents an original individual protein fiber or a bundle of these fibers. The juxtaposition of this layer to the eggshell inner side coupled with the morphological change from a compact to a somewhat more aerated textural composition is congruent with the appearance of the MT in modem non-avian reptiles and birds (Packard and DeMarco, 1995). The white compact layer of calcite is characterized by dome-like structures on its upper surface (Fig. 2.3A). Although occasionally standing as single structures, they mostly form groups of two or more (Fig. 2.3A). The outside diameter of a dome averages 0.034 mm and all possess apical pits. The base of each dome does not extend lower than the domes themselves (Fig. 2.3B). Furthermore, these pits display fine negative imprints of acicular crystals radiating from their centers. These imprints persist onto the rim of the pits but are not present on the outer surface of the domes. A slight red coloration of the calcite is visible around the perimeter of the domes. Although different in color (green-pistachio), this feature was noted by Koslesnikov and Sochava (1972) in unidentified dinosaur eggs from the Gobi Desert. A small fragment of eggshell was pried off the white MT in specimen MCF- PVPH 441 (Fig. 2.2B) to test the hypothesis that these domes could be the negative imprints of the inner apex of eggshell units. The inner surface of this eggshell shows a thin layer of white calcite similar to that of the MT, which is randomly pierced by the black base of the eggshell units (Fig. 2.3C). The inner apexes of these eggshell 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.4 A, rhombohedric calcium carbonate crystals infilling voids outside the MT of MCF- PVPH 441. B, rod-like calcium carbonate crystals infilling voids within the MT of MCF-PVPH 441. Note the striking difference of crystallization according to the position of the voids. Within the MT the crystallization is speculated to be induced or controlled by bacteria, while outside this region the crystallization is abiotic in origin. C, higher magnification of the rod-like crystals within the MT of MCF-PVPH 441. Note the discoloration at the tip of each crystal as previously noted by Folk and Lynch (1994) in ooids suggesting according to these authors a bacterial precursor to crystallization. D, thin section of MCF-PVPH 445 clearly showing the preserved filamentous MT. E, magnification of figure 4 D showing in greater detail strands of MT (arrow 1) and eggshell unit nucleation centers (arrow 2). F, greater magnification of the MT strands in MCF-PVPH 445. 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. units are grouped in the same pattern as the domes of the MT. Fine acicular calcite crystals radiate from the apex of each eggshell unit (these radial centers are also named “organic cores” or “nucleation centers”). Around the domes on the calcite layer, there is also a red coloration encircling each eggshell unit. However, the red tinge is itself surrounded by a thin blue concentric layer (Fig. 2.3D). The presence of the two juxtaposed discoloration areas on either sample is reminiscent of stratified, multilayered microbial communities observed in modem microbial environments (Walter et al., 1992; Stal and Caumette, 1994; Stal, 2000). The black apexes, by their geometry and acicular striations, perfectly fit inside the pits of the white domes located on the surface of the MT. Furthermore, the spatial grouping of the black striated nucleation centers evidence that they would have been opposed to and in contact with the white domes of the calcified MT. The spatial arrangement of apexes and domes in these fossilized eggs is analogous to the position of eggshell unit nucleation sites in the MT of extant reptiles (Board and Sparks, 1995), hence further supporting the MT identity of the fossilized white layer in these titanosaurid eggs. The morphology of the two fossilized morphs coupled with their placement on the inner eggshell surface, their similarity to both previously described fossilized dinosaur and extant reptiles MTs evidence that the white calcite and the mesh-like layers are the MT of these titanosaurid eggs. A small section from MCF-PVPH 441 consisting of the MT layer and underlying material was broken off and its radial section was finely polished (Fig. 2.3E) to possibly discover the process behind such an exquisite preservation. Scanning at 1200 dpi resolution with a flat bed scanner reveals that the MT is divided into two distinct layers, whereas the innermost section is the thinnest and the uppermost section is crisscrossed by braided strands (Fig. 2.3E). BSEM reveals these braids to consist of long calcific strands 2 to 3 microns wide laterally flanked by voids (Fig. 2.3F). Examination of the voids throughout the entire section shows an abrupt change of crystallization from the outer to the inner surfaces 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the MT. Within the MT, the voids are packed with rod-like or baton-like structures with a slight discoloration at their distal tips (Figs. 2.4B, C) a carbonate structure previously mentioned by Folk and Lynch (2001) in their description of ooid formation. Flowever, at the border and outside of the MT, voids are filled with rhombohedric calcite crystals more characteristic of abiotic precipitation (Fig. 2.4A). DISCUSSION Previous geological descriptions of the Auca Mahuevo site (Chiappe et al., 2000; Dingus et al., 2000; Chiappe and Dingus, 2001) attributed the excellent preservation of its eggs and embryos to the deposition of very fine sediments from the flooding of paleorivers. The silt and mud would have gently covered the riparian nesting grounds and smothered the developing embryos in ovo. The lack of energy and lack of exposure to air are paramount for the preservation of the MT in ovo and in situ. This interpretation is compatible with the fossilization of the MT which would have otherwise desiccated and become detached from the eggshell if exposed to air for a week or more (Hayward et al., 2000). Although the weathering of the eggshell outer surface could be driven by abiotic processes such as chemical dissolution the eggshell surface (Hayward et al., 1991), the preservation of micron size protein strands is rare in the fossil record and therefore could be the result of exceptional taphonomic processes resulting from microbial mediation (e.g. Bottjer et al., 2002). Kohring (1999) and Kohring and Sachs (1997) already speculated that previously known MT preservations were bacterially mediated during an early stage of aerobic decay. However, an early stage of aerobic decay is not a prerequisite for bacteria to induce or control the calcium precipitation onto the MT protein fibers (Reid et al., 2000; Visscher et al., 2000). Eggs by their nature are excellent reservoirs of sulfur, phosphate and other organic and inorganic compounds necessary for the developing embryo (White, 1995; Noble, 1995; Palmer and Guillette, 1995). In addition, the 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. chemical composition of most reptile eggshell consists of calcium carbonate and calcium phosphate), with minute and specifically related amount of trace elements (Board and Sparks, 1995). Although these inorganic compounds amount for 98% of the shell composition, it is well recognized that the remaining 2 % are proteins (Board and Sparks, 1995). Under anaerobic conditions, sulfate or sulfur-reducing bacteria could transform these compounds into energy source by anaerobic respiration of sulfate or sulfur (Nealson, 1997). A by-product of bacterial sulfate reduction is increased alkalinity that could foster the precipitation of calcium carbonate. Such microbial mediation has already been proposed by Reid et al. (2000) and Visscher et al. (2000) for the lithification of stromatolites. These originally soft microbial mats are exceptionally preserved because autotroph and heterotroph bacteria precipitate calcium carbonate ions onto micron-size exopolymer fibers during sedimentation intervals. Folk (1994) demonstrated empirically that bacteria are more likely to precipitate calcium carbonates in hypersaturated solutions onto organic matter than on copper or ceramic. Also, organic matter has been noted to form a template for early silicification versus abiotic precipitation (e.g., Maliva et al., 1989). In our sample, the rod-like crystals within the MT voids are morphologically similar to those seen in ooids by Folk and Lynch (2001). These crystals could plausibly originate from bacteria feeding upon organic matter thus creating microenvironments oversaturated in calcium carbonate or these same negatively charged bacteria offer a “potency spot” attracting cations (Folk and Lynch, 2001). EDS of these MT crystals show a slight amount of silica (S i2 + ) which is not observed in the rhombohedric crystals outside the MT, thus possibly supporting bacterial mediation for the fossilization of the MT in eggs from Auca Mahuevo. Thus, the rod-like carbonate structures filling the voids within the MT suggest a bacterial mediation (Folk and Lynch, 2001). This strikingly 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. differs from abiotic rhombohedric calcite crystals found at the border and outside of the MT, characterized by the absence of organic matter. Although the origin of the calcium dissolution from the outer eggshell surface could simply come from abiotic dissolution of the eggshell, it is important to consider that lithotroph cyanobacteria are known to drill through crystals and jump from crystal to crystal to utilize inorganic compounds (Nienow and Friedmann, 1993; Reid et al., 2000). The result is that rock surfaces display at a microscopic level a pitted appearance and mineral crystals are welded together (Reid et al., 2000). This process could explain the observed surficial pitting by bacterial involvement. The pore apertures and canals of eggs are natural channels for bacterial invasion as reported by Hayward et al. (1997) in the case of black bacilli biofilms centered in pores of Gallus gallus eggshell after eggs were submerged and buried in marine sediments. The location of the concentric colorations around the domes and eggshell units in the calcified MT morph correspond with the inner extremities of the pore channels of the titanosaurid eggs. Therefore, reenforcing the interpretation of a microbial driven process for the MT preservation. CONCLUSION A probable interpretation of the preservation of the MT of the Auca Mahuevo eggs is that it results from biomediation by microbial activity under anoxic conditions at the time of burial and decomposition of the inorganic and organic compounds within the eggs. This microbial driven biochemical process could also explain the recovery in ovo of permineralized embryonic skin patches (Chiappe et al, 1998). Bacterial mediation was previously suggested as an agent of fossilization for the mesh-like MT of the megaloolithid eggshells from the Maastrichtian red beds of the Tremp Basin in northeastern Spain (Kohring, 1999). However, this author invoked an aerobic phase, a phase that we now know is not essential because bacteria can adapt 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to extreme environments transforming the organic compounds into energy source by anaerobic respiration of sulfate or sulfur (Nienow and Friedmann, 1993; Nealson, 1997). Kolesnikov and Sochava (1972) described the MT of the multicanalicular Albian-Cenomanian eggshells from Asia with the same morphology as the compact calcified MT from Auca Mahuevo, but no rod-like crystals were reported at the time. Exceptional conservation in Lagerstatten is often attributed to stagnation, obrution, and microbial mediation (Bottjer et al, 2002). The episodic flooding and sedimentary events in Auca Mahuevo could offer a Lagerstatten setting for some of the egg clutches that are assumed to be the product of seasonal reproductive events. Under these conditions, the eggs were not crushed at the time of burial and the presence of organic compounds within, could have provided the necessary organic compounds to foster a microbial mediation in the preservation process. However, it is still unclear if these taphonomic conditions were specifically restricted to the environments of all or some clutches, and all or some eggs within a clutch. Likewise, the presence of two MT morphs is still unresolved and variation in the taphonomic processes leading to the two preservation states needs to be further investigated. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER III Titanosaurid eggs from Auca Mahuevo (Agentina) (Taphonomic and taxonomic levels 1) INTRODUCTION Dinosaur localities preserving embryonic remains in ovo are extremely rare. An exception is Auca Mahuevo, in Argentina (Neuquen Province, northwestern Patagonia), a site contained within the Late Cretaceous Anacleto Formation (early Campanian; see Dingus et al. 2000), where embryonic skeletal remains and skin casts are abundant (Chiappe et al. 1998 1999 2000 2001a). The initial study of the Auca Mahuevo embryos in ovo (Chiappe et al. 1998) documented several cranial characters diagnostic of Neosauropoda, a cosmopolitan group of sauropod dinosaurs comprising the common ancestor of Diplodocus and Saltasaurus plus all its descendants (Wilson and Sereno 1998). This study provided only a brief description of the eggshell structure, identifying it as Megaloolithidae, a parataxonomic group (Sahni et al. 1994; Vianey-Liaud et al. 1994; Bravo et al. 2000; Magalhaes Ribeiro 2000; Mohabey 2000). Recently, Chiappe et al. (2001b) reported on several new eggs containing nearly complete embryonic skulls. The cranial morphology of these embryos allowed fine-tuning of the initial identification of these eggs as those of titanosaurids, a subgroup of neosauropod dinosaurs. Despite the fact that megaloolithid eggs have been often identified as those of sauropods (Powell 1992; Faccio 1994; Sahni et al. 1994; Vianey-Liaud et al. 1994; Calvo et al. 1997; Bravo et al. 2000; Cousin and Breton 2000; Magalhaes Ribeiro 2000; Mohabey 2000; Peitz 2000), the Auca Mahuevo eggs are the only ones identified on the basis of skeletal remains in ovo. In this chapter I provide a detail description of the shell structure and variation of the titanosaurid eggs from Auca Mahuevo based on five eggs that contain 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. diagnostic material (Chiappe et al. 2001b) and several subsequent phylogenetic and paleobiological interpretations. In addition, the morphology of these eggs and eggshell is compared to that of other South American Late Cretaceous eggs alleged to be also of titanosaurid neosauropods. Material The studied eggs (MCF-PVPH 147, 250, 262, 263, 264) and eggshells (MCF-PVPH 442, 445, 444) were recovered from a quarry contained within Auca Mahueo’s egg- bed 3, one of the four egg-bed layers recognized within the stratigraphic section of this locality (Chiappe et al. 2000). These eggs were part of what were considered to be clutches during the excavation process, and efforts were made to unearth and preserve the entirety of these assemblages. The selected eggs (MCF-PVPH 147, 250, 262, 263, 264) were further prepared and their cranial skeletal content was recently described (Chiappe et al. 2001b). The studied eggshell fragments (one or two samples/ specimen) were removed during the preparation of these embryos. DESCRIPTION MCF-PVPH 147, 250, 262, 263, 264 displayed the same spherical to subspherical general shape as other preserved eggs from Auca Mahuevo egg-bed 3 (Fig. 3.1 A). They show signs of fractures and various degrees of compression (Fig. 3. IB), interpreted as the result of sedimentary compaction. It is worth mentioning that mini slickenslides are visible at the quarry site vertically displacing some eggs. Eggs located at, or close to a vertical displacement horizon are distorted in the same direction as that of the displacement, thus demonstrating a certain amount of plasticity. Two diameters of the eggs were measured. The greatest diameter range between 125 and 140 mm, the smallest diameter is 10 to 20% smaller. However, these two diameters might be slightly underestimated due to the degree of compaction. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Eggshell samples were taken from stacks of eggshell fragments (arrangement resulting from taphonomic processes) randomly assorted in multiple layers surrounding the embryonic skulls (Fig. 3.1C). These samples display tremendous diagenetic variation, manifesting itself by the calcium carbonate dissolution from the internal structure of the eggshell, the pore canals and apertures, as well as the outer eggshell surface. Calcium carbonate after being mobilized was re-deposited within the pore canals in the form of two very distinct crystal shapes but both belonging to the rhombohedric form (Figs. 3.2A, B). The outer eggshell surface displays nodes visible to the naked eye, with some coalescing into a longer structure (Fig. 3.ID). However, MCF-PVPH 250 nodular surficial ornamentation is totally flattened leaving only a rough and pitted appearance (Fig. 3.2C). MCF-PVPH 262 displays a similar chemical weathering to that of MCF-PVPH 250, but its eggshell surface still displays an undulatory appearance, representing remnants of preexisting nodes (Fig. 3.2D). MCF-PVPH 147 suffers from a similar alteration, although to a lesser degree. In this specimen, only the apex of the nodes are affected, displaying a concave appearance (Fig. 3.2E). The dissolution of MCF-PVPH 263 is limited to a moderate pitting of the surficial nodes (Fig. 3.2F). MCF-PVPH 264 is the best preserved of all the samples and shows no surficial alteration (Fig. 3.3A). In MCF-PVPH 264 and 263, the average diameter of the nodes equals 0.58 mm and the average nodular height measured from the base to the apex is 0.28 mm. Due to biases of observation, nodes that coalesce were not observed in SEM images, but only in light microscopy (Fig. 3. ID). It seems that the width of these tubercles, however, does not exceed that of a single node. The few intemodular measurements performed on MCF-PVPH 263 and 264 range from 0.52 to 0.87 mm. The eggshell thickness of the studied specimens varies according to the degree of diagenesis they have been submitted to. MCF-PVPH 147 thickness averages only 0.65 mm; MCF-PVPH 250, 0.73mm; MCF-PVPH 262, 0.82 mm; MCF-PVPH 263 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.1 A. Egg clutch recovered from the quarry site in egg-bed 3 (LACM 149648). B. Detail of Fig. 1. Note the sub-spheric aspect of these eggs (LACM 149648). C. Embryonic remains in ovo (MCF-PVPH 147). D. Eggshell surface with single and coalescent nodes (MCF-PVPH 442). Scale the side Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.1 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.2 A. Diagenetic calcite infilling in and at the base of pore canals of MCF-PVPH 262. B. Diagenetic rhombohedric calcite crystals infilling in and at the base of pore canals of MCF-PVPH 147. C. The eggshell outer surface of MCF-PVPH 250 displays an extreme weathering (scaling) from purely chemical or, as suggested in another publication, of biochemical origin. D. MCF-PVPH 262, same as MCF-PVPH 250 but not as altered. Pore orifices can still be observed (arrow). E. Eggshell outer surface of MCF-PVPH 147 with moderate weathering. The nodular ornamentation although pitted (arrow) is clearly visible. F. Pristine surficial ornamentation of MCF-PVPH 263. However, note the infilling of the pore channel by secondary calcite deposits (arrow). 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.2 ■ M H H I H n H H I h a W t m m m m 500(,im 5OX ■ a i f f i g l g . OHK S > V > * * 50 ^m 500X j g * | 20|ira 1000X M M 200jxm 75X _ . _ . _ _ _ _ _ r _ _ hMMNb ^ V * - . -*r -■.." '" .i* . 1 * *. 500)im 5 OX 5 0 0 n m 5 OX 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and 264 have the same value of 1.31 mm. In radial section, the eggshell is composed of a single structural layer consisting of acicular calcitic crystals radiating from nucleation centers (former organic cores) located above the MT (Fig. 3.3C). The core of MCF-PVPH 263 measures 0.034 mm at its widest diameter (Fig. 3D), and as observed in MCF-PVPH 444, is located 0.19 mm above the MT (Fig. 3.3B). The calcite crystals in specimens MCF-PVPH 263 and 264 are particularly devoid of pitting caused by acidification during the fossilization process. Close observation of these specimens at high magnification reveals that the crystals radiating from the core are of rhombohedric form (Grellet-Tinner 2000b). The diameter of the rhombohedrons proximal to the core averages 0.004 mm (Fig. 3.3E) and each crystal is subdivided in four long coalescent smaller crystals (Fig. 3.3F). The regular grouping of all the acicular crystals radiating out of any given core and their vertical extension define the eggshell unit and the thickness of the eggshell. Unless they coalesce one another, each of these eggshell units forms a round node at the outer eggshell surface. Eggshell units are best observable in MCF-PVPH 263 and 264 (Fig. 3.2F and Figs. 3.3 A, B) and their geometry is best seen when radial sections cut through their center. Beginning at the inner surface, the units flare out at approximately 40 degrees from both sides of the vertical axis of the core. As they reach a third of the total eggshell thickness (Fig. 3.2F and Figs. 3.3A, B), their lateral margins still flare out but to a lesser degree, thus appearing nearly parallel in some instances. The maximum width of any given eggshell unit can only be obtained if the cut of a thin section or the break of an SEM sample is centered at the sagital plan. The measured width of MCF- PVPH 264 (Fig. 3.3 A) is close to 0.63 mm. This is congruent with the measurement of one of the eggshell units in a thin section of a sample of MCF-PVPH 444 (Fig. 3.3B). The apparent change of obliquicity in the geometry of the eggshell units results from the presence of a network of horizontal pore canals, connected to the vertical 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pores, that develops at the base of the eggshell units. This network is particularly evident (Fig. 3.4A) when the MT is preserved in situ. It is the development of such a network that gives each eggshell unit the appearance of a massive pillar separated by horizontal canals from that the adjacent units (Fig. 3.4B). Thin sections reveal that the eggshell units are transversally crossed by thin and compact growth lines (Fig. 3.3B). These lines are slightly convex at the base of the eggshell unit following the primary radiation of the acicular crystals from the core. They progressively flatten higher up in the eggshell units and become horizontal. Once at the level of the apex of the cavity created by the network of horizontal canals, the growth lines more frequently cut through adjacent eggshell units, thus becoming continuous across them. This condition becomes the norm closer to the outer surface of the eggshell. The overall pore-canal network consists of vertical and horizontal canals intersecting on a one to one basis at and between the base of each eggshell unit (Fig. 3.4A). Mostly, the vertical section of these pores is straight, although they sometime diverge forming a Y pattern (Fig. 3.4 A). The diameters of the vertical canals are not precisely measurable in these specimens due to the fracture orientation of the eggshells. Nevertheless, it is possible to determine that their minimum width is 0.08 mm -as previously observed by Chiappe et al. (1998) -in MCF-PVPH 262 and that they do not exceed the maximum diameter of 0.2 mm. As they reach the eggshell surface, pore canals are funnel-shaped (Figs. 3.4C, D), creating round depressions between the surficial nodes. In the studied specimens, the diameters of pore apertures vary from 0.15 to 0.29 mm. The innertips of the eggshell units are particularly visible in MCF-PVPH 263 (Fig. 3.4E). Contrary to expectations they do not display a typical concave appearance resulting from the calcium absorption by a developing embryo, but rather exhibit a perfect conic shape. In extant birds, calcium carbonate dissolution takes place on the 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.3 A. Pristine surficial ornamentation of MCF-PVPH 264 and no secondary calcite deposits in pore canal. B. Thin section of MCF-PVPH 445. The membrana testacea (arrows) forms a fibrous mat underlaying the eggshell units. C. Calcite crystals radiate from a nucleation center (arrow), also called organic core, at the base of each egghsell unit (MCF-PVPH 263). D. Detail view of a nucleation center (MCF-PVPH 263). E. Rhombohedric acicular calcite crystals in eggshell units (MCF-PVPH 263). F. The fabric of each shell unit only consists of fine acicular radiating acicular calcite 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.3 SOOum SOX / > T ~ \ rm ■ -------------------- Q H H M N B 3 " ^ ■ H M ! . 1 ™ 50pm 500X 60 um 350X H U H JB. lOum 2000X _ _ - f l M r i H I 60|Um 1000X 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.4 A. Total view of the radial section of a titanosaurid eggshell. Note the network of vertical and horizontal pore channels. Each eggshell unit forms a massive pillar (arrows) at its base delineating horizontal pore canals. Node the node on the outer eggshell surface surrounded by 3 pore apertures and the 2 branches (arrows) of a pore canal, one ending at the aperture (MCF-PVPH 444). B. View of the junction of the base of three eggshell units with a poorly preserved MT (arrow) showing the horizontal network of pore canals infilled with secondary calcite (MCF-PVPH 444). C. Detail of two pore apertures at the base of a node on an outer-eggshell surface (MCF-PVPH 444). D. Detailed view of the circular, cup-like shape of a pore aperture (MCF-PVPH 444). E. View of the inner eggshell surface. Note the base of many eggshell units (MCF- 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.4 200 m joox 200m m75X lOOfim 150X lOO^trn 200X1 1 HI SOOum 50X Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. innertips of the shell units (mammillary cones) during embryogenesis by a process where the carbonic acid results from the C 02 exhaled by the breathing embryo. The lack of calcium resorption coupled with the presence of a well-developed skull in MCF-PVPH- 263 suggests that calcium utilization from the eggshell by the titanosaurid embryo must have happened at a later ontogenic stage than that observed in modern birds. DISCUSSION As noted in the above description, the recovered eggs are mostly subspherical, a shape that could be either characteristic of this oospecies or the result of post-burial compression forces. Experimentation on recent eggs, however, suggests that the latter is probably the case. Hayward et al. (2000) documented that when chicken {Gallus gallus) eggs are buried in sand under controlled environments and submitted to a compressive force of 700 kg, fresh and hollow (blown) eggs were compressed to 85 % and 82.1 %, respectively, of their original shape. It is possible that egg resistance to high compressive pressures be correlated to the number of structural layers in a given eggshell. Studies on mollusk shells have documented an increased resistance to pressure as the shell accumulates more structural layers, particularly when the internal crystallization varies in direction from layer to layer (Kamat et al. 2000). The presence of structural multilayered shells in birds ( Grellet-Tinner 2001; Grellet-Tinner and Chiappe 2004; Mikhailov, K. E. 1991), suggests avian eggshells may be more resistant to pressure than the single structural layer of titanosaurid and other non- theropod dinosaur eggs. If seemingly more resistant eggshells such as those of Gallus gallus were significantly compressed by sedimentary compaction, it is likely that the difference between the two diameters observed in each of the studied titanosaurid eggs, with only one eggshell structural layer, also results from compressive burial forces. 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As described in chapter 2, the diagenetic variation affecting these Auca Mahuevo eggs greatly varies among the examined eggshell fragments. The surface pitting of MCF-PVPH 147, 262, 263 can be attributed to acid erosion, a process reported and described particularly by Hirsch and Lopez-Jurado (1987) and Hayward et al. (1991). As previously observed with other archosaur eggshell (Kohring 1999), in Auca Mahuevo the mobilized calcium carbonate from the outermost part of the eggshells was precipitated into the pore system (Figs. 3.3E, F) and on the MT (Grellet- Tinner in press). These weathering and precipitation processes in Auca Mahuevo seem to result from biotic mediation, an interpretation sustained by experimental investigations (Hayward et al. 1997) and previous examinations (Grellet-Tinner in press). Pore canals assure the diffusion of gases and water vapor through the eggshell (Paganelli 1980). Their size, geometry, and number reflect a specialization to the habitat where nesting occurs (Williams et al. 1984; Cousin and Breton 2000). Williams et al. (1984) described twelve megaloothid eggs that they labeled as “Type 31” from the Late Maastrichtian of France with a pore system matching in size, shape, and geometry those of Auca Mahuevo eggs. According to these authors, these French eggs would have a vapor conductance equal to twenty four times the calculated values for avian eggs. Independent studies (Packard et al. 1979) have also shown that hard- shelled eggs of extant non-avian reptiles deposited in high moisture environments have vapor conductances from 2 to 5.5 times higher than those of avian eggs. In light of this, Williams et al.’s (1984) conclusion that the French Late Cretaceous megaloolithid eggs were incubated in high moisture conditions seems reasonable and such a conclusion could be extended to the titanosaurid eggs from Auca Mahuevo, on the basis of similarities of the pore system. This interpretation would thus suggest the presence of high moisture content in the Auca Mahuevo titanosaurid nests. Other characteristics of eggshells that have been used to infer the 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. paleoenvironmental context of dinosaur eggs include the surficial ornamentation. Cousin and Breton (2000) used the characteristics of the nodular appearance of eggshell from the Late Cretaceous of France as an indicator of the substrate, or nesting material surrounding the eggs. Thus, the nodular surface of the titanosaurid eggs from Auca Mahuevo could be interpreted as a specialization to facilitate gas and water vapor conductance through the pores that are located around and at the base of each node, by avoiding nesting debris plugging their apertures (e.g., Sabath, 1991). Following this assumption, it is reasonable to hypothesize that in order to achieve maximum efficiency in facilitating gas and water vapor conductance, the inter-nodular distance should be less than the size of the nesting material, otherwise the pores would be plugged. Multiple measurements taken from eggshell fragments not suffering from diagenetic or taphonomic alterations show that the inter-nodular distance of the Auca Mahuevo eggs varies from 0.6 to 1.6 mm. Although this inter-nodular variation is substantial, each eggshell fragment has itself a nearly constant inter-nodular value. Notwithstanding these caveats, these inter-nodular values suggest that the nesting material should optimally exceed 1.6 mm in size. However, the granulometry of the silicoclastic sediments at the Auca Mahuevo site is smaller than the minimal required intemodal distance for the pores to be functional, and therefore it is feasible to speculate that the Auca Mahuevo nests could have been lined and covered by vegetation. This interpretation seems to be supported by the recent discovery of carbon remains, presumably of plants, inside some of the nesting depression excavated by the adult titanosaurids from Auca Mahuevo titanosaurids (Chiappe et al. 2004). Furthermore, nests covered by vegetation, have already been suggested for other Late Cretaceous megaloolithid French sites such as those of Rennes-Le-Chateau (Cousin, 1997; Cousin and Breton, 2000; Erben et al., 1970), and at l’Arc (Kerourio, 1981). Numerous South American isolated eggshell fragments and complete eggs with various degrees of diagenetic alterations have been referred to the megaloolithid 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. parataxonomic group. Among those, some have been labeled as titanosaurid dinosaur eggs (e.g., Erben 1970; Powell 1992; Sahni et al. 1994; Faccio 1994; Moratalla and Powell 1994; Vianey-Liaud et al. 1994, 1997; Calvo et al. 1997), although the lack of diagnostic embryonic remains in ovo made each of these taxonomic identifications unreliable. The Auca Mahuevo eggs provide the opportunity to compare the morphology of unquestionable titanosaurid eggs with that of eggs alleged to belong to this dinosaur group. The following oological charaters; sub-spherical eggs with a greater diameter of 12.5 to 14 cm, eggshell ornamentation consisting of mostly of single but also coalescent hemispheric nodes of 0.58 by 0.28 mm, funnel-shaped pore aperture in between each nodes that are connected to straight or Y-shaped vertical pore canals transecting the entire eggshell thickness; each of these vertical canals coincides with a horizontal counterpart of equivalent diameter; the eggshell thickness, although varying according to the degree of diagenesis, averages 1.31 mm for well-preserved eggs and is composed of one single structural layer (Table 3.1). A brief summary of comparisons with other South American oospecimens putatively assigned to titanosaurid is given below. Calvo et al. (1997) reported a partial egg (MUCPv-245) and three crushed eggs (MUCPv-246) from rocks of the Late Cretaceous (early Campanian; see Dingus et al. 2000) Anacleto Formation in the vicinity of Neuquen, the same stratigraphic unit that contains Auca Mahuevo although located some 200 km towards the southeast. In most respects, the Auca Mahuevo eggs are remarkably identical to those from this locality described as Megaloolithuspatagonicus by Calvo et al. (1997) (Table 3.2). This similarity includes the surficial ornamentation, pore diameter, pattern of growth lines within the eggshell unit, and the apparent presence of a horizontal canal network. The only noticeable differences involve the diameter of the eggs, which is slightly larger in the Neuquen eggs (160 mm), and a thicker eggshell 1.7-2.1 mm according to Calvo et al. (1997). Differences in diameter, however, could be easily explained by a greater 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.1: Titanosaurid egg and eggshell parameters egg shape egg greater diameter egg smaller diameter egg diameter ratio ornamentation morphology ornamentation size pore canal morphology pore canal size pore aperture morphology intern odular space pore aperture size eggshell thickness number of layers layer 1 morphology layer 1 size ratio Ll/total thicknes core location core size eggshell units morphology eggshell units-thin section monoautochronic paired eggs egg spatial position in clutch air cell clutch spherical to sub-spherical 13.25 cm 11.25 85% nodular-single nodes and coalesecent nodes 0=0.58 mm H=0.28mm straight or Y shape 0 = 0.08 -0.2 mm round and funnel shape 0.52-0.87 mm 0.15-0.29mm 1.31mm 1 acicular 1.31 mm 1 0.19 mm above MT 0.034 mm 40° flare up to 30% of height to nearly // convex growth lines to flat higher no no random ? random Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Egg shape Egg greater diameter Egg smaller E g g sfte lP t& ie ss Ornamentation (morphology) Ornamentation (size) Pore canal (morphology) Pore canal PorePpelture (morphology) Internodular SDHCC pore aperture size Number of layers Layer 1 (morphology) Layer 1 Corporation Core size Auca Mahuevo titanosaurid eggs spherical to sub-spherical 12.5 to 14 cm 10 to 20% smaller 0.6 to 1.31mm pristine 1.31mm nodular-single nodes and coalesecent nodes 0=0.58 mm H=0.28mm straight or Y shape 0 = 008 -0.2 mm round and funnel shape 0.52-0.87 mm 0.15-.029mm 1 acicular 1.31 mm 0.19 mm above MT 0.034 mm U l C O o c < ■ - * • tr > 3 o > Neuquem, Soriano Salitral Moreno Laguna Mayo Calvo et al., 1997 Faccio, 1994 Powel, 1992 Sige, 1968 spherical to spherical spherical ? sub-spherical 16 cm 17-20 cm 15-23.5 cm ? ? ? ? ? 1.7-2.1 mm 2.5-5 mm, mostly 5mm and 1mm 0.81-1,44mm nodular-single nodes 5 mm nodular-single nodular-single nodes nodular-single and coalesecent nodes and and coalesecent nodes nodes and nodes coalesecent nodes smooth shell coalesecent nodes 0=0.5-0.9 mm H=0.8-1.3 mm ? ? straight straight and ? straight 0 = 007-0.110 mm multicanicular ? ? ? ? ? ? round-elleptical ? ? ? and funnel shape ? ? 0.3-0.4 mm ? ? 1 2 ? 1 acicular microfibrous ? acicular 1.7-2.1 mm? ? ? 0.81-1.44mm 0.19 mm above MT ? ? ? 0.35 mm?? 0.39-0.64 mm ? ? Table 3.2: Comparison o f putative titanosaur e g g a n d eggshell characters among five degree of compaction of the Auca Mahuevo eggs or as differences in value estimation since none of the eggs reported by Calvo et al. (1997) is whole. Likewise, differences in the eggshell thickness could be explained by the limited sample measured by Calvo etal. (1997). Faccio (1994) reported the presence of 27 spherical eggs from the Mercedes Formation (Senonian) in Soriano (Uruguay). The eggs were found with intact lower halves, and upper halves reduced to fragments orientated concave up at the bottom of the eggs, a taphanomy suggesting the embryos hatched before burial, creating what is referred to as a hatching window in the eggs (Breton et al. 1986; Faccio 1994). The diameter of these eggs ranges between 170-200 mm and their eggshell thickness varies from 2.5 to 5 mm, although in most samples it is close to 5 mm (Table 3.2). These values, much greater than those of the Auca Mahuevo eggs, easily distinguish the eggs of the Mercedes Formation. In addition, the Mercedes Formation eggs display a nodular surficial ornamentation, in which nodes are 75% higher than those from Auca Mahuevo. Pores in the Uruguayan eggs are far more numerous and differ from those of the Auca Mahuevo eggs by their shape and ramification pattern. Although Faccio (1994) mentions the existence of two structural layers in the eggshell of the Mercedes Formation eggs, close examination of the published SEM images questions such a condition. The eggs from the Mercedes Formation differ in many characters from those of Auca Mahuevo and could hardly be associated with the same dinosaur species. Powell (1992) briefly reported titanosaurids eggs found in two stratigraphic layers of the Allen Formation (Maastrichtian) in the locality of Salitral Moreno, Rio Negro Province (Argentina). Published SEM images and Powell’s (1992) succinct description indicates that two types of eggshell structures are present among the studied sample from this locality. Both egg types exhibit a nodular surficial ornamentation and diameters ranging from 150 to 235 mm (Table 3.2). Yet, while 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. one type has eggshell thickness values of 5 mm and structural organization similar to the eggs from the Mercedes Formation of Uruguay, the thickness of the other type of eggshell measures only 1 mm. According to Powell (1992), the latter type of eggshell can be subdivided into three different morphotypes on the basis of their surficial ornamentation, which varies from simple nodes to coalescent nodes to a smooth eggshell. Although the limited information provided by Powell (1992) hampers further comparisons with the Auca Mahuevo eggs, the larger size of some of these eggs and thicker eggshell clearly differentiate them from those of the latter locality. Comparisons with the thinner shelled type are more tentative because the lower value of the size range approaches that of the Auca Mahuevo eggs and its eggshell thickness falls within the Auca Mahuevo range. Sige (1968) reported dinosaur eggshell material from the Early Maastrichtian Vilquechico Formation at Laguna Umayo, a site near the Peruvian town of Puno. More material from that locality was collected and described by Kerourio and Sige (1984). Even though no dinosaurian remains were directly associated with the eggshell, Vianey-Liaud et al. (1997) re-evaluated 21 eggshell fragments (HEC 380-1, 381-1, 382-1, and 433) from this site and assigned them to the cosmopolitan oospecies Megaloolithus pseudomamillare, an oologic species traditionally related to titanosaurids. The thickness of these samples varies from 0.81 to 1.44 mm according to their degree of weathering (Table 3.2). Surficial sculpting shows single nodes and other nodes that coalesce into longer tubercles. The straight narrow pores open at the surface creating a funnel-shape depression. However, no other detailed description is supplied, but SEM and thin section images provided in the 1997 publication coupled with the brief structural description could lead to the interpretation that these Peruvian eggshells share enough similarites with those from Auca Mahuevo and on this basis could perhaps warrant their affiliation with titanosaurs. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONCLUSION The abundance of eggs and eggshells in Auca Mahuevo provide an unique opportunity to assess the variation of many oological characters, whether these variations are the product of taphonomic and particularly diagenetic processes. The tremendous variations observed in the eggshell thickness and surficial ornamentation are sufficient to warrant against the erection of new parataxonomic oological species without prior extensive and thorough sampling of the material. Eggs and their eggshells are useful indicators of the paleobiological environments and conditions where they were laid and incubated. For example it can be inferred from the internal and surficial eggshell structures that Auca Mahuevo titanosaurs may have laid and incubated their eggs in nests covered with vegetal mounds creating an environment with an elevated moisture content. Furthermore, the titanosaurid eggs from Auca Mahuevo provide oological synapomorphies to the skeletal characters diagnosing this dinosaur group. These are the presence of: sub-spherical eggs with a greater diameter of 12.5 to 14 cm, eggshell ornamentation consisting mostly of single nodes of 0.58 by 0.28 mm but with occasional coalescent hemispheric nodes; funnel-shaped pore aperture in between each node that are connected to straight or Y-shaped vertical pore canals transecting the entire eggshell thickness; each of these vertical canals coincide with a horizontal counterpart of equivalent diameter; the eggshell thickness, although varying according to the degree of diagenesis, averages 1.31 mm for well-preserved eggs and is composed of one single calcitic acicular structural layer. 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Considering the character list associated with the titanosaurid eggs of Auca Mahuevo and comparing that with puplished descriptions of unassociated oological material in South America putatively assigned to this group of dinosaurs, leads to the conclusion that presently only the oological material from Neuquen (Megaloolithus patagonicus ) and that from Peru (M pseudomamillare) could be interpreted as belonging to titanosaurid dinosaurs. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IV Egg, nesting behavior and egshell structure of the dromaeosaurid Deinonychus antirrhopus: (Taphonomic level 2 and taxonomic level 1) INTRODUCTION In 1931 an American Museum of Natural History (AMNH) field party explored the Early Cretaceous Cloverly Formation of southern Montana under the leadership of Barnum Brown. Near Cashen Ranch, on the Crow Indian Reservation, they discovered a partial skeleton of the dromaeosaurid Deinonychus antirrhopus (AMNH 3015) associated with a specimen of the omithopod Tenontosaurus tilletti (AMNH 3014) as substantiated by Barnum Brown’s 1931 field report (AMNH paleontology archives). An aerial photograph published by Ostrom (1970) and only a few details of the quarry site are preserved making future field work difficult at that site. However, most of the recovered skeletal elements of both specimens have since been prepared, and the D. antirrhopus individual is among the most complete described specimens of this taxon (Ostrom, 1969), is now on display at the American Museum of Natural History. For unknown reasons, a number of weathered, surface-collected bones and some small pieces of the limey matrix with bone fragments were left unprepared in AMNH collections. Examination of these small blocks revealed the presence of large quantities of eggshell as well as the circular cross-sections of some rod-like bones in two small blocks of matrix. Although a hand-written label identified these skeletal fragments as sections of ossified tendons, further preparation revealed these bones to be from the caudal midline of a theropod gastral basket, and direct comparison with the hypodigm of D. antirrhopus specimens confirmed their identity as of this species (Makovicky and Grellet-Tinner, 2000). In this chapter, the egg and eggshell morphology are described, and the 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. associated skeleton is used to infer the paleobiology of this dromaeosaurid. This new information is important because despite a string of recent discoveries of maniraptoran eggs intimately associated with either embryos (Norell et al., 1994; Mateus et al., 1997; Varricchio et al., 2002) or adults (Norell et al., 1995; Varricchio et al., 1997; Clark et al., 1999; Varricchio et al., 2002), AMNH 3015 is the only associated oological material of dromaeosaurid dinosaurs yet described. Given the phylogenetic proximity between this clade and birds (Gauthier, 1986; Sereno, 1997; Makovicky and Sues, 1998), these observations may have broad implications toward understanding the distribution of oological (egg, clutch, and eggshell) characters among maniraptoran theropods, aspects of eggshell evolution, and levels of homoplasy in both birds and their nearest relatives (Makovicky and Grellet-Tinner, 2000; Varricchio et al., 2002) Material Materials: parts of articulated gastralia, and a crushed egg and eggshell fragments within four small blocks of matrix associated with the partial skeleton of D. antirrhopus (AMNH 3015). This material derives from a carbonaceous claystone, near the middle of the Cloverly Fm. (Member V of Ostrom, 1970), in the “Cashen Pocket” locality on the Crow Indian reservation, southern Montana, T: 4S R: 29 E, Big Horn County (Ostrom, 1969, fig. 1). DESCRIPTION The largest block of matrix holds the medial tips of two left lateral elements, the medial parts of three medial gastral elements of the left side, and two of the right side as well as what appears to be a fused midline elements as is sometimes seen at the caudal end of theropod gastral baskets (Claessens, 2004). The gastralia are preserved in articulation, and the terminal midline ends of the left lateral gastralia articulate 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. along a shallow groove on the anteroventral face of the two anteriormost medial left gastralia. The medial gastralia from opposite sides of the midline meet each other at acute angles in an imbricate pattern as in the posterior part of the gastral cuirasse in Velociraptor mongoliensis (IGM 100/ 985), omithomimids (Sternberg, 1933) and tyrannosaurids (Lambe, 1917). Isolated gastralia of D. antirrhopus (YPM 5247, 5251) bear a small oval articular facet just posterior to such a process or tab near the midline end (Ostrom, 1969; Makovicky and Grellet-Tinner, 2000). The last segment has a U-shaped apex and appears to consist of fused midline gastralia. A similar U-shaped gastralia is known from other theropods including Troodon formosus (Russell, 1969) and tyrannosaurids (Makovicky and Currie, 1998; Makovicky and Grellet-Tinner, 2000; Claessens, 2004). In addition to the articulated gastralia, the largest block contains a large piece (ca. 2cm x 4cm) of cortical bone, which presumably originated from the tibia of the specimen, which is the only large hind limb bone collected from this specimen. Unfortunately, the specimen is currently inaccessible, so it is impossible to match the section of cortex with any part of the tibia in order to attempt a postmortem reconstruction of the relative position of the hindlimb to the abdomen. Several thin sections of the gastral fragments were made, originally with the goal of discerning them from omithischian epaxial tendons, which they had tentatively been identified as in the collections (Makovicky and Grellet-Tinner, 2000). Two sections of the D. antirrhopus gastralia showed a pattern in which most of the central part of the gastralia were heavily remodeled by several generations of secondary osteon growth and the cortical part of the bone displays six tightly packed rings of dense, lamellar bone, adjacent to the primary osteons where remodeling has not occurred. Annular counts in gastralia have been demonstrated to be reliable indicators of age in alligators and other reptiles (Erickson et al., in press), which suggests that AMNH 3015 was at least in its seventh year at the time of death. Varrichio (1993) 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. showed that the maniraptoran T . formosus, with roughly the same body size and mass as D. antirrhopus, would have reached somatic maturity within 3-5 years if lines of arrested growth are interpreted as annual markers. The crushed egg fragments are preserved in small blocks of buff, limey mudstone matrix, and occur in accumulations composed of up to three layers of eggshells, sometimes including a zone of diagenetic calcite crystals between two of these layers (Fig. 4.1 A). In the initial study (Makovicky and Grellet-Tinner, 2000), a near continuous curve of eggshell fragments in the largest preserved block of matrix was interpreted as the cross-section of a single egg that was crushed between the articulated gastral segment and the large piece of cortical bone described above. The fusilinid shape of this cross-section (Fig. 4. IB), with its multiple stacked eggshell fragments on its side, is reminiscent of the preservation of eggs that have been buried intact and then crushed by compressive forces at a later taphonomic stage (Mueller- Towe et al., 2002). Subsequent preparation revealed a large, equatorial section of a single egg underlying and in contact with the preserved midline series of gastralia (Fig 4.1C). Unfortunately, the polar ends of this egg were lost during excavation, leaving a distinct cross section at either end. The smaller of the two cross-sections is toward the caudal end of the gastral basket and measures roughly 3.05 by 0.71 cm, whereas the larger measures approximately 3.48 by 1.83 cm. The lower surface of the egg appears to be crushed and folded inward, as evidenced by multiple layers of stacked eggshell fragments, and the cross sectional measurements thus represent underestimates. The egg is oriented with the external surface convex up against the gastralia. The exposed eggshell surface does not exceed 7.53 cm by 4.55 cm. The eggshell is criss-crossed by white veins of a diagenetic mineral indicative of a diagenetic fluid phase after compaction (Fig. 4. ID). The eggshell surface is sculpted with an ornamentation consisting of anastomosed bumps and ridges, (Figs. 4.1C, D, E, F) which are not aligned with any major symmetry axis of the egg, a 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.1 A. TLM view of random accumulations of D. antirrhopus eggshell. B. Fusilinid cross-section (black arrows) with multiple stacked eggshell fragments (white arrows) indicate that D. antirrhopus egg was intact at the time of burial but was crushed during fossilization. C. Equatorial egg section (black arrows) closely apposed to gastralia (white arrows). D. White veins of calcitic (black arrows) criss-cross the eggshell indicative of at least two diagenetic phases. E. Eggshell outer surface with anastomosed bumps and ridges. F. SEM view of the eggshell surface with presence of two pore orifices (black arrows) in the recesses of the ornamentation. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.1 m m m m 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. condition unlike that observed in the Late Cretaceous oviraptorid Citipati osmolka (see Chapter 5). In SEM radial view this ornamentation appears as wavy ridges (Fig. 4.2A). Although rare, the pore apertures seem to be evenly and randomly distributed in the troughs between or on the apices of the surficial ornamentation. Observed pore openings are elliptical and the pore measurements show little variation averaging 130 by 60 pm in size (Fig. 4.2B). Only a single pore canal is visible for its entire cross section through the eggshell. It runs perpendicular to the eggshell from most of its length before turning obliquely to exit on the side of one of the external ridges (Fig. 4.2C). Although filled by a secondary diagenetic mineral, its walls do not display sign of chemical dissolution and its diameter is relatively constant averaging 15 pm. One organic filament crosses through the pore midway through the eggshell, testifying to the very high degree of preservation (figs. 4.2C, D). The average eggshell thickness is 0.44 mm but could reach 0.6 mm when accounting for the height of the surficial ornamentation. SEM examination of eggshell radial section reveals that the eggshell is composed of two layers (Table 4.1). The contact between these layers is aprismatic (Grellet-Tinner and Nor ell, 2002), a type of contact where layer boundaries are clearly defined, and eggshell units are not visible in the outer layer 2 (fig. 4.2D). The shell units in the inner layer (layer 1) are clearly visible on either the exposed or the freshly broken surfaces (figs. 4.2D, E) and consist of fan shaped cones (mammillae) 0.054 mm wide at the level of the organic core and 0.173 mm thick at the boundary between layers (Table 4.1). The mammillae average height equals their widest diameter (0.173 mm). The fan shaped cones are quite uniform in shape and consist of acicular calcite crystals with rhombohedral cleavage, a character also observed in oviraptorids (see Chapter 5). No resorption pits are observed between the organic cores, indicating that the egg was either freshly laid near the time of death and burial, or that it may have been infertile. Layer 2 is sharply differentiated from Layer 1 and composes the rest of the eggshell in radial 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.1: Deinonychus antirrhopus egg and eggshell parameters egg shape elongated egg greater diameter ? egg smaller diameter 6.5 - 7 cm (estimated) egg diameter ratio ornamentation morphology ostomozed bumps and ridges ornamentation size pore canal morphology straight with an elbow close to the surface pore canal size 0 = 15 microns pore aperture morphology elliptical internodular space NA pore aperture size 130 by 60 microns eggshell thickness 440 microns number of layers 2 eggshell character state aprismatic layer 1 morphology acicular layer 2 morphology continuous at 90 degrees layer 3 morphology NA layer 1 size 131 microns layer 2 size 180- 390 microns layer 3 size NA ratio total thickness/Ll 3.35 ratio L1/L2 0.34 ratio L3/L2 NA core location core size eggshell units morphology fan shaped, 54 at their base to 173 micron at eggshell units-thin section spaced and thick°m1?i microns, compact and thin to the. outer surface monoautochronic yes (inferred) paired eggs yes (inferred) egg spatial position in clutch ? air cell yes (inferred) clutch no empty space in center Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.2 A. SEM view of the eggshell cross-section (white arrows). Note the wavy surficial ornamentation. B. Note the common elliptical shape (130 by 60 microns) of a pore orifice in SEM. C. Note the elbow and then straight shapes of a pore canal (black arrows) in TLM. D. A fine organic filament (black arrow) crosses the width of the pore canal attesting to the high degree of preservation of this eggshell. E and F. TLM and SEM views of the same eggshell section. Note that vesicules (white arrows) in SEM are analoguous to organic line (white arrows) in TLM. Note the sharp boundary (black arrow) between layers 1 and 2, defined as aprismatic by Grellet-Tinner and Norell (2002). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.2 ■ B B B B M fcfcT r k 5?w « . ! ? » ' j w ,»gW )- j w r a K ^ Hf - b J m h I L , . , ■ ■ " -1 " ? a * - A ' ■ ■ ' / I jR jf f I 15.0 KV 1 0 0 u m 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. section (figs. 4.2C, E, F). The outer surface appears undulatory in radial view due to the surface ornamentation (figs. 4.1 A, 2 C). The thickness of this layer varies from 0.18 mm to 0.39 mm between the top of a tubercle and the bottom of a trough. The eggshell is relatively thin and compares favorably with oviraptorid eggshell in a number of crystallographic and ultrastructural details (Norell et al., 1994; Dong and Currie, 1996). Light microscopic examination of eggshell thin sections confirms the shape, size and acicular composition of the base of the eggshell units in the inner layer (Figs. 4.2C, 3A). However in addition to the SEM observation, polarized light microscopy reveals the presence of former organic lines oriented parallel to the eggshell surface (Fig. 4.3B). They are thin, tightly compacted and they connect with each other at the apex of the cones of layer 1 (Fig. 4.3C). In one thin section (Fig. 4.3B) the spacing of these lines, which fluoresces under CL, matches the spacing between a linear array of voids in the SEM images of the corresponding sample lending weight to our interpretation of these lines as voids (figs. 4.2E,F). The micron size of these voids also supports the idea that they are the molds of the original protein fibers embedded in the eggshell. The contact between the cones and layer 2 is abrupt without any transgression of the acicular crystals in the adjacent layer (figs. 4.2C, E, F, and 4.3 A, B). Layer 2 can be differentiated into two distinct sections based on the thickness and density of the organic lines. The inner section adjacent to layer 1 forms a band that is nearly constant in thickness averaging 130 pm in width (fig. 4.3C) and is marked by thick and widely separated organic lines. The outermost boundary of this sub-section only follows the undulatory outline of the outer eggshell to a moderate degree (Fig. 4.3C). The outer section of layer 2 is characterized by thinner and more tightly spaced organic lines than its inner sub-section (fig. 4.3D). Its maximum thickness varies according to troughs and ridges of the eggshell outer surface. 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.3 A. TLM view of layer 1 and lower section of layer 2. Note the barrel shape of the eggshell units (white arrows) and the concentration of compact and thin organic lines at the apex of theses eggshell units. B. TLM view of layers 1 and 2. Note how thicker layer 2 is and the presence a two sub-layers within layer 2 based on the size and compaction of organic lines. C. Detail TLM view of the demarcation between the two sub-layers and of the size of the organic lines. D. Note the smaller size of the sub-layer closer to layer 1 (white double arrow). Organic lines stay horizontal (do not follow the profile of the eggshell surface). E and F. Exquisite preservation of the eggshell structure. Taphonomic and diagenetic processes did not alter the internal structure of these eggshell fragments despite, as indicated by CL, some localized replacement of its former organic component by diagenetic material that fluoresces similarly to the encasing matrix. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4.3 '■ ■ ' ■ -v''’^ L - 1 * * * ; V . > ' . : E E gjjsB-. - 15.0 KV 167 urn Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Overall, the preservation of the eggshell structure is exquisite and the effects of taphonomic and diagenetic processes did not alter the internal structure of these eggshell fragments despite, as indicated by CL, some localized replacement of its former organic component by diagenetic material that fluoresces similarly to the encasing matrix. CL clearly supports the presence of two aprismatic structural layers. In addition, the strong luminescence follows in layer 2 and lines that are parallel to each others and to eggshell surface, while the luminescence of layer 1 follows paths that are perpendicular the layer’s 2 lines and those created by the radiating acicular crystals and the thin organic among them. Considering both the exquisite preservation of this material and that luminescence mostly follows the paths of ghost protein matrix in the eggshell, there is a strong possibility that the diagenetic process was biomediated. This process, where bacteria consume the original interwoven proteins and fix the cations from the surrounding pore fluids, during fossilization has been recognized in multiple instances in the fossil record (Reid et al., 2000; Visscher et al., 2000; Maliva et al., 1989; Grellet-Tinner, in press). DISCUSSION Three alternative hypotheses for the association between the egg beneath theropod skeleton were considered by Makovicky and Grellet-Tinner (2000). The first was a chance association between the two, an explanation that seems at least somewhat plausible given the purported close association of the D. antirrhopus specimen with a specimen of T . tilleti. The second alternative was that the egg was either laid by D. antirrhopus and that the animal died while brooding it. The theropod synapomorphies of the eggshell and the precise location of the egg directly in contact with the articulated gastralia support this view. The third hypothesis considered, mostly on historical grounds, is that D. antirrhopus was feeding on the egg at the time of its death. However, this alternative which was deemed unlikely since theropod 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. specimens historically interpreted as dying in the act of egg feeding (Osborn, 1923) were subsequently reinterpreted as brooding their alleged food (Norell et al., 1994; Dong and Currie, 1996). Here these hypotheses will be reconsidered in greater detail, applying new taphonomic and structural evidence both from the egg and the associated skeleton as well as from recent findings on other associations between eggs and adult specimens. Although a purely random association between AMNH 3015 and the aposed egg cannot be directly falsified several taphonomic considerations render it unlikely. First, the articulated nature of the gastralia, which in life are imbedded in the abdominal wall and easily disarticulated or lost after death, suggests that the specimen was at least partially articulated when discovered. Well-preserved gastral baskets are only found in fully or almost fully articulated specimens (e.g. Lambe, 1917; Osborn, 1917; Sternberg 1933; Chure and Madsen, 1996; Norell and Makovicky, 1997; Kobayashi et al., 1999), and these delicate elements are among the first parts of the skeleton to disintegrate during decay. Even in well preserved and largely articulated specimens such as the holotypes of Caudipteryx zoui (Ji et al., 1998) and Compsognathus longipes (Ostrom, 1978; Griffiths, 1993), the gastral basket is disarticulated, thus suggesting that the degree of articulation is a sensitive indicator of how well articulated a specimen was at the time of burial. Second, the articulated nature of the gastralia of AMNH 3015 suggests that the specimen did not experience long subaerial exposure or transport prior to burial and that most likely it died at or near the point of discovery. In addition, its occurrence in a fine clay stone that is generally devoid of allochtonous clasts (Ostrom, 1970) is indicative of little or no transport. Taphonomic considerations such as its crushed and infolded nature suggest that the egg beneath AMNH 3015 was intact at time of deposition. The crushing indicates that the egg was not infilled and therefore neither cracked nor opened prior to and 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. during burial. The overlap of the eggshell fragments near the equatorial portion of AMNH 3015 suggests the lateral expension of the shell was limited as the egg was crushed because of an early lithification of the encasing sediment (Mueller-Towe et al., 2002), thus further supports the coeval sedimentation of the egg and the associated skeletal remains. Furthermore, the lack of resorption craters at the tip of the cones in layer 1, indicates that the egg was unhatched (Grellet-Tinner, 2000), and therefore corroborates the possibility that it was intact at the time of burial. To sum up, the taphonomic characteristics of both the skeleton and the egg strongly advocate that the egg was in place beneath the skeleton at the time of deposition. Therefore, although the association of the egg with AMNH 3015 fits only in the taphonomic level 2 as described in Chapter 1, this specimen is regarded here as a taxonomic level 1. This conclusion is further supported by the presence of two or more eggshell layers is a synapomorphy of at least Tetanurae (Makovicky and Grellet-Tinner, 2000; Varricchio et al., 2002) which clearly identifies it as that of a theropod egg. Recent discoveries of brooding behavior in the oviraptorids Oviraptor mongoliensis, C. osmolka and the troodontid T . formosus indicate that brooding had already evolved among primitive maniraptorans (Norell et al. 1995; Dong and Currie, 1996; Clark et al., 1999; Varrichio et al, 1997), and that it would be expected in Dromaeosauridae. In this respect, it is worth noting that, as for the egg of D. antirrhopus, several eggs are in contact with the gastralia of C. osmolka (Clark et al., 2001) or only separated from the eggs by a thin layer (< 3cm) of sand (Clark et al., 1999). Similarly, the lithology of a preserved nesting trace of T . formosus interpreted together with the find of an adult in direct contact with the eggs (Varrichio et al., 1999) indicates that one pole of the eggs was aerially exposed and probably brooded in direct contact with the parent. Although the egg beneath AMNH 3015 is not complete extrapolations of its size and volume can be made on the basis of the pelvic opening of the adult 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dinosaur. Egg diameter is related to the age of the female and the size of the pelvic opening (Carpenter, 1999). Younger females have a tendency to lay smaller eggs than mature females of the same species. Withstanding this bias, it is well accepted that the maximum egg diameter cannot exceed that of the pelvic opening of the female (Carpenter, 1999). Ostrom (1976) estimated the pelvic opening of D. antirrhopus at 7 cm after restoration. This value gives an upper limit to the egg maximium diameter, a dimension well within the 6.5 to 7.2 cm range of C. osmolka egg (Clark at al., 1999). In addition to this shared similarity, these two theropod species share the same eggshell thickness and structure. D, antirrhopus and C. osmolka body mass could be estimated by proxy of their femoral diameters. D. antirrhopus 122 and C. osmolka 110 femoral diameters advocate that these theropods had comparable body mass. Hence, in view of the similarity of body masses of D. antirrhopus and C. osmolka, their nearly identical eggshell structure, coupled with the estimated diameter of D. antirrhopus egg, it is likely that AMNH 3015 egg volume would be comparable to that of C. osmolka, thus implying that the length of D. antirrhopus egg would average 18 cm as that of C. osmolka. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONCLUSION The only known egg of D. antirrhopus (AMNH 3015) is the oldest and only occurrence of an associated dromaeosaurid egg in the fossil record. Although limited, this oological material provides important information for a better understanding of the theropod phylogeny and the nesting behavior of dromaeosaurids. In particular, the spatial position of the egg in relation to the gastralia advocates the existence of brooding behavior in Dromaeosauridae, an interpretation congruent with phylogenetic predictions. These discoveries emphasize the importance of oology in the evolution of dinosaurs and support the hypothesis that brooding behaviors was present in many non-avian theropods that predate the rise of birds. In addition, the eggshell structure of AMNH 3015 indicates that a closer phylogenetic proximity to oviraptorids than troodontids. 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER V Oviraptorid dinosaurs and the species Citipati osmolka from the Gobi desert (Taphonomic 1 and taxonomic level 1) INTRODUCTION The phylogenetic relationship between oviraptorids and other dinosaur lineages is still debated and until recently rests solely on osteological data. Maryanska et al. (2001) considered Oviraptorosauria as an avian clade in contrast to the non- avian maniraptors, a placement usually assigned to this group (Makovicky and Sues, 1998). Others have placed this group within a polytomy including Troodontidae and Ornithomimosauria (Bakker et al., 1988). However, a large amount of informative oological material has been found since the 1920’s. In particular, three life assemblages of oviraptorids brooding their nests and an oviraptorid embryo in ovo (Norell et al., 1994, 2001) have been collected from the Campanian Djadokhta Formation of Mongolia and from its equivalent in Chinese Inner Mongolia (Dong and Currie, 1996). This chapter offers a novel examination of a spectacular fossil egg clutch (IGM 100/979) (fig. 5.1 A), an embryo in ovo (IGM 100/971) (fig. 5.IB) of Citipati osmolka, both belonging to the same oviraptorid species, with the aim of augmenting the phylogenetic character data set of this clade and to better understand the nesting behavior of this remarkable theropod species. Material Taxonomic identification of this oological material rests upon the data collected from IGM 100/979, IGM 100/971, IGM 100/1125 Comparison of the egg shape, and the eggshell external and internal morphologies within these various specimens not only insures that they each belong to the same species but also permits us to gather 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. information from one that might not present in another. Although, IGM 100/971 is not a complete egg (Fig. 5.1 B), it preserves an embryo in ovo allowing its taxonomically identification to the oviraptorid species C. osmolka (Clark et al., 1999), and has enough eggshell surficial (fig. 5.1C) and structural characters to positively compare it with IGM 100/979 and IGM 100/1125. Furthermore, the postcranial skeletal remains of an adult C. osmolka (fig. 1 A) found in a brooding position over its egg clutch (IGM 100/979) supports a level 1 taphonomic association. CL examination of this material shows no luminescence, thus indicating no noticeable diagenetic replacement in the eggshell structure. IGM 100/971. C. osmolka embryo found in a partial egg, collected in 1993 from the exposures of the Djadokhta Formation at Ukhaa Tolgod. IGM 100/979. Partial skeleton of C. osmolka found on the top of an egg clutch composed of at least of two superposed egg rows, collected in 1993 from the exposures of the Djadokhta Formation in Ukhaa Tolgod. IGM 100/1004. Partial skeleton of C. osmolka found on top of an egg clutch composed of at least of two superposed egg rows, collected in 1995 from the exposures of the Djadokhta Formation at Ukhaa Tolgod. IGM 100/1125. A pair of C. osmolka isolated eggs DESCRIPTION The egg type of IGM 100/971, 979, 1004, 1125 have been assigned to the elongaloolithid oofamily (e.g., Zhao, 1975; Mikhailov et al., 1994), a parataxonomic classification mostly based on the egg shape and eggshell surficial ornamentation. The outer eggshell surface of the eggs of IGM 100/971, 979, 1004, 1125 displays a series of fine longitudinally oriented ridges (linearituberculate in the traditional parataxonomic classification) that grades into dimples or smooth areas near or at the poles (fig. 5.1C). The brood (IGM 100/979) consists of fifteen eggs that are partially 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exposed below the skeleton of the adult oviraptor and arranged in pairs. Although the clutch is probably not complete as evidenced by the presence of an isolated egg (Clark et al., 1999), it is fair to assume that few unexposed eggs remain hidden in the matrix. Paired eggs lay sub-horizontally (fig. 5.1 A) and are radially aligned in two superposed layers (Clark et al., 1999). In contrast to the Troodon formosus and Byronosaurus jaffei clutches (see following 2 chapters), the eggs of C. osmolka leave an empty space in the center of the clutch. Although none of these eggs are completely exposed, their shape appears elongate with parallel sides and with two rounded poles. Originally, one of the poles was reported less rounded than the other (Norell et al., 1995) but a subsequent description implies that they are similar in shape and size (Clark et al., 1999). According to Clark et al. (1999), the most exposed egg is at least 18 cm but less than 19 cm in length and a width variation from 6.5 to 7.2 cm in width is noted from egg to egg. IGM 100/971, the partial egg containing an embryo, is too incomplete to obtain a more accurate description of the egg shape (Norell et al., 1994) but IGM 100/1125 (a pair of eggs) enables to determine that these eggs clearly show that the two poles are not identical (fig. ID) as noted in Clark et al., (1999). One pole, that facing down (sub-horizontal position in the clutch) is slightly more tapered than the other facing up that likely contained a proto-air cell. Eggshell fragments were removed from the three IGM specimens. The eggshell is relatively thin averaging 570 pm and ranging from 500 to 641 pm according to whether the measurement was taken at the apex of a ridge or in at the bottom of a trough, and compares favorably with that of D. antirrhopus (Makovicky and Grellet-Tinner, 2000; Chapter 4). Its radial section viewed in SEM displays a bi-layered structure (Fig. 5.IE) where layers 1 and 2 average 169 pm and 401.59 pm respectively. The contact between the two structural layers is aprismatic (Fig. 5.IE), character as defined by Grellet-Tinner and Norell (2002). In addition to the aprismatic 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5.1 A. IGM 100/979. Life assemblage with at least 2 superposed rows of eggs and a brooding (oviraptor) Citipati osmolka. B. IGM 100/971. C. osmolka embryo in ovo. C. IGM 100/971 external surface. Note the overlapping eggshell fragments indicative of compressive forces after burial. D. IGM 100/1125. Paired eggs indicative of monoautochronic ovideposition. Note the slight angles between these two eggs also observed in Macroelongatoolothid eggs, and the minor but visible difference in size between the two poles indicative of the presence of proto-air cells. E. Presence of two aprismatic layers (with arrows show the demarcation) with layer 1 consisiting of shell units made of acicular and radiating clacitic crystals. F. Shell units radiate out of a former organic core (black arrow). 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5.1 WIND *, * » 15.0 KV 42.9 um Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contact, there is a net increase in amount of vacuoles in layer 2 comparatively to layer 1 (Fig. 5.IE), a feature that helps to discriminate layer 1 from layer 2. The base of the shell units in layer 1 are clearly visible on both the exposed and freshly broken surfaces (Fig. 5. IE) and consist of bundles of acicular crystals of rhombohedric form that create fans of equal width and height radiating from the cores (Fig. 5. IF) to the boundary between layers 1 and 2 (Fig. 5.2A). Although their shape slightly differs with that of D. antirrhopus (Makovicky and Grellet-Tinner, 2000; Chapter 4), these fans average 169 pm in height, a value that favorably compares with that of D. antirrhopus. TLM of this eggshell confirms SEM observations of the shape, size and acicular composition of the base of the eggshell units in layer 1 (Fig. 5.2A). However in addition to the SEM observation, TLM reveals the presence of organic lines perpendicularly oriented to the cones and shell units. This sharply differentiated layer composes the rest of the eggshell radial section (Fig. 5. IE) and consists of calcitic crystals with a C-axis that are orientated at a 90 degree angle with respect to those of layer 1. The outermost section of this layer appears undulatory (Fig. 5.2B) due to the linearituberculate surficial ornamentation aligned with the long axis of the egg, a character congruent with previous observation (Norell et al., 2001). The thickness of this layer varies according to the surficial ornamentation from 331 pm to 472 pm. Black organic lines are visible in TLM throughout layer 2 (Table 5.1). Although called organic lines, these lines are likely the voids left by a proteineous mesh originally intertwined with the inorganic calcium carbonate of the eggshell. This observation is supported by SEM examination where the matching emplacement of these lines visible in light microscopy, consists of vesicles lined up in straight lines (Fig 5.2 D). The micron size of these vesicles supports the interpretation that they are the molds of the original protein fibers embedded in the eggshell. In contrast to those of D. antirrhopus, these black lines are equally compacted throughout layer 2 (Fig. 5.2C) and their wavy or sinusoidal appearance that does not always parallel the 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5.1: Citipati osmolka egg and eggshell parameters egg shape spherical to sub-spherical egg greater diameter 18.5 cm egg smaller diameter 6.7 egg diameter ratio 36% ornamentation morphology linearituberulate ornamentation size pore canal morphology straight pore canal size pore aperture morphology round internodular space pore aperture size 0 = 78 microns eggshell thickness 570 microns number of layers 2 eggshell character state aprismatic layer 1 morphology acicular layer 2 morphology continuous at 90 degrees layer 3 morphology layer 1 size 169 microns layer 2 size 401 microns layer 3 size ratio total thickness/Ll 3.37 ratio L1/L2 0.42 ratio L3/L2 core location core size eggshell units morphology fan shaped 169 microns at contact of LI eggshell units-thin section equally distributee?1?ut undulation in monoautochronic opposite to the surficial ornamentation yes paired eggs yes egg spatial position in clutch in circles and superposed up to 3 layers air cell small clutch empty space in center Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. troughs and apex of the eggshell outer surface. This feature was previously interpreted (Norell et al., 2001) as the result of erosion, but is viewed here as a character of biological origin. Although rare, a few pore canals and apertures are visible. These apertures are round (Fig. 5.2E) and the obtained measurements are consistent averaging 78.47pm in size. Pore canals cross the eggshell at right angle and directly connect the internal and external surfaces (Fig. 5.2F). DISCUSSION The combined discoveries of the oviraptor C. osmolka (IGM 100/979) on an egg clutch and of a co-specific embryo in ovo (IGM 100/971) changed the historical interpretation of the theropod behavior of oviraptorid theropods from one perceived as an egg robber to one interpreted as dinosaurs displaying an avian-like type of parental care (Clark at al., 2001). In addition, because the definite taxonomic identity of eggs requires the presence of identifiable embryonic remains in ovo IGM 100/971 provides an important bridge with other oviraptorid specimens (IGM 100/979-1004- 1125) and a link with the elongaloolithid oofamily they have been assigned to in the parataxonomic classification. Several taphonomic and behavioral inferences can be made from observing IGM 100/979, IGM 100/971, and IGM 100/1004. The presence of multiple superimposed eggshell fragments in IGM 100/971 suggests that the egg and its embryo were intact when buried but that they were submitted to compressive forces afterward. However, the presence of sediment identical to the surrounding matrix inside the egg would imply that the egg was subsequently open and filled up after burial. This taphonomic hypothesis is congruent with a first stage of microbial mediation that possibly contributed to the preservation of the articulated embryo in ovo and that is analogous with the preservation of soft tissue fossilization in the titanosaurid eggs from Auca Mahuevo (see chapters 2 and 3). This interpretation 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5.2 A. TLM view of the eggshell units that are as wide as high. Note the ubiquitous line (white arrows) between layers 1 and 2. B. SEM of the eggshell outer surface. Note the undulatory aspect of the linearituberculate ornamentation that differs with that of D. antirrhopus by the lack of anastomosed ridges. C. TLm view of the organic lines (black arrows) in layer 2. In contrast to those of D. antirrhopus these lines are equally compacted and either follow the aspect or are in opposition of the eggshell surficial ornamentation (as here shown). D. SEM view of lines of vesicules (black arrows) analoguous to the organic lines observed in TLM. E. SEM of two round pore apertures (black arrows). F. SEM view of pore canals that cross the eggshell thickness in a straight line and normal to both eggshell surfaces. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5.2 9QD um a ' / * : . 15.0 KV 15.0 KV 600 um Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. slightly differs from that proposed by Norell et al. (2001) where the opening of the egg prior to burial would have been a prerequisite to the preservation of the articulated embryo. Preservation of IGM 100/979 shows that there was little disturbance of the clutch and the skeleton on it, thus suggests that the adult specimen is preserved in the position in which it died. Recently, Loope at al. (1998) reinterpreted the depositional setting of the Djadokhta Formation fossiliferous sands as low energy debris flows from semistable eolian dunes, a geological setting congruent with the latitudinal paleobearing of the Djadokhta Formation estimated at 30 degrees north, a latitude known for desert environments. The burying sediments would come from destabilized sand dunes during large rainstorm events. When brooding, most modern birds would tend to abandon their brood at the approach of a deadly threat. The preservation of life assemblages such IGM 100/979 coupled with our knowledge about its depositional environment and behavior of modem birds suggests that the brooding oviraptorid could have been dead or nearly dead at the time of burial. Alternatively, modem paleognath birds namely D. novahollandiae are known to incubate their brood for 40 days with very limited food or water intake, initiating an energy depletion that could reach lethargic levels. If oviraptorids brooded their clutches for similarly long periods of time, it could explain their burial atop their egg clutches. At the same time this would advocate that the behavior of long brooding period predates the most recent common ancestor of modem birds. IGM 100/979 and similar other life assemblages (Dong and Currie, 1996) provides clear evidence that the center of the nest is devoid of any eggs. Oviraptorid eggs and those classified as elongatoolithids are commonly recovered in clutches where they are paired and in an horizontal to sub-horizontal position in one or more superposed circles with an empty space in the center. This arrangement contrasts with that seen in troodontids (see Chapters 6 and 7) and is possibly a synapomorphy 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the lineages laying elongaloolithid eggs. Several paleobiological interpretations could be inferred from these observations. First, oviraptorids and their allies had two functioning ovaries and produced paired eggs in a monoautochronous sequence (Clark et al., 2001). Second, the presence of multiple and superimposed egg rows could suggest that oviraptorids had a polyhandric nesting behavior coomparable to modem ostriches and rheas, whereas many females lay their eggs in a common nest that is protected by an incubating male. However, without further evidence this interpretation remains an interesting but very untested hypothesis. Third, the double aprismatic structure of the eggshell is cut by straight or nearly straight pores that are mostly located near the blunt pole. The sub-horizontal orientation of the eggs with the blunt pole up having a higher pore concentration coupled with the linearituberculate ornamentation in between the poles, thought to act as an air conveyor (Sabbath, 1991), may suggest that these eggshell structures facilitated the air circulation along the eggs to their blunt poles where the air cell was located. Such an implication suggests that the eggs were likely laid in a nest covered by plants and topped by the brooding parent. Fourth, the studies on shells of mollusks have shown that resistance to pressure increases with the number of structural layers, particularly when the internal crystallization varies in direction from layer to layer (Kamat et al., 2000). C . osmolka eggshells is composed of two aprismatic structural layers suggesting a coeval evolution between brooding behavior and accretion of a second layer in the eggshell structure. CONCLUSION The combined discoveries of IGM 100/979, IGM 100/1007, and IGM 100/971 provide a rare opportunity to learn the reproduction and nesting behaviors of C. osmolka. IGM 100/971 with its embryo in ovo is paramount to the taxonomic 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. identification of similar eggs and eggshells. The similarities of the eggshell structure of each specimen of C. osmolka indicates that IGM 100/979 was brooding its eggs and advocates that oviraptorids were not egg robbers as implied by their name. To date the slightly asymmetric eggs of C. osmolka depart from the shape of any other identified non-avian dinosaur eggs beside those of troodontids, indicating the presence of a proto-aircell. In addition, the paired and sub-horizontally positioned arrangement of the eggs in clutches is so far apomorphic to C. osmolka and their elongatoolithid allies strongly suggests that these dinosaurs still had two ovaries that functioned in a monoautochronous sequence. Whether brooding was performed by a male sitting on eggs laid by multiple females remains speculative, but the presence of multiple superimposed rows of eggs favor this interpretation. Regardless, it is clear that brooding, a precursor to avian incubation, was present in C. osmolka (Clark et al., 2002), and presumably all other oviraptorids, and evolved in parallel with the accretion of a second aprismatic eggshell layer, a mean to cope with increased external stress. C. osmolka live assemblages, eggs, eggshell structure, and clutches suggest that this oviraptor went a step forward in the direction of avian reproduction and nesting. However, our present oological knowledge does not support for the phylogenetic placement of C. osmolka within the Avian clade, in fact it supports the opposite. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VI Re-examination of Troodon formosus, a troodontid dinosaur from Montana (Taphonomic and taxononomic levels 1) INTRODUCTION Troodon formosus was the first discovered species of the troodontid family, a middle-size coelurosaur from the Late Cretaceous of North America first described by Leidi in 1856. Although there are no problems related to the taxonomic identity of adult troodontids (REF), the embryonic and oological materials have been the source of many debates (Homer and Weishampel, 1996) and confusion (Varricchio et al., 1997). Embryos, eggs and eggshells today accepted as of T . formosus, were previously assigned to the omithischian dinosaur hypsipholodont Orodromeus (Horner, 1982; Hirsch and Quinn, 1990; Horner and weishampel, 1988). As a consequence of this original misidentification, similar oological material was until recently assigned to the egg type “dinosaurid-prismatic” a parataxonomic group that is adjacent to, but does not include most of the avian and non-avian theropods (Hirsch and Quinn, 1990; Mikhailov, 1996, 1997) under the assumption that egg with similar eggshells were the product of hypsipholodont dinosaurs (Zhao and Li, 1993). Aside these internal issues, there is still a greater phylogenetic debate questioning whether T . formosus and the six troodontid species present in North America and North Asia during the Late Mesozoic (Clark et al., 2002) are the closest relatives to basal Mesozoic avians (Currie, 1985, 1987; Osmolka and Barsbold, 1990; Currie et al., 1993; Currie and Dong, 2001; Russel et al., 1993; Foster et al., 1998, Clark at al., 2002). Although numerous specimens of T . formosus were found associated with eggs, eggshell fragments, and nests (Varrichio et al., 1997; 2002), thus potentially bringing a wealth of characters to help in resolving the phylogenetic position of troondontids, it is not clear that this oological material has been correctly interpreted. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Therefore, a revision and new interpretation of T . formosus is offered here. Material MOR specimens were collected from the Late Cretaceous North American Two Medicine Formation in the Egg Mountain and Island localities in Montana. These specimens include: MOR 393 with 22 eggs suggesting egg parity MOR 363 with 22 eggs suggesting egg parity MOR 246 eggs with at least one embryo in ovo and forming an oval shape clutch of 90 by 50 cm MOR 748 with a partial adult on a clutch MOR 750 the clutch under the brooding adult MOR 393 with 22 eggs suggesting egg parity MOR 963 with 24 eggs in a rimmed nesting trace Personal SEM observations were made on an eggshell fragment from MOR 246 and clutch photos from MOR 676 and 675. It is necessary to point out a discrepancy, likely without major consequences, in the citation of the material used by Varricchio et al. (2002). Varricchio et al. (2002) description refers to a clutch (Mor 750) supposedly described in their 1997 paper. However, the 1997 paper does not mention anywhere MOR 750 but concentrates on MOR 363. Could MOR 363 be synonymous to MOR 750? Or are these two clutches unrelated and this is an overview of the authors? Assumption is made in this section that Mor 750 and 363 are the same. IGM 100/1003x. Byronosaurus jaffei egg clutch and a partial juvenile of the same species. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DESCRIPTION The overall egg length of T . formosus ranges from 12 to 16 cm and the eggs are elongated and asymmetrical (Figs. 6.1 A, B) and according to previous authors (Varricchio et al., 2002) their narrow and blunt poles averaging respectively 3 cm in diameter and their blunt end 7 cm. However, using their own figure and scale (Figure 4 B, page 570) the pointed pole would be only 2 cm and the blunt one 3 cm measuring the sample at their flattest sections. Notwithstanding the issue related to the pole size, the significant difference in shape and size between the two poles of T . formosus egg advocate for the presence of a bigger air cell than that inferred for C. Osmolka. Furthermore, the smaller pole of T . formosus eggs are geometrically identical to those of B. jaffei (submitted). Previous studies of T . formosus eggs highlighted the effect of taphonomic and diagenetic alterations on some eggs, namely striations on the outer eggshell surface (Varricchio et al., 2002) analogous to micro slikenslides for the former and a secondary calcite deposition for the latter (Fig. 6.1C). These two features can be misleading. For instance, the secondary calcite deposit could be interpreted as a third eggshell structural layer and the striation as part of eggshell ornamentation. SEM images of the studied sample (MOR 246) shows no striations, but it does not void the existence of those in other specimens (Varricchio et al., 2002). Although none of the MOR clutches are probably complete, the largest one (MOR 963) contains up to 24 eggs that stand vertically or at a slight angle with their narrow poles embedded in the sediments (Fig. 6.1 A). These 24 eggs (Fig. 6.ID) occupy an oval shape space of 45 cm by 56 cm within a 1 by 1 m rimed nest (Varricchio et al., 1999). Mor 963 reveals a long and narrow empty space in the center of the egg clutch giving it a bilateral symmetry (Varricchio et al., 1997). No such a central empty space was reported in Mor 246 (Varricchio et al., 2002), a clutch where the eggs stand sub-vertically leaning towards the center. Clutches of ^ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T . formosus would therefore exhibit two conditions; one with and another space in the center. The U-shaped rim surrounding the clutch of MOR 963 is composed of positioned of a different lithology than the fill sediments (Varricchio et al., 1997). The rim is approximatively 10 cm high and 21 cm wide delineates a 1 m by 1 m nest (Fig. 6.ID). To date, T . formosus is the only known theropod species for which the nesting structure is preserved. Egg parity has been mentioned for T . formosus (Varricchio et al., 1999, 2002) and supported by statistical analysis (Varricchio et al., 1997). This claim relies on the spacing among the narrow end of the eggs, their trend and plunge angles, and also on an apriori assumption that the eggs were laid in pairs (Varricchio et al., 1997). However, another statistical analysis by these authors (Varricchio et al., 1997) relying on the blunt poles (the poles not buried in sediments) came to the conclusion that parity could not be supported and the eggs were randomly positioned in a bowl-like depression. The arrangement of this clutch although seemingly negligible is important for paleobiological and phylogenetic considerations. Presence of paired eggs could imply that T . formosus retained the plesiomorphic condition of its ancestors by keeping two functional ovaries but having a monoautochronous ovideposition as C osmolka. However, unpaired eggs coupled with their spatial distribution could be interpreted as the result of having only one functional ovary, a condition that if confirmed, would be shared with extant birds but not with any other archosaurs. The interpretation of these authors whereas T . formosus eggs are paired is further addressed in the discussion section of this chapter. An eggshell fragment of MOR 246 was the basis of the micro structural analysis. The eggshell is relatively thin averaging only 965 pm and its radial section viewed in SEM displays a bi-layered structure (Fig. 6.IE). The contact between the two structural layers is gradual giving this eggshell a prismatic character (Grellet- Tinner and Norell, 2002) making difficult to distinguishing layer 1 from layer 2. 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.1 A. Egg clutch of Troodon formosus (MOR 675) prepared from below, as such the exposed sections of the eggs are their lower sections normally buried in the substrate. Note the conic and asymmetrical shape of these eggs, and the tremendous difference in their pole sizes. In contrats to oviraptorid eggs, T . formosus eggs are “planted” with their long axis at an angle at, or around 90 degrees to the substrate. B. Detail of some of (MOR 675) eggs where the analoguous sections of B. jaffei eggs are copy-pasted. Although there is difference in size between the eggs of these two troodontid species, their shape perfectly matches. C. TLM view of T . formosus eggshell (MOR 246). Note the ubiquitous presence of a surficial diagenetic layer of calcite that could be confused with a structural eggshell layer. D. (MOR 963), egg clutch in a rimed nest. E. SEM view of T . formosus eggshell. Note the presence of two prismatic eggshell structural layers which contact is unclear. Double black arrows highlight these two layers, and single black arrows point to the concave base oo the eggshell units. F. Increased number and size of vesicules (black arrows) help to discriminate layer 1 from layer 2 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.1 A n . \ 1 inm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, the boundary between these two layers can be better approximated by net increase in amount of vacuoles in layer 2 (Fig. 6.IF), a feature visible in SEM. The 191 pm thick layer 1 consists of long blade-shaped calcite crystals that are the prolongation of the basal spherulites (Figs. 6.2A, B). The spherulites form elongated, laterally compressed, and truncated barrel-shaped cones of 70.3 by 122.5 pm (Table 6.1). These structures originate from the nucleation centers at the base of each eggshell unit (Figs. 6.2A, B). Congruent with the process of calcium absorption resulting from embryogenesis the inner surface of few mammillae have been dissolved giving a concave appearance to the base of some eggshell units, a process likely amplified by diagenetic dissolution (Fig. 6.IE). Layer 2 averages 721 pm and consists of calcific crystals with their C-axis angled with respect to the bladed-shaped crystals of layer 1. Yet, this differential orientation does not completely mask the prismatic columns of layer 1 (Figs. 6.1C, E), as it would be in the case of the aprismatic eggshells of C. osmolka and D. antirrhopus. As previously mentioned layer 2 has a greater concentration of micron size vesicles than layer 1 and the ratio of this layer to layer 1 is in the order of 4:1. Only two pore canals were observed; one in a SEM image and another in a TLM (Fig. 6.2C). Both cross obliquely layers 1 and 2 and based on the TLM observations it seems that the pore canal widens midway in layer 2 to 36 pm and pinches down to 18 pm as it reaches the contact between layers 2 and 1 (Fig. 6.2C). This configuration of the pore canals was already noticed by Varricchio et al. (2002), although its width estimation exceeds this one by 20 pm, an artifact easily attributed to pore fluid weathering and calcite infilling. Although reports (Varricchio et al., 2002) suggest that T . formosus could have three structural layers, close examination of Tformosus eggshell structure indicates otherwise. SEM observation reveals a funnel-like pore opening below what could be interpreted as a third structural layer. The thickness of this putative third layer varies from 129 to 78 pm. It consists of diagenetic calcite that exhibits needle like crystal habit (Fig. 6.2D) similar to 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.1: Troodon formosus egg and eggshell parameters egg shape egg greater diameter egg smaller diameter egg diameter ratio ornamentation morphology ornamentation size pore canal morphology pore canal size pore aperture morphology internodular space pore aperture size eggshell thickness number of layers eggshell character state layer 1 morphology layer 2 morphology layer 3 morphology layer 1 size layer 2 size layer 3 size ratio total thickness/Ll ratio L1/L2 ratio L3/L2 core location core size eggshell units morphology eggshell units-thin section monoautochronic paired eggs egg spatial position in clutch air cell clutch oval/pear 13 cm 7.71 cm 59% smooth oblique 36 thinning to 18 round, funnel-like 914.8 micron 2 prismatic blade shape continuous but not at 90 degrees 191.4 microns 721.5 microns 4.77 0.26 153 microns (w) by 275 microns (h) convex growth lines to flat higher yes no vertical yes no empty space in center Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. aragonitic needles growing on micrite upon grains in marine environments (Tucker and Wright, 1990). Thus, the crystalline make up thus refutes a biological origin for this possible third layer. A different source of error could stem from the observation of the upper section of layer 2 that seems is separated from the rest of layer 2 under CL (Fig. 6.2E). However, the separation line fluoresces similarly as the diagenetic calcite layer that blankets the eggshell outer surface, the micrite that invaded a pore canal, and some former organic lines that are most adjacent to the external eggshell surface, thus suggesting that this line is a diagenetic artifact. SEM observations denote no structural difference, TLM and PLM observations denote no crystallographic or extinction patterns change at or around this demarcation line, adding more support to the CL observations and refuting the presence of three layers in this eggshell. DISCUSSION Paleoecologic and paleobiologic considerations Egg clutches of T . formosus are relatively small in contrast to those of known oviraptors or those associated with elongaloolithid eggs. The eggs stand with their long axis at 60 to 90 degrees to the horizon and fill the entire clutch surface without forming multiple superposed egg layers. Egg pairing as inferred by Varrichio et al. (1997), was based on a circular argument. A further issue against this claim of egg parity involves the geometry of the nest and the eggs. According to the authors (Varrichio et al., 1997, 1999, 2002), the nest is a bowl-shaped depression and we know that eggs are conic with their narrow ends down and their blunt ends in contact of each other filling the top of the nesting structure. Having the widest section of these eggs in close contact and filling up the space of the bowl-shaped depression implies that the position of their lower and narrower sections is random and not according to an ovideposition pattern. Therefore, any analysis based on their narrow pole spatial 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.2 A and B. SEM and TLM views of the long blade-shaped crystals at the base of each eggshell unit. C. TLM view. Note the pore canal that obliquely crosses the eggshell thickness and widens in its lower section. D. SEM of the diagenetic calcite layer that blankets that outer eggshell surface. Note the needle-like crystal habit similar to aragonitic needles growing on micritic grain in marine environments. E and F. CL and TLM views. Note the line (white arrows) that separates layer 2 in two sub-sections, possibly indicative a third structural layer. However, close examination of these two images refutes the presence of such a third layer by the lack of change of crystallographic or extinction patterns below and above the line and by the presence of the same diagenetic material that has invaded this line, the pore canal, and covered the eggshell surface. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.2 . t - v 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. arrangement is biased. J. Homer (1987) even suggested a manipulation either with the snout or the hands by the parent dinosaur after ovodeposition. This argument would further refute any conclusion inferring egg parity based on their spatial arrangement. Although caution should be warranted before delving into speculations resting on multiple levels of inferences, it is worth noting that there is a reported monoautochronic ovideposition for C. osmolka (Clark et al., 2001), observed in Chinese elongaloolithid eggs. Assuming a lack of egg parity in T . formosus egg clutch could suggest that this troodon already had a single functional ovary as modern birds do. Alternatively, the lack of egg parity could be interpreted as a character reversal whereas T . formosus would have to lay its entire clutch in a single ovideposition as observed in extant crocodilian species for instance, a less parsimonious explanation considering T . formosus phylogenetic position. Although a final decision about whether T . formosus had a single ovary may require further evidence, the asymmetrical and elongated-oval eggs of this extinct theropod resemble more the eggs of modem birds than those of other non-avian theropod families. In modem birds, the asymmetrical eggs are known to be rotated within the oviduct prior to ovideposition. Such a peristaltic movement is facilitated by having one single egg at the time in the oviduct. Reptilian and avian nests are viewed as microenvironments that optimize egg incubation and hatching and are species specific. Understanding the behaviors of extinct animals can only be extrapolated from observations of their living relatives. Among modem birds, the pluvianidsnot only (1) display a suite of interesting nesting behaviors which result in patterns similar to those observed in troodontids, but also (2) they live in environments that compare favorably with those inferred for T . formosus. The pluvianid, Pluvianus aegyptius, partially buries its eggs by scratching a depression in the sand of sandbars (Howel, 1979). Its eggs form a clutch filling up a bowl-like depression without leaving any empty space and they are positioned 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in a vertical to sub-vertical position, pointed pole down (as observed T . formosus clutches) with only the blunt pole of these eggs exposed periodically. Although the physiology of avian eggs (32-39° centigrade) is better suited for hot climatic conditions than their non-avian reptile counterparts, they still require parental care. Eggs partially buried or buried only a few centimeters below the surface in hot and arid environments could easily reach lethal levels of 40° centigrade if not attended by the parents whose primary thermoregulation strategy is to cool down their clutch with water from nearby water sources (Grant, 1982). P . aegyptius follows this incubating model by transferring heat at night and by cooling down the eggs during the day by wetting them with water transported in the plumage (Howel, 1979). This interesting incubating strategy consists of utilization of thermo-inertia from the sand as a heat source and active involvement of the parent to insure that its eggs stay below lethal levels during extreme hot conditions (Howel, 1979). The paleoclimate of the Two Medicine Formation is best described as arid with summer convective showers (Varricchio, 1993). Egg Mountain and Egg Island, the two localities where T . formosus have been recovered were islands in a local lacustrine system. The similarity of the paleoenvironments, coupled with nesting structure and clutch arrangement that resemble each other, could indicate that T . formosus and modern pluvianids could share the same nesting behaviors. This further infers that T . formosus could have incubated rather than the brooded its clutch. In addition to nesting behavior, the eggshell structure of T . formosus could support its incubating strategy. Studies on shells of mollusks have shown that there is a net increased resistance to pressure as shell accumulates more structural layers (Kamat et al., 2000). Although the bi-layered eggshell structure is not equivalent to the trilaminated structure in modern birds, the vertical arrangement of T . formosus eggs coupled with their partial burial into the substrate offers an innovative compromise whereas the weight of the “incubating” parent is transferred onto one 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pole of the egg that are at least partially buried eliminating the risk of damaging the clutch (Zhao, 1997). Hence, the mechanical properties inferred from the eggshell microstructure coupled with the egg spatial position and burial would further advocate a proto-avian incubating behavior for T . formosus. CONCLUSION T . formosus nests, clutches, eggs, and eggshell structure provide a unique window to the past by providing clues about the reproduction and nesting behaviors of this dinosaur. To date the marked asymmetric and elongated eggs of T . formosus depart from the shape of any other known non-avian dinosaurs and suggest the presence of a fully formed air cell. In addition, the vertical to sub-vertical position of the eggs within the clutch is also particular to T . formosus and B. jaffei (see Chapter 7) and strongly parallels that observed in birds such as P . aegyptius. P . aegyptius thermoregulates its clutch which is partially buried by cooling with water rather than incubating its eggs by heat transfer as most of the modern birds do. Whether T . formosus nesting behavior is totally analogous to that of P . aegyptius is still unclear but the observations at hand would favor this hypothesis. Brooding, a precursor to incubation was already suggested for the Asiatic Oviraptorid C. osmolka (Norell et al., 1995, Clark et al., 2002). T . formosus eggs, eggshell structure, clutches, and nests suggest that troodontids, and therefore dromaeosaurs went a step further toward the avian condition by perhaps developing an incubating behavior in a form of thermoregulatory mechanism for shading eggs in hot and arid environments. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VII An Egg Clutch of the Troodontid Byronosaurus jaffei from the Gobi Desert: Novel Perspectives on the Origin of the Avian Reproductive Physiology (Taphonomic level 2 and taxonomic level 1) INTRODUCTION A spectacular egg clutch of the Mongolian troodontid Byronosaurus jaffei (Figs. 7.1 A, B) was found in close proximity of two hatchling skulls (Norell at al., 1994), one of those preserved with eggshell fragments on its dentary and maxilla (Fig. 7.1C), and a partial skeleton belonging to a juvenile theropod, all of the same species (Fig. 7. ID). The four fossils were discovered in the late Cretacaous Djadokhta Formation of Ukhaa Tolgod (Gobi Desert) during 1993 and 1995 expeditions of the American Museum of Natural Flistory in association with the Institute of Geology of Mongolia. At the same time these specimens could provide new characters for phylogenetic use, they could shed light on the reproductive behaviors and indirectly on the physiology of this troodontid. Furthermore, comparison with the North American T . formosus will test whether the hypotheses set forth in the previous chapter are relevant for both troodontid species which are part of a cosmopolitan group of dinosaurs that roamed in North America and North Asia during the Late Mesozoic (Clark et al., 2002). Material IGM 100/972. Nearly complete Byronosaurus jaffei skull with eggshell on the maxilla and dentary, collected in 1993 from the exposures of the Djadokhta Formation in Ukhaa Tolgod, Xanadu sublocality. IGM 100/1003x. Byronosaurus jaffei egg clutch and a partial juvenile of the same 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. species. Collected in 1995 from the exposures of the Djadokhta Formation in Ukhaa Tolgod, Xanadu sub-loclaity at the same location than IGM 100/972 and a few meters uphill. The specimens consist of a clutch of 16 eggs in situ and three isolated eggs found in close proximity of the nest. None of the eggshell material reacts to CL, thus implying no diagenetic replacement in the crystalline structure. DESCRIPTION The preserved sections of B. jaffei eggs show no signs of fracture and very little taphonomic or diagenetic deformation besides a line above the sediment level following the circumference of eggs 5, 1, 3, 6, and 14 IGM 100/1003x. Although it is difficult at this time to speculate on the exact origin of these lines, it is worth noting that the only affected eggs are located on the periphery of the clutch. Although the clutch is probably not complete as evidenced by the presence of the three isolated eggs, it is fair to assume that the original size IGM 100/1003x does not markedly differs from the presently recorded dimensions of 33.5 cm by 35 cm. This assumption is based not only on the spatial arrangement of IGM 100/1003x eggs, their size, but also with a comparison of a 24 egg clutch of the three meter long T . formosus (MOR 963) (see Chapter 6). The entire clutch fills up the whole surface leaving no empty space in the center. Eggs are positioned with their narrow-end down within the sediment with their long axis at 45 to 90 degrees but mostly at 90 degrees to the substrate. The pole preserved in the sediment are small and rounded, and the eggs flare out to the point where they are broken, giving them a conical shape (Figs. 7.1 A, E). There is no obvious egg parity in IGM 100/1003x. Spacing among the eggs was obtained by measuring the distance between the topmost preserved circumferences of adjacent eggs. However, due to the egg flaring up, the measured spaces are consequently greater than the real distance between the uppermost section of the eggs when complete. Therefore, it is likely that the blunt poles of these eggs were in 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contact forming a continuous surface. To examine and compare the eggs of the IGM 100/1003x clutch their contour was drawn then scanned and brought to scale in the Photoshop 5.5. Five random drawings were subsequently imported as layers and stacked (Fig. 7.IE). When superposed, their contour lines fit closely implying an identical shape. To examine and compare B. jaffei eggs with those of T . formosus eggs, the narrow poles of B. jaffei from IGM 100/1003x were superimposed on the photo of Mor 675 and 676 eggs (Fig. 7. IB). It is worth noting that further comparative measurements of these narrow poles between these North American and Central Asian troodontid species show that T . formosus narrow poles are twice as large as B. jaffeVs. The two pictures were reduced to a similar resolution and B. jaffei eggs were brought up to T . formosus size. When superposed, their contour lines fit closely implying that the aspect of B. jaffei narrow egg pole sections conforms to the T . formosus general aspect (Fig. 7.IB). Considering that T . formosus and B. jaffei share the same egg geometry for their narrow pole and preserved equatorial sections, its seems plausible that the missing bigger pole of B. jaffei eggs would also conform to the known shape of those of T . formosus. One eggshell fragment from IGM 100/972 and two from IGM 100/1003x were removed to examine the eggshell microstructure. As indicated by its curvature, it is likely that eggshell fragment aposed on the dentary of IGM 100/972 is a fragment of the hatching window glued on the snout of this B. jaffei embryo during hatching. The eggshell is relatively thin, averaging only 428.05 pm and its radial section viewed in SEM displays a bi-layered structure (Fig. 7. IF). The contact between the two structural layers is fuzzy giving this eggshell a prismatic character (Grellet-Tinner and Norell, 2002). As for T . formosus there is a net increase in amount of vesicles in layer 2 (Fig. 7. IF) that helps to distinguish layer 1 from layer 2 in IGM 100/972 (see Chapter 6). The entire 158.25 pm thick layer 1 (Table 7.1) consists of long blade-shaped calcite crystals (Fig. IF) that surround nucleation centers at the base of 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 7.1: Byronosaurus jaffei egg and eggshell parameters egg sbape egg greater ammeter oval/pear egg smaller diameter egg diameter ratio 59% ornamentation morphology smooth ornamentation size na pore canal morphology ? pore canal size ? pore aperture morphology round and funnel shape internodular space na pore aperture size ? eggshell thickness 466.7 microns number of layers 2 eggshell character state prismatic layer 1 morphology blade shape layer 2 morphology continuous but not at 90 degrees layer 3 morphology layer 1 size 159.9 microns layer 2 size 305.5 microns layer 3 size ratio total thickness/Ll 2.91 ratio L1/L2 0.52 ratio L3/L2 core location core size eggshell units morphology 40° flare up to 30% o f height to nearly eggshell units-thin section monoautochronic yes paired eggs no egg spatial position in clutch vertical air cell yes clutch no empty space in center Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.1 A. IGM 100/1003x, an egg clutch (prepared from the bottom) of the troodontid B. jaffei. The clutch was discovered in the Djadokta Fm. in Ukhaa Tolgod, Mongolia. B. Same egg clutch but prepared from the top. Regardless of the way this clutch was prepared, there is no obvious parity of the eggs as speculated by Varricchio et al. (1997, 1999) for T . formosus. The lack of parity, combined with the egg spatial arrangement, and the egg shape suggest that B. jaffei might already have one single (instead of two) functional ovary as modem birds have. C. IGM 100/972, juvenile B. jaffei skull with eggshell on its maxilla and dentary. This skull was found a few meters away from IGM 100/1003x during another field season. D. Cast of IGM 100/1003x with the post-cranial remains of a B. jaffei juvenile found above the clutch. The clutch had 16 in situ eggs and 3 eggs were recovered at the same time nearby. E. To evaluate the homogeneity of the 19 eggs, their contours were drawn and, randomly, five of those were stacked together. As observed by the superposed lines, their shapes are similar. F. As that of T . formosus, the eggshell structure of B. jaffei is bi-laminated (black double arrows) and has a prismatic contact between the two layers. Note the concave base of the eggshell units that attests for the calcium resorbtion during the development of the embryo. 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.1 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. each eggshell unit. Some nucleation centers have been dissolved giving a concave appearance to the base of a few eggshell units, a process congruent with calcium resorbtion attributed to late stage of embryonic development and congruent with the “hatching window” hypothesis. Layer 2 averages 261.81 pm and consists of calcitic crystals with C-axis that are orientated at a different angle than those of the bladed- shaped crystals of layer 1. Although there are no readily visible pores, a bias likely due to the small size of the studied sample, an aperture on the eggshell surface may be a putative pore aperture. This aperture is aligned with discontinuous openings in the radial section of IGM 100/972 and the trend of the pore canal appears slightly oblique. An eggshell sample was removed from IGM 100/1003x egg #1 to objectively compare both specimens and assess their co-specific relationship. SEM examination confirms what is observed in IGM 100/972. As for IGM 100/972,the internal structure of IGM 100/1003x egg #1 is also bi-layered and prismatic (Figs. 7.2A, B). Layer 1 averages 159.94 pm a value comparable as that of IGM 100/972. The total thickness of this sample (466.71 pm) seems to slightly exceed the previous observation by 40 pm. This incongruence could easily be attributed to either a difference of eggshell thickness between the pole where the air cell is located and the rest of the egg, or to a diagenetic artifact because IGM 100/972 outersurface was glued to the dentary of the embryo during fossilization. Layer 2 measures 305.50 pm and reflects the same thickness variation as that of the total eggshell thickness. As for IGM 100/972, layer 2 of IGM 100/1003x is distinctly different from layer 1 not only by the change of crystallographic orientation but also by the presence of a greater amount of vacuoles (Figs. 1.2 A, B), some of them exceeding a micron in size. No pores are visible in this specimen and as in IGM 100/972 many eggshell unit nucleation centers are concave (Figs. 7.2C, D), congruent with the interpretation of calcium resorption during the late stage of embryogenesis, thus supporting at least the presence of embryos at an advanced ontogenetic stage in these eggs. I ll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.2 A. SEM view of an eggshell fragment from egg # 1 in the IGM 100/1003x clutch. The observed structure favorably compares with that of IGM 100/972. B. BSEM view of the same eggshell fragment reveals in greater detail the bi-laminated eggshell structure (double black arrows) and the presence of more numerous and bigger size vesicles in layer 2. C and D. SEM and BSEM views. As the base of the eggshell units of IGM 100/972, those of IGM 100/1003x display the same concavity, indicative of calcium resorbtion, a pattern congruent with the hatching of the juveniles. E. SEM view of an excessive acidification of the internal eggshell section due to pore fluids from surrounding siliceous sediments. This diagenetic process helps to deliminate the eggshell units. F. Eggs of an extant pluvianid that are spatially buried within a rimed nest in similar sediments and environments as those of the Cretaceous T . formosus and B. jaffei. This pluvianid utilizes the thermoemergy from the sediment to incubate its eggs and thermoregulates excessive temperature (over 40 degrees) by wetting its clutch with water transported in its plumage. All the oological and reproductive analogies suggest that troodontids may have already adopted a proto-incubation similar to the modem pluvianid incubation behavior. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.2 W m m ig m 15.0 KV m um 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The third eggshell sample from one of the three isolated eggs confirms what has already been observed under SEM in the previous samples (Fig. 7.2E). This third sample exhibits a greater amount of dissolution likely due to pore fluids originating from the surrounding siliceous sediments (Faccio, 1994). DISCUSSION The discovery of brooding oviraptors (Chapter 5) changed the historical interpretation of the behavior of oviraptorid theropods from egg robbers to avian-like brooders (Clark at ah, 2001). The study of the oological material of B. Jaffei allows a step further in the understanding of theropod nesting behavior and physiology. In contrast to C. osmolka, B. jaffei laid relatively small clutches with a single layer of eggs that stand with their long axis at 60-90 degrees to the substrate and fill the entire clutch surface. Evidence of egg parity is lacking in IGM 100/10003, which could confirm this hypothesis about T . formosus (see previous chapter). Considering the already reported monoautochronic ovideposition of T . formosus (Varrichio et al., 1997. 1999), C. osmolka (Clark et al., 2001), the lack of egg parity in B. jaffei is interpreted here as evidence of atrophy of one ovary, a condition previously to be exclusively avian. Alternatively, the lack of egg parity in IGM 100/1003 could be interpreted as a character reversal whereas B. jaffei would have to lay its entire clutch in a single ovideposition as observed in extant crocodilians and titanosaurids (Chapters 2 and 3). However, this alternate interpretation would not account for the regular and constant spatial arrangement of the eggs in the nest, or for the character reversal of the monoautochronic ovideposition in T. Formosus (Varricchio et al., 1997,1999, 2002), oviraptorids (Clark et al., 2002), and unassigned elongatoolithid eggs of various size (Carpenter, 1999). Another line of evidence supporting the single ovary hypothesis comes from the fact that unlike basal reptiles, birds lay asymmetrical eggs with a well- 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. developed air cell located in the largest pole. This elongated-oval pyriform egg shape is characteristic of avians where eggs are known to be rotated within the oviduct prior to ovideposition, a peristaltic movement that is more easily accomplished by having one single functioning ovary. Although a slight asymmetry has already been reported in oviraptorid eggs to accommodate a proto-air cell (Chapter 5), B. jaffei and T . formosus eggs depart from the elongatoolithid long shape eggs by having a more asymmetrical and compact avian-like aspect. In turn, these more avian-like eggs probably reflect a smaller and more compact abdominal cavity (see Carpenter, 1999). Thus, the proposed ovary reduction in troodontids can be viewed as a pre-adaptation for flight that evolved in concert with highly pneumatic skeletons and feathered bodies prior to the rise of birds. Although the eggshell micro structure could only attest that embryos at a late ontogenetic stage developed in the eggs, the incompleteness of the eggs in the IGM 100/1003 clutch and the three isolated eggs recovered in its proximity are likely the result of successful hatching. This conclusion is not only supported by the presence of a hatchling skull with eggshell fragments on its dentary and maxilla, and juvenile skeleton recovered nearby, but also because of other oological evidence. The presence of few eggshell fragments found randomly positioned in the sandstone infilling the eggs 3, 6, 13, and 15, the preservation of the lower section of each egg without visible crushed and cracked eggshell on the egg surface coupled with the presence of sandstone infilling advocate that these eggs hatched and were filled with the surrounding matrix prior to fossilization. As reported in the descriptive section, the presence of the intact pointed pole coupled with the flaring up portion of these eggs undoubtedly implies that the missing section is the blunted pole in IGM 100/1003x. No pores are visible under SEM or thin section microscopy supporting the interpretation that the blunt poles, where most of the pores are concentrated (Varricchio et a., 1997), are missing. Hatching strategy in modern seagulls shows that 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the polar region where the chick pips in the air cell is generally destroyed as well as the region immediately adjacent to the air cell, and only the lower section of the eggs stays intact (Hayward et al., 2000). As for T . formosus, (Chapter 6) interpretations related to avian physiology are similarly set forth for B. Jaffei. Nesting environments of pluvianids (e.g. P . aegyptius) (Fig. 7.2F) are comparable to the desert settings of the estimated 30 degrees north latitudinal paleobearing of the Djadokhta Formation. Although avian eggs, in general, by their physiology are well suited for hot climatic conditions, they still require parental care. P . aegyptius eggs partially buried or buried only a few centimeters below the sand surface in hot and arid environments could reach lethal levels of 40° centigrade if not attended by the parents whose primary incubating strategy is to cool down their clutch (Grant, 1982). In addition to this extrinsic heat, metabolic heat is produced by the embryos during embryogenesis increasing the existing heat, thus further reducing the need of the parent to incubate the clutch in the conventional sense (heat transfer), but probably requiring more parental care to cool down the eggs at heat pick times so the eggs would stay within an acceptable 32-39° centigrade temperature range. This type of incubation results in a more active participation of the parents in cooling down rather than transferring body heat to eggs, could be viewed as a phylogenetic precursor to the incubation of modem birds. If correct, this hypothesis would advocate that the B. jaffei and T. formosus developed a parental care that exceeds the brooding behavior of their contemporaneous oviraptorid theropods but not yet fully resembles that of most neomithines. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONCLUSION The Central Asiatic specimens IGM 100/1003x and IGM 100/972 provide clues about the reproduction and nesting behaviors of B. jaffei that confirm those inferred in the previous chapter for the North American T. formosus. To date the marked asymmetric and conic shape of B. jaffei and T . formosus eggs suggests a well- developed, avian-like air cell. This egg shape coupled with the spatial distribution of the eggs in the clutch and nest suggest that both troodons had a single functioning ovary. In addition, the typical vertical to sub-vertical position of the troodontid eggs resembles what is observed in P . aegyptius, a modem bird that lives in hot climates. P . aegyptius incubating strategy is characterized by a greater investment to cool down rather than transferring heat to its egg clutch. This strategy could be regarded as a precursor to the incubation of most neomithines, thus implying that avian incubation would predate that advent of avialans 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER VIII “Baby Louie”, a putative oviraptorid dinosaur embryo from China (taphonomic level 2) and comparable non-associated oospecies (taphonomic level 4) INTRODUCTION While conducting a geological exploration in the Xixia County (Fang et al., 1994) in 1974, the 12th Geological Team of the Henan Geological Bureau discovered several spherical objects with an ornamented surface that were soon after identified as dinosaur eggs. Since then, numerous fossil dinosaur eggs and eggshells have been discovered in these Upper Cretaceous Chinese red beds (Fang et al., 1994). These eggs come in different shapes and preservation states, but none of them were originally reported in close association with embryonic remains. Consequently, Chinese researchers (Zhao, 1975) adopted a binominal nomenclature to classify (parataxonomic classification) this egg and eggshell cornucopia. Among these are the elongaloolithid group with the Elongaloolithus, Macroolithus, genera and the Macroelongatoolithus genus (from the from Xixia County) which measures approximate 45 cm by 16 cm. Many of these eggs have come to light as a consequence of worldwide illegal trading practices. A clutch of at least four eggs so obtained with an associated early juvenile atop provides important information for the taxonomic identification of these Macroelongatoolithid egg. This clutch, known as “Baby Louie”, is currently at the Indianapolis Children Museum (a donation by C. Magovem of the Stone Company) and a cast and original eggshell fragments are housed at the LACM (Fig. 8.1 A). Preliminary observations of the juvenile favor its identification as a big oviraptorid (pers.com. Norell and Clark), or a therizinosaur (Fig. 8.IB) (National Geographic, 1996). An unidentified and sizeable clutch from Sanlimiao (Xixia County) of approximatively 75 inches in diameter comprising at least 13 pair of eggs was also featured in the 1996 issue of the National Geographic 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Fig. 8.1C). Cursory inspection of this egg clutch in the Nanyang Museum, Henan Province (Fig. 8. ID) shows that its eggs are of the same dimension and shape as “Baby Louie” and other macroelongatoolithid eggs. In view of these interesting specimens, a description of macroelongatoolithid eggshell structure, egg shape and size, and egg distribution is proposed in this chapter to better understand the taxonomic identification of these eggs and evolution of oological characters at large. This study could also examine the paleobiologic and phylogenetic implications that stem from the comparison of these macroelongatoolithid specimens with other elongatoolithid eggs, namely oviraptorids. Material This material (“Baby Louie”) is perfectly associated (taphonomic association level 2) but not yet well identified. LACM 7477/149736a. “Baby Louie”; A cast of 4 macroelongaloolithid eggs with a flattened embryo atop an egg pair recovered in the Zoumagang FM (Maastrichtian), 10 km NE of Sanlimiao, Xixia County, Henan Province, China. Egghells were set aside during the preparation of the original specimen. LACM 7477/149736b. Eggshell fragments saved during preparation of the specimen. Macroelogatoolithus xixia (Nat. Geo. 1996) a clutch of at least 13 pairs of eggs (taphonomic association level 4) laid in a large ring of 75 inches in diameter from the Zoumagang FM (Maastrichtian). No eggshell directly from that specimen but eggshell fragments from GMY material from the same locality now housed at the Nanyang Museum. F.A.3818.2001-1. LACM accession number for an oblong dinosaur egg (taphonomic association level 4) from the Zoumagang FM (Maastrichtian), 10 km NE of Sanlimiao, Xixia County, Henan Province, China. Eggshell fragments from near the oblong pole. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.1 A. LACM 7477/149736a. “Baby Louie”; A cast of 4 macroelongaloolithid eggs with a juvenile atop a pair of eggs recovered in the Zoumagang FM (Maastrichtian), 10 km NE of Sanlimiao, Xixia County, Henan Province, China. Egghells were separated during the preparation of the original specimen. B. A pair of identical eggs interpreted by P. Currie in the 1996 National Geographic Issue as those of a therizinosaur. C. Macroelogatoolithus xixia, a clutch from the Zoumagang FM (Maastrichtian), of at least 13 pairs of eggs that were laid in a large ring of 75 inches in diameter (as portayed by the National Geographic 1996 Issue). D. Same egg clutch re-visited by the author. E and F. SEM and PLM views of LACM 7477/149736b, an eggshell fragment of Bay Louie. Note the low amplitude sinusoidal aspect of the eggshell outer surface (white arrows), the presence of a round pore aperture (black arrow), and the presence of two eggshell structural layers with aprismatic contact. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.1 500[im SOX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DESCRIPTION “Baby Louie” eggs and eggshell structure Although the eggs are flattened and only partially surfaced from the matrix, they exhibit an unquestioned parity. Two of them are well enough preserved to estimate their length at 40 to 41 cm (Fig. 8.1 A). The eggshell surficial ornamentation varies from low ridges aligned with the long axis near the equatorial regions to a more pronounced reticulated ornamentation near the poles. However, the eggshell is nearly smooth at the poles. Aside from these few observations, very little more can be said about the shape of these eggs. The juvenile lays atop two eggs and is nearly complete. Based on preliminary observations, the shape of the dentary, in particular its posterio-antero thickness, indicates an oviraptorid affinity but taxonomic identity of this specimen must await further descriptions. The studied eggshell fragments range from 1198 pm to 1581 |xm in thickness with an average of 1406.27 pm. These measurements exclude the surficial ornamentation and are based on thin section and SEM observations (Table 8.1). In radial section the ornamentation appears as a low amplitude sinusoidal line and is composed of a parallel network of undulating low rounded ridges that show signs in SEM or TLM of chemical weathering due to preparation of the specimen or related to diagenetic (Figs. .8 IE, F). There is a distinct possibility that this weathering affects the surficial ornamentation and thereby biases the present observations pertinent to this subject. CL examinations confirm a degree of alteration in the eggshell from diagenetic origin. Luminescence follows a vertical orientation paralleling the eggshell units and also reveals that the eggshell is blanketed by a thin calcite layer that invades the troughs of the surficial ornamentation and conceals its true nature (Figs. 8.2A, B). As observed in CL, this ornamentation consists of nodes of equal height with a slight undercut at their base, giving them a mushroom-like aspect, and qualifying this type99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 8.1: Baby Louie and F.A.3818.2001-1 combined egg and eggshell parameters egg shape egg greater diameter egg smaller diameter egg diameter ratio ornamentation morphology ornamentation size pore canal morphology pore canal size pore aperture morphology internodular space pore aperture size eggshell thickness number of layers eggshell character state layer 1 morphology layer 2 morphology layer 3 morphology layer 1 size layer 2 size layer 3 size ratio total thickness/Ll ratio L1/L2 ratio L3/L2 core location core size eggshell units morphology eggshell units-thin section monoautochronic paired eggs egg spatial position in clutch air cell clutch elongated 35 cm 14 cm 40% tubercles to linearituberculate and nodular 68.8 microns round funnel shaped 298 microns 1406.2 microns 2 aprismatic intermediate long bladed continuous but not at 90 degrees NA 376.9 microns 1026.5 microns NA 3.73 0.36 long bladed growth lines nto as compacted close to the external surface yes yes in a circle small empty space in center Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of ornamentation as dispersituberculate (nodes that are dispersed over the eggshell surface) (Figs. 8.2C, D). Only one pore aperture was visible in SEM and its shaped as a round depression with an outside diameter of 298 pm (Fig. 8. IE). However, this pore aperture is not aligned with a visible pore canal. Only CL reveals the presence of pore canals perpendicular to the outer and inner eggshell surfaces connecting those in a bee-line path. Although radial SEM section, TLM, and PLM support the presence of two structural layers (Figs. 8.IE, 8.2E) with aprismatic condition by Grellet-Tinner and Norell (2001), the demarcation between these two layers is not as obvious as expected for an aprismatic condition, as observed in C. osmolka or D. antirrhopus, likely due diagenic alteration and other intrinsic factors. Layer 1 is 376.35 pm thick and consists of calcitic crystals in prolongation of the basal spherulites (Fig. 8.2E). The crystals in layer 1 of “Baby Louie” eggshell are long and slender and do not fall neatly in the arbitrary division between acicular and bladed-shape. This division reflects the two extremes of a spectrum where the aspect ratio between length and width is either very small or very large. It is generally agreed that needle like or acicular crystals have a high length to width high ratio, and blade-shaped is a small ratio. The aspect ratio of “Baby Louie” crystals are still consistent with the bladed-shape category, but their long and slender shape would place them at the border of this category. The spherulites at the base of layer 1 form the lower section of the shell units. They are composed of calcite crystals that are a miniaturized version of the larger ones observed in the same layer and are V-shaped with branches forming a 60 degree angle from the horizontal on either side (Fig. 8.2F). The shell units disappear at the boundary of layers 1 and 2. Layer 2 differs from layer 1 by its tabular crystallography where the crystals are stacked with their contact surfaces generating lines similar to those observed in the foliation associated with the metamorphic rock slate (Figs. 8.IE, 8.2E). These lines are at an angle of 45 degrees from the vertical axis of the bladed- 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.2 A and B. Same eggshell LACM 7477/149736b in TLM and CL. CL reveals the true aspect of the surficial eggshell ornamentation by exposing the diagenetic layer that surrounds the base of the nodes (white arrows). Also pore canals connecting both eggshell surfaces in a bee-line pattern are seen in CL while invisible in TLM, PLM, or SEM. C and D. Greater detail of the as previously observed features. Note the undercut at the base of the node giving them a mushroom-like aspect that has already been reported for macroelongatoolithid eggs. E. Although aprismatic by nature the demarcation between layers 1 and 2 is not as well defined as in the eggs of C. osmolka or D. antirrhopus. Note the extremely elongated blade-shaped crystals in layer 1 (black arrows) and the abundance of vesicles in layer 2. F. SEM view of the base of a “V”-shaped eggshell unit where also the primary spherulites are composed of long and narrow blade-shaped crystals (black arrows). 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.2 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. shape crystals of layer 1. In addition to its different crystallographic nature, layer 2 has a higher vesicle concentration than layer 1. These vesicles, when seen in SEM, follow the contact surface of the tabular crystals and, in TLM, organic lines that are regularly distributed throughout layer 2 forming multiple side by side rounded inverted V-shaped. The thickness of layer 2 averages 1026.64 (am but this dimension increases by 266 (im if the surficial ornamentation is included. F.A.3818.2001-1 egg and eggshell structure This egg was recovered from the Zoumagang FM (Maastrichtian), 10 km NE of Sanlimiao, Xixia County, Flenan Province, China, the same location and horizon as LACM 7477/149736a. Although only its size and the eggshell morphology can be compared with LACM 7477/149736a, F.A.3818.2001-1 could well belong to the same oospecies. If its eggshell characters match those observed in LACM 7477/149736b, F.A.3818.2001-1 could contribute characters that are presently missing in “Baby Louie”. F.A.3818.2001-1 is oval shape and lays on its long axis half embedded in the original reddish matrix (Fig. 8.3A). Several interpretations could be immediately drawn based on the fact that the gray-green eggshell is fractured and at places overlap, and also based on the presence of a thin layer of white calcite that blankets the exposed surface (Fig. 8.3B). The eggshell is fractured into several medium and small size fragments that have been dislocated from their original position with a minor amount of overlap (Figs. 8.3A, B). The pattern of fractures suggests that the egg was submitted to extrinsic forces after been buried and after the matrix had already solidified thus limiting any possible lateral extension. It also advocates that LACM 7477/149736 was complete and possibly not fractured at burial. The white calcite blanket denotes a second diagenetic phase took place not only after the compression phase but also after the egg was buried and the silty matrix transformed into siltstone. This second 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. phase was likely associated with hypersaturated pore fluids in respect to CaCo3 and its precipitation. The direct consequences of these processes are that the eggshell ornamentation have been eroded and/or is partially masked by diagenetic CaCo3 as observed in CL (Figs. 8.3C, D) and the egg dimension might be biased. As such, the egg measures 14 cm by 35 cm, and its length to width ratio is 2.5. None of these dimensions match the oospecies included in the elongatoolithid family as described by Li et al. (1995), Fang et al. (1994) and Carpenter (1999). According to Fang et al. (1994), Macroelongatoolithus xixiaensis is known to reach 45 cm in length and 16.5 cm in width. A range of 38 to 60 cm in length for this oospecies is also mentioned and concurs with dimension given by other authors (Li et al., 1995; Carpenter, 1999). Among the other large oospecies generally included in the elongatoolithid family (e.g. Elongatoolithus, Macroolithus), none are as long as F.A.3818.2001-1 and are within the lower 20 cm range. Without further information, two interpretations can be put forward: F.A.3818.2001-1 is a new oospecies or the range of the length of M. xixiaensis needs to be extended. Despite not been a criteria by itself but just an indicator, the horizon and localities where M. xixiaensis are recovered are congruent with those of F.A.3818.2001-1 (Li et al, 1995). Another character that this egg displays, is a faint morphological difference between its two poles. In order to qualify this difference in shape, the perimeter of F.A.3818.2001-1 was outlined, cut into two exact halves, and one half was rotated at 180 degrees then overlaid over the other. The flaring angle at which the egg expends from the blunt pole toward its equator is wider than the other. When dashed lines are added to connect the eggshell fragments that are dislocated and/or overlapping at the blunt end, the asymmetry is even more apparent (Fig. 8.3E). This observation concurs with Fang et al.’s (1997) description of M. xixiaensis “one end obtuse and the other taper” and also supports that of Carpenter’s (1999) elongatoolithid description “One end is usually rounded, while the other is somewhat pointed”. Finally, where the eggshell is pristine or with an extremely 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.3 A. F.A.3818.2001-1, an oblong egg from the Zoumagang FM (Maastrichtian), Xixia County, Henan Province. Although middly altered as substanciated by the presence of overlapping eggshell fragments, this egg is well enough preserved to obtain its 14 by 35 cm dimensions. B. Detail of the overlapping eggshell fragments indicating that F.A.3818.2001-1 was buried whole and subsequently submitted to taphonomic forces that compressed and cracked its surface. The white squarre shows the place where an eggshell fragment was taken for microstructural examination. Although dispersituberculate on the rest of the specimen, the surficial ornamentation is smooth at the poles. Cand D. TLM and CL combined reveal that, as LACM 7477/149736b, the outer surface of F.A.3818.2001-1 is blanketed by a layer of diagenetic calcite (white arrows). Note the “Y”-shaped pore canal that bends midway between the eggshell surfaces. E. As C. osmolka, F.A.3818.2001-1 poles display a faint asymmetry. The F.A.3818.2001-1 contour was drawn, then bisected and the two halves were overlapped to evaluate their degree of asymmetry. F. As LACM 7477/149736b, the eggshell structure of F.A.3818.2001-1 is bi laminated with an aprismatic contact between the two layers as observed on this SEM image. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.3 m * ilk ^ ? 1 * - < * t t m H e ..... . . . „ . . . . . . . . 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. thin coating, it possible to notice a variation of the ornamentation from reticulated and randomly orientated, low vermiculate ridges at the blunt pole, to ridges that are oriented parallel to the long axis toward the equatorial regions, finishing as low and somewhat connected nodes at the pointed pole. These observations are congruent with Fang et al. (1994) description of M. xixiaensis. Eggshell fragments were sampled from the blunt pole area. The eggshell is relatively thick averaging 1373 pm without accounting for the external ornamentation and compares favorably with LACM 7477/149736b. F.A.3818.2001-1 radial section viewed in SEM displays a bi-layered structure (Fig. 8.3F), where layers 1 and 2 respectively average 307 pm and 1066 pm. The contact between the two structural layers is similar to that of LACM 7477/149736b. Likewise, layer 2 of F.A.3818.2001- 1 also displays an increase in number of vesicles comparatively to layer 1 (Fig. 8.4 A) and shows a change of crystallographic orientation (Figs. 3F, 4A). The bases of the shell units in layer 1 consist of long and thin spherulitic crystals that have nearly an acicular aspect (Figs. 8.4B, C). Furthermore, these crystals radiate from the cores forming uniform and semi-circular fans resembling the tail feathers of a turkey as to those of LACM 7477/149736b. The rest of layer 1 is composed of elongate and narrow calcific crystals which shape and spatial arrangement match those of LACM 7477/149736b (Fig. 8.4D). Although everywhere in layer 2, the organic lines are less compact as they approach the eggshell outer surface (Fig. 8.4E). The surficial ornamentation is dispersituberculate and display mushroom-like nodes with slight indentations at their bases, a feature quite noticeable in TLM (Fig. 8.4E) and CL (Fig. 8.4F), again resembling that of LACM 7477/149736b. The outer and inner surfaces are connected with straight pore canals that sometime branch in a shallow “Y” near the outer eggshell surface. Thus, the microstructure of this eggshell also supports the interpretation of this egg as co-specific to those of Baby Louie and of M. xixiaensis. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.4 A. This SEM image shows a greater amount of vesicules (black arrows) in layer 2 of F.A.3818.2001-1. B. In SEM, the shape of the spherulite crystals (black arrows) that forms the base of the eggshell units of F.A.3818.2001-1 favorably compares with that of Baby Louie. C. The blade-shaped spherulites of F.A.3818.2001-1 as those of Baby Louie nearly approach an acicular aspect due to their length by width high ratio, as observed in this TLM. D. SEM of long blade-shaped crystals that are in prologation of the primary spherulites. E. As observed in D. antirrhopus layer 2, the organic lines of F.A.3818.2001-1 divide this layer in two sub-layers. However, the upper section of layer two, in contrast to D. antirrhopus, is thinner than the lower sub-section. F. Cl examination reveals diagenetic processes have similarly affected both F.A.3818.2001-1 and LACM 7477/149736b and that these two eggs have identical eggshell structures. 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 8.4 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION The first interpretation based on eggshell microstructural examinations of F.A.3818.2001-1 and LACM 7477/149736b is that they both belong to the same dinosaur species. As such the characters of F.A.3818.2001-1 that are not presently observable in “Baby Louie” can now be inferred for this specimen and M. xixiaensis. Although the taxonomic identification of LACM 7477/149736a must await the detailed study of the early juvenile skeletal features, the interpretation of these eggs as part of the same oospecies provides oological information that in the future would be able to be linked to a particular dinosaur. Within that context, the two M. xixiaensis fossil assemblages are particularly important for phylogenetic and physiologic considerations: M. xixiaensis (National Geographic, 1996) a clutch of at least 13 pairs of eggs from the Zoumagang FM (Maastrichtian) and a partial clutch of 6 paired eggs described by Youxing et al. (1995). Youxing et al. (1995) reported six M. xixiaensis eggs which dimensions range between 39 and 51 cm in length and 13 to 17.9 cm in width. These authors noted that the eggs were distinctively paired and almost parallel to each other or forming a 13 degree angle among themselves, a feature they attributed to the muscular contraction of the dinosaur fallopian tubes during (monoautochronic) ovideposition. A similar egg pairing pattern is observed in the other clutch containing 13 pairs of eggs featured in the 1996 National Geographic Edition and now housed at the Nanyang Museum. The eggs of this clutch average 39-40 cm in length, a measure also congruent with what has been already reported and observed in other M. xixiaensis specimens. In light of our knowledge of bird physiology, egg diameter is related to the age of the dinosaur and its pelvic opening (Carpenter, 1999). In addition, younger birds would have a tendency to lay smaller eggs than mature females of the same species (pers. obs.) but overall, egg size is related to the size of the parent and whether the offsprings are precocial or altricial. Precocial chicks require more energy to 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. develop than altricial dinosaurs, thus more albumen in the egg. Hence a bigger egg is required to support precociality and/or a large bird species. The only means to increase the volume of an egg is by elongating its shape when considering the size limitation exerted on the diameter of an egg by the diameter of the pelvic opening. This elongation is a common feature to all elongatoolithid eggs suggesting that M. xixiaensis hatchlings were likely precotial and laid by sizeable dinosaur species lineage. Comparison with the oviraptorid C. osmolka (Clark et al., 1999), shows that M. xixiaensis and C. osmolka share the following features: two aprismatic structural eggshell layers, surficial eggshell ornamentation, egg pairing, egg shape, a proto-air cell denoted by a slight asymmetry between the two poles, the absence of eggs in the center of the clutch, and the narrow poles of the eggs pointing down toward the center of the clutch. However some features differentiate the clutches of these two species: the absence of multiple superimposed egg rows in M. xixiaensis as observed in C. osmolka and its allies (Clark et al., 1999; Dong and Currie, 1996), the size difference in the eggs and empty space in the center of the clutch. Although there are differences that could easily be attributed to the bigger size of the dinosaur species that laid M. xixiaensis egg, all the shared characters between these two lineages, suggest the dinosaurs that laid M. xixiaensis eggs were theropods and that they brood their eggs like to oviraptorids sitting in the center of their clutch. 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONCLUSION Eggshell microstructural observations support the identification of the Baby Louie specimen as M. xixiaensis. By the association of a neonate skeleton with oological material, Baby Louie provides another link between elongatoolithid eggs and a theropod species. The observation of M. xixiaensis clutch and its comparison the oviraptorid C. osmolka advocate that, although likely larger, the theropod species that laid M. xixiaensis eggs are closely related to Oviraptosauria and likewise shared similar egg laying and incubating strategies. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER IX Theropod eggs from Phu Phok (Thailand) (Taphonomic level 1, low taxonomic level) INTRODUCTION Recent discoveries from the Early Cretaceous of China reveal that, besides the acquisition of feathers, body size reduction was paramount in the transition from non- avian theropods to birds (Xu et al., 2000, 2003; Zhang et al., 2002) and that in some instances the primitive condition for deinonychosaurs might be small. Hypothetically, this skeletal reduction should also have been mirrored at the reproductive level by egg size reduction. However, no positively identified theropod egg displaying such a size reduction had yet been reported until four very small eggs about the size of a goldfinch’s (18 mm by 11 mm) egg were discovered in the Lower Cretaceous Sao Khua Formation of Thailand. One of them still contains a theropod embryo in ovo (taphonomic association level 1) although the fragmentary nature of this embryonic material prevents its identification beyond Theropoda. Material SKI-1: Four fossil eggs were found in 2002 and 2003 in the Lower Cretaceous red siltstones Sao Khua Formation (Raecy et al., 1996) at Phu Phok, Sakhon Nakhon Province, northeastern Thailand. As they were surface-collected, together with abundant bones and teeth of fishes and tetrapods (Buffetaut et al., 2003) eroding out of the sediment, no evidence of nesting structure was discovered. Yet, as they all come from the same small area and are morphologically identical, they are likely from the same clutch. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DESCRIPTION The four fossil eggs have undergone crushing, as substantiated by the eggshell cross section examination, but two of them are however sufficiently preserved to assess their original size and shape. The asymmetric oval shape best compares with that of Gallus gallus and its dimensions of 18 mm (length) by 11 mm (equatorial diameter) with an estimated (Sabath, 1991) volume of 1.15 cm3 match that of Carduelis carduelis (Goldfinch) or Parus major (Great Tit). The presence of multiple layers of eggshell (Fig. 1 A) indicates that the eggs were not infilled, and thus neither cracked, or opened, prior to and during burial (Mueller-Towe et al., 2002). The overlap of the eggshell fragments near the equatorial portion suggests the lateral expansion of the shell was limited as the egg was crushed after early lithification of the encasing sediment (Mueller-Towe et al., 2002). Crushing was followed by weathering (pore fluid) and calcite re-deposition, as determined in light microscopy by a faint black line outlining the eggshell outersurface (Figs. 9.1 A, D, F). Closer examination reveals this line consists of calcitic epitaxial growth possibly of microbial origin (Tucker and Wright, 1990). CL observations show the multiple diagenetic phases to which this material has been submitted by not only displaying a total luminescence but also by the variation of its intensity in the eggshell structure. The shell thickness excluding surficial ornamentation averages 353.9 pm. This value approaches that of Gallus gallus and it is 30% thicker than a possible enantiomithine from the Late Creataceous of Argentina (Schweitzer et al., 2002). The eggshell outer surface is deeply ornamented with nodes that display a bimodial distribution (Fig. 9.IB). The taller nodes average 183 pm, while the more numerous and smaller ones are half that size (92.3 pm). Funnel-like structures at the level of the taller nodes may be interpreted as pore canals, but they are more likely fissures resulting from compression during a third diagenetic phase (Fig. 9.1 A). Despite intense recrystallization, three prismatic structural layers are still visible in 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radial section (Figs. 9.1C-F), a character presently known to be synapomorphic of “enantiomithines” (Schweitzer et al., 2002) and more modem birds (Grellet-Tinner, 2000). Layer 1 is poorly preserved and only a few truncated mammillary crystals are visible at its base (Fig. 9.1C). The width and cross-section of some of these crystals show that they are not acicular but blade-like (Table 9.1). Layer 1 averages 207 pm and is 1.7 times larger than layer 2 (Figs. 9.1C, D, F), a character shared with the Neuquen eggs (per. obs.), but not observed in modern birds. Layer 3 is conspicuous and is 32 pm thick (Figs. 9.1C-F). Transitions between the layers are gradual rather than sharp representing a prismatic condition (Grellet-Tinner and Norell, 2002) as present in troodontids (Varricchio et al., 2002; Per. obs.) and modem neognath birds (Grellet-Tinner, 2000; Grellet-Tinner and Chiappe, 2004). Inside one of the eggs from Phu Phok (SKI-1), three small bones can be seen in cross-section in the hard calcitic matrix which fills the crushed shell. One of the bones just shows the outline of a hollow subcircular shaft. The other two appear to be associated in what may be roughly their original anatomical position. The larger bone shows a hollow oval shaft continued on one side by a long curved tapering process. In cross-section the greatest dimension of this bone is 0.37 mm (at the level of the process). The associated bone, placed laterally, is subcrescentic in cross-section. The combination of a hollow, relatively thin-walled bones advocate that these embryonic bones are those of a small theropod, however, without real evidence to ascertain its avian affinity. DISCUSSION When compared with the eggs of other well-diagnosed theropods, the eggshell of the Phu Phok eggs is similar in its prismatic structure to T . formosus eggshell. However, T . formosus differs in its thick bi-layered eggshell (700 to 1000 pm) and lack of surficial ornamentation. In addition, T . formosus eggs have a much larger 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 9.1: Phu Phok egg and eggshell parameters egg shape oval/pyriform egg greater diameter 0.18 cm egg smaller diameter 0.11 cm egg diameter ratio 61% ornamentation morphology bi-modial nodular ornamentation size H=92. 6 microns and H=183.3 i pore canal morphology pore canal size pore aperture morphology round/funnel-shaped? internodular space pore aperture size eggshell thickness 353.9 microns number of layers 3 eggshell character state prismatic layer 1 morphology Fan-shaped and bladed layer 2 morphology continuous layer 3 morphology Vertical long crystals layer 1 size 207.8 microns layer 2 size 136.3 microns layer 3 size 32.8 microns ratio total thickness/Ll 1.7 ratio L1/L2 1.52 ratio L3/L2 0.24 core location core size eggshell units morphology eggshell units-thin section monoautochronic ? paired eggs ? egg spatial position in clutch ? air cell yes clutch ? Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 9.1 A. Thin section of an entire egg. Note the multilayered and multidirectional arrangement of the eggshell fragments indicative of a compression of the egg, as it was still whole. Black arrows point to some of the rarer and taller surficial nodes. White arrows point to fractures in the eggshell that likely occurred during a third diagenetic stage. B. SEM of the nodular surficial ornamentation. Note the bimodal distribution of the height of the nodes. This type of ornamentation has been previously reported with elongatoolithid eggs, a parataxonomic family that has been traditionally associated with non-avian theropods. C. SEM of eggshell radial section. Note the distinct presence of three eggshell structural layers, a pore canal and eggshell units broken at their base, leaving only a conic stub at their point of origination. D. Thin section of eggshell radial section. As for the SEM, note the presence of the three structural eggshell layers and also of a black diagenetic line (dl) here interpreted as bacterially mediated micrite. E. SEM showing eggshell structural layer 3 in more detail. F. Thin section showing the difference of crystallographic orientation between L2 and L3. Note also the presence of the micrite diagenetic line on the outer eggshell surface Abbreviations: LI, eggshell structural layer 1; L2, eggshell structural layer 2; L3, eggshell structural layer 3; pc, pore canal; nc, eggshell units nucleation centers; dl, diagenetic line. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 9.1 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. volume of 450 cm 3 (Varricchio et al., 2002) and are elongated (12 to 16 cm) with one pole (7 cm) at least twice as big as the other (3 cm). C. osmolka eggs are long and wide (18 by 7 cm), possess a 0.50 to 0.64 mm thick bi-layered eggshell that displays an aprismatic structure with acicular crystals in layer 1, the eggshell surface is characterized by a linearituberculate ornamentation. Therefore, C. osmolka eggs bare no resemblance with the Thai eggs. Although admittedly very little is known about the eggs of basal birds, the Phu Phok eggs share a similar shape with the previously reported avian eggs from Neuquen (Argentina), a trilaminated, prismatic eggshell structure where layer 1 that is at least 1.4 times bigger than layer 2, a character not yet observed in non-avian theropods or Neomithines. However, they differ from these Patagonian eggs by their nodular ornamentation. To sum up, the Phu Phock eggs display advanced avian features that are shared with the enantiomithine eggs from Neuquen but still retain a pronounced surficial ornamentation that is presently typical of non-avian theropods. In addition to the observed oological characters, the miniaturization of the Thai eggs suggests that Skl-1 was either a small bird more primitive than the Neuquen specimen, or a small, non-avian theropod such as Microraptor zhaoianus (Xu et al., 2000), Epidendrosaurus ningchengensis (Zhang et al., 2002), and Scansoriopteryx heilmanni (Czerkas and Yuan, 2002) that display some characters seen in early birds such as body size reduction and presence of feathers. However little more can be said, because the embryonic bones show few diagnostic features, their systematic position among theropods cannot be very precisely determined. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONCLUSION The small size of these eggs suggests that in the transition from non-avian theropods to birds, eggs followed the body size miniaturisation of the egg-laying species and reach avian size prior or at the level of enantiomithine. The three- layered prismatic eggshell with its layer 1 thicker than layer 2, oval shape, and thin eggshell strongly suggest a basal avian relationship. Interestingly, the pronounced ornamentation of the eggshell is likely a retention of a feature typical of non-avian dinosaurs. Thus, although whether these eggs were laid by an early avialan or by a non-theropod close to the ancestry of remains presently unclear. 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER X Possible Enantiornithine eggs from Neuquem (Patagonia, Argentina) (Taphonomic level 1, taxonomic level only bracketed to two phylogenetic nodes) INTRODUCTION Dozen of small asymmetrical eggs have been found in Cretaceous exposures of the Bajo de la Carpa Formation of Neuquen city, Patagonia, Argentina. A recent publication describes these eggs and associated embryonic remains in ovo the phylogenetic position of which has been bracketed between the two avian Ornithothoraces and Omithuromorpha nodes (Schweitzer at al., 2002) These are groups of basal avialans that had a cosmopolitan distribution in during the Mesozoic (Chiappe and Dyke, 2002). This discovery is important because it permits to connect the morphology of avian Mesozoic eggs with a particular clade of basal birds for the first time. Samples of these eggs are here re-examinaed and discrepancies with the original work are highlighted. This material is important for a better understanding of the distribution of oological and reproductive characters across maniraptoran theropods and thus illuminates aspects of eggshell evolution in modem avians and their closest relatives. Material MUCPv: Eight fossil eggs were found in a non-marine sandstone unit of the Bajo de la Carpa Formation (Campanian) in the city of Neuquen, Patagonia, Argentina. No evidence of nesting stmcture or clutch assemblage was ever been reported but all the eggs were presumed to belong to the same species (Schweitzer et al., 2002), as they all come from the same area and are morphologically identical. Samples consist of small eggshell fragments that were loosely attached to two eggs. Because of extreme 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fragility and thinness only a few attempts to examine this material at the SEM were successful, thus the limited amount of SEM images used to illustrate this study. DESCRIPTION The eggs have undergone only slight taphonomic deformation as attested by their well-preserved shape. Their asymmetric, oval shape (Fig. 10.1A) compares best with that of Gallus gallus but the dimensions of 45 mm (length) by 27 mm (equatorial diameter), as reported by Schweitzer et al. (2002) implies a lesser volume (about 150% smaller) than Gallus gallus. Measurements of the complete egg on loan at the LACM are 20% smaller than those reported by these authors. This discrepancy is easily justified by the absence of preserved eggshell on the egg surface coupled with a possible erosion of the sandstone endocast during diagenesis. To test this hypothesis, another egg which section was fully shelled was digitally “lasso” and copied-pasted onto the whole egg (Fig. 10. IB). Two conclusions came forth from this test: the shape is proportionally identical between the corresponding sections of these two specimens, and eggs that are only partially preserved consist of the halves that have a tapered pole (without air cell). This taphonomic feature coupled with the presence of sandstone infilling and the absence of visible crushed and cracked eggshell on the egg surface advocate that these now partial eggs were hatched then filled up with the surrounding sandy matrix prior to been fossilized. A similar taphonomic process has been reported in modem seagulls where the portions of the eggs that are left intact resting on the ground surface correspond to those without the air cell (Hayward et al., 2000). The polar region that contains the air cell where the chick pips, as well as the region immediately adjacent to the air cell is generally destroyed during hatching. To further test the hypothesis that the missing pole of MUCPv eggs is the most obtuse (the one containing the air cell), a complete fossil egg was digitally bisected in the center and the two poles were superimposed. This independent test supports the previous 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conclusion that the missing poles in the five eggs are indeed those that contained the air cell. The eggshell structure is composed of three distinct structural layers (Fig. 10.1C). Its thickness excluding the surficial diagenetic layer material that exhibits euhedron crystal habits (Fig. 10.ID) averages 285 pm (table 10.1). Layer 3, the outermost layer, averages 49.26 pm in thickness, a value somewhat exceeding by 12 pm that reported by Schweitzer et al. (2002). In the studied specimens, layer 3 is also characterized by having only a few vesicles and by sharing a very distinct boundary with layer 2 (aprismatic condition, see Grellet-Tinner and Norell, 2002). These two features were omitted and misinterpreted respectively in Schweitzer et al. (2002). In contrast to the layers 2 and 3, the contact between layers 2 and 3 is undefined (prismatic condition,, see Grellet-Tinner and Norell, 2002). Nevertheless, layer 2 differs from layer 1 by having a net increase number of vesicles that are also larger than those in layer 1 (Figs. 10.1C, E). The 95.56 pm thicknesss of layer 2 is nearly twice that of layer 3 and markedly differs from the 132 pm value mentioned in Schweitzer et al. (2002). The aspect of the inner sections of this eggshell displays an appearance that results from diagenetic alteration. As a consequence, the eegshell units originating at the base of the layer 1 are well defined (Fig. 10. IF) and extend without major crystallographic interruption into layer 2 but abruptly stopping at the boundary of layers 2 and 3 (Fig, 10.1C). Considering the silicoclastic nature of the egg endocasts, it is probable that some acidification occurred during the fossilization process, a pattern already observed in similar sedimentary contexts for the Late Cretaceous megaloolithid eggs from Uruguay (Faccio, 1994) and for extant gull colonies of seagulls (Hayward et al., 1997). Yet, this weathering offers a natural opportunity to observe the elongate and narrow shape of the eggshell units. The base of layer 1 normally composed of the core of the shell units also suffers from acidification as denoted by the interval in between each unit. The cores themselves 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 10.1: Patagonian “enantiomithine” egg and eggshell parameters egg shape oval/pyriform egg greater diameter egg smaller diameter egg diameter ratio ornamentation morphology ornamentation size pore canal morphology pore canal size pore aperture morphology internodular space pore aperture size eggshell thickness 285 microns number of layers 3 eggshell character state prismatic but aprismatic for L3 layer 1 morphology bladed layer 2 morphology continous butnot at 90 degrees layer 3 morphology amorph layer 1 size 137.2 microns layer 2 size 95.56 microns layer 3 size 49.26 microns ratio total thickness/Ll 2.07 ratio L1/L2 1.43 ratio L3/L2 0.51 core location core size eggshell units morphology slender eggshell units-thin section monoautochronic ? paired eggs ? egg spatial position in clutch ? air cell yes clutch ? 4.5 cm 2.7 cm 60% smooth NA ? ? ? ? ? Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 10.1 A. MUCPv eggs from Bajo de la Carpa Fm. (Campanian), in Neuquen, Patagonia. Although most of the specimens consist only of preserved halves a few complete eggs display a shape like those of extant chicken. They are oval, with one pole more inflated than the other to accommodate an air cell. They average 45 by 27 mm. B. In order to estimate the asymmetry between the two poles, the pictures of the few eggs that were intact were inverted and pasted over the commonly found broken halves. This confirmed the presence of asymmetrical poles but also that the commonly preserved halves were the pointed poles, which is not surprising considering that most modem birds destroy the air cell and the surrounding section during hatching. C. As those from Phu Phock, the MUCPv eggshell structure is tri-laminated (black double arrows in this SEM) with prismatic contact between layers 1 and 2 and aprismatic contact between layers 2 and 3. The latter could be a diagenetic artifact. Equally important, the thickness of layer 1 far exceeds that of layer 2,as already observed in the Phu Phock eggs. D. SEM of the eggshell surface with a cover of euhedral crystals supporting a diagenetic phase that could have modified the contact between alyers 2 and 3. E. SEM shows a greater amount of vesicles (black arrows) present in layer 2 that helps to distinghish this layer from layer 1. F. SEM shows that the bases of the eggshell units display a substantial amount of acidification, a pattern congment with the previously mentioned diagenetic hypothesis. This results in eggshell units truncated at their base and flaring at a 45 degrees angle. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 10.1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. have weathered out and consequently each eggshell unit seems to originate from a flat, horizontal line that flares out on either side at a 45 degrees angled-V-shape. The thickness of layer 1 averages 137.2 pm, a value that also severely departs from the reported 91 pm (Schweitzer et al. (2002). No, pore canals or openings are readily visible in the observed specimens. A positive taxonomical identification of these eggs was provided by the morphology of the embryonic remains in ovo of the MUCPv 284 egg (Schweitzer et al., 2002). Although this association should be optimal for the taxonomic identification of these eggs, the partial ossification of the embryonic remains in ovo and the fact that most of the bones were not preserved precludes a definite taxonomic affiliation of the embryo. Nevertheless, the combined skeletal features allowed these specimens to be bracketed between Omithothoraces and Omithuromorpha and as “most likely an enantiomithine” bird (page 193, line 2 in Schweitzer et al., 2002). Two birds are known from the same location as the eggs: the basal omithuromorph Patagopteryx deferrariisi (Alvarenga and Bonaparte, 1992) and the enantiomithine Neuquenornis volans (Chiappe and Calvo, 1994. Schweitzer at al. (2002) argued that while the embryonic anatomy precludes the identification as the eggs of Patagopteryx, their assignement to Neuquenornis was inconclusive. DISCUSSION Although over all well-preserved, these eggs suffered diagenetic alteration during the fossilization process as evidenced by the dissolution of the original calcite in the structural layers 1 and 2. Layer 3 differs totally from these two layers by its homogenous consistency, few vesicles and very distinct boundary with layer 2. The previous description of these eggs (Schweitzer et al., 2002) did not recognise these features. A possible explanation is that the degree at which diagenesis affected the eggshell varies from egg to egg or within an egg and that this variation could lead to 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different observations. Layer 3 is particularly interesting because presently there is no record of eggshell displaying simultaneously a prismatic condition between layers 1 and 2 and an aprismatic condition between layers 2 and 3. Although unlikely, this feature not yet reported could be linked to the degree of diagenetic re-crystallization and a preferential calcium carbonate deposition in layer 3, but nevertheless it does not exclude a biological origin. The lack of congruence of the respective values of layers 1 and 2 obtained in this research and those previously published (Schweitzer et al., 2002) are more difficult to explian unless they were inverted in the 2002 publication. At best, the erosion of layer 1 could possibly reflect the disparity between its values but layer 2 in either case is sandwiched between layers 1 and 3, thus cannot be thinned out by erosion process . The aprismatic character at the boundary of layers 2 and 3 establishes an unquestionable upper limit for layer 2 and the marked increase of the size and number of vesicles coupled with the change of crystal orientation between layers 1 and 2 provide an unequivocal lower limit for layer 2. In light of these observations and the congruent measurements if inverted, it is possible that the previous reported values inverted the respective dimension of layers 1 and 2. Notwithstanding these incongruences, the presence of a larger layer 1 than layer 2 as observed in this study is congruent with that observed in the tri-laminated eggshell of the derived theropod from Phu Phok but not characteristic of modem birds. The presence of three layers in an eggshell is a synapomorphy of modem birds (Grellet- Tinner, 2000) yet it is observed in the Phu Phok eggs, as opposed to the two-layered condition seen in other non-avian theropod eggs (Grellet-Tinner and Chiappe, 2004). This raises an interesting question about the optimization of this character. Is the coelurosaurian who produced the Phu Phok eggs a close relative to basal birds and thus the presence of a tri-laminated eggshell structure should be considered a synapomorphy to this ceolusaurian clade and all the descendants of its most recent common ancestor? Is a tri-laminated eggshell a homoplastic character that appeared 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in different lineages should the Phu Phok ceolusaurian not be in close phylogenetic proximity to Aves? Admittedly, very little is known about the eggs of basal birds. The fossil record has not yielded sufficient assemblages of basal avians associated with their eggs. Hence, polarization of this character is presently impossible. Furthermore, the feature shared by Phu Phok and Neuquem eggs of having a layer 1 thicker than layer 2 is not present in modem birds or non-avian theropods. This morphometric character suggests that it is synapomorphic to the group including Phu Phok dinosaur and the Neuquem enantiomithine. Other important conclusions can be derived from the morphology of the Neuquen eggs. Two birds were recovered from the Rio Coloado; P . deferrariisi (Alvarenga and Bonaparte, 1992) and N volans (Chiappe and Calvo, 1994) and pursuing the proposed hypothesis by Schweitzer et al. (2002) that one of these birds could father the MUCPv 284 embryo, requires a closer examination of all the data at hand. P . deferrariisi is a chicken size flightless bird and N. volans is a volant bird morphogically smaller than P . deferrariisi. Considering that the shape and size of a modem bird egg is compromise between environmental forces that have influenced the evolution of a species and the size of the hen pelvis, N. volans would be considered a better candidate than P . deferrariisi as the parent lineage for the MUCPv 284. According the Schweitzer at al. (2002), no definite skeletal character permits the assignment of the embryo in MUCPv 284 to N. volans. However again following their set hypothesis, the corroboration between hen and egg size (Rahn et al., 1975) provides an oological character that could establish the membership of these MUCPv specimens to the enantiomithine N. volans. CONCLUSION MUCPv eggs display avian characters in their three-layered eggshell, their oval egg shape with an expended pole for the air cell, and the lack of surficial 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ornamentation. Because the MUCPv embryonic bones show sufficient diagnostic features to associate them with Ornithothoraces, they offer an unprecedented opportunity to connect the observed oological characters with a group of basal birds. In addition, a higher taxonomic identification can be achieved if associating these eggs with one of the two primitive birds recovered from the same location the eggs were discovered. Including oological data in the debate allows us to constrain the identification of these eggs to the enantiomithine N volans. Aside their smooth eggshell ornamentation, the MUCPv eggs share multiple features with the Phu Phock, thus advocating a close phylogenetic proximity between these two lineages. Overall, the Neuquen eggs illustrate the acquisition of egg and eggshell avian characters that took place during the evolution of Mesozoic basal birds. 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER XI The fossil bird Lithornis from England and Montana (Taphonomic level 3, taxonomic level 1) INTRODUCTION Lithornis is one of the first fossil birds discovered. It was described in 1840 by Sir Richard Owen (Owen, 1840) from partial fossil remains preserved within a clay nodule (figured for the first time by Owen, 1841: 206) collected from the Lower Eocene London Clay Formation at the Isle-of-Sheppey (Kent, U.K). This specimen, subsequently to become the holotype of Lithornis vulturinus, was housed in the Museum of the Royal College of Surgeons before its destruction in the bombing of London during World War II. Little more information about Lithornis was published until Houde (1986, 1988) reported the discovery of additional fossil specimens from the Tertiary of North America, which they considered closely, related to L. vulturinus. On the basis of this new material, coupled with re-evaluation of existing fossils not previously recognized as “lithomithid” (Harrison and Walker, 1977), Houde (1988) designated a neotype specimen for L. vulturinus, and named a number of additional species including the North American Lithornis celetius. This taxon, from the Paleocene Fort Union Formation of Montana (Houde, 1988), was placed within the family Lithomithidae along with two additional genera also named by Houde (1988) - Paracathartes and Pseudocrypterus. The systematic position of Lithornis remains questionable. Owen (1841, 1846) considered these birds to be related to extant cathartids (Cathartidae; i.e., extant vultures), whereas Harrison and Walker (1977) placed them closer to the musophagids (Musophagidae; i.e., extant turacos). Most recently, Houde (1988) suggested that 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lithornis is more likely a basal palaeognath, closely related to extant Tinamidae. The goal of this chapter is to re-evaluate the taxonomic status and phylogenetic affinities of L. celetius and L. vulturinus by combining characters from both osteology and oology (eggs and eggshell morphologies) to investigate the evolutionary relationships of a fossil bird. M aterial and Analytical M ethod PVPH 263: Titanosauria eggshell PVPH 264: Titanosauria eggshell PVPH 272: Titanosauria eggshell IGM 100/972; B. jaffei IGM 100/971: C. osmolka TMM(M) 758: Meleagris gallopavo TMM(M) 7592: Anser anser TMM(M) 7588: Anser anser TMM(M) 7585: Struthio camelus TMM(M) 7587: Rhea americana MSC 704: Rhea americana TMM(M) 7594: Dromaius novaehollandiae LACM (to be accessedO: Lithornis vulturinus YPM PU 16961: Lithornis celetius AMNH 13376: Rhynchotus. rufescens AMNH 15478: Ttinamus sp Both oologic specimens are based on their co-occurrence of diagnostic skeletal remains of lithornitid birds (L. vulturinus and L. celetius), these associations are considered as an a priori working hypothesis for this research. Only a single eggshell fragment putatively assigned to L. vulturinus and a few assigned to L. celetius were 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. available. Therefore, each of the samples was broken into two fragments for SEM and thin section observations. A novel approach is used in this chapter to acertain the taxonomic identity of these two eggshells by conducting a cladistic analysis including eggshell of well-identified specimens. The majority of the oological characters used here were taken from Grellet-Tinner’s (submitted) phylogenetic analysis of paleognaths along with a number of new ones that pertain to “lithomithid” eggshells from the Lower Paleocene Fort Union Formation of Montana (Houde, 1988; associated with specimens of L. celetius) and from eggshell associated with a fossil specimen of L. vulturinus from the London Clay Formation of England. The osteological character codings are based on previously described specimens of L. vulturinus and L. celetius. In total, a matrix of 15 taxa and 15 eggshell characters was analyzed using the heuristic search algorithm in PAUP 4.0b4a (Swofford, 2001). Two outgroups (Maddison et al., 1984)—titanosaurid sauropods from Patagonia and an Asian troodontid of uncertain specific affinity—were used to determine the polarity of character transformations. In addition, a sample of anseriform (ducks, geese and relatives) and galliform (chickens, pheasants and relatives) birds was added to the analysis. These two clades are considered to be basal within neomithine phylogeny (Cracraft, 1988). Following initial searches, characters were weighed according to the rescaled consistency index (Carpenter, 1988); this method reweighting offers the benefits of avoiding a priori choices therefore, minimizing researcher bias and choice between alternative and equally parsimonious cladograms based on the most consistent characters. This technique is particularly applicable for small matrices (Carpenter, 1988). DESCRIPTION The osteological characters were taken primarily from the previous descriptions presented by Lee et al. (1997) as part of their analysis of palaeognath 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. inter-relationships. In addition, a number of osteological characters were added to the matrix presented by Lee et al. (1997) as outlined by Dyke (2003). The data matrix and codings for taxa are as given by Dyke (2003). Characters and taxa were coded for the three basal clades of Neomithes (modem birds) (sensu Cracraft, 1988). Codings for Lithornis were based on described specimens of L. vulturinus and L. celetius (i.e., BMNH A 5204, neotype of Lithornis vulturinus)', BMNH A 38935 (holotype of Promusophaga magnified Harrison and Walker; referred to L. vulturinus by Houde, 1988); BMNH A 38934 (referred to P . magnified by Harrison and Walker, 1977; referred to L. vulturinus by Houde, 1988); and BMNH A 5425 (listed as BMNH A 5424 and referred to L. vulturinus by Houde, 1988). The observations of L. vulturinus are restricted to an eggshell fragment collected from the Lower Eocene London Clay Formation at Walton-on-the-Naze, Essex, England. Although this specimen was not found as part of a complete egg containing a Lithornis embryo in ovo, it was closely associated with postcranial material certainly referable to this taxon (Dyke, unpublished data). This specimen was found preserved in a soft clay nodule and thus suffered no diagenetic transformation detrimental to our analysis. In tangential view, the smooth outer surface of this specimen displays a number of circular pore orifices with diameters that range from 73 to 80 pm (Fig. 11.1A). However, no pore canals were visible in cross section, an important character considering that tinamous, Galliformes, Anseriformes and rheas display a straight single pore from the inner to the outer eggshell surfaces, whereas Struthio exhibits a branching pore canal system (Grellet-Tinner, 2001). Cross sectional view clearly reveals the presence of three structural aprismatic layers (with a clear delimitation between each layer, e.g, Grellet-Tinner and Norell, 2002), the second of which is by far the most conspicuous (Fig. 11. IB). Layer 3 is amorphous in respect that no distinctive crystallization present is visible, but nevertheless the general crystallographic orientation is vertical and neatly differs from layer 2 (Figs.IB, C). 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The thickness of Layer 3 based on eight measurements averages 24.68 pm, a value slightly less than that of layer 1. Layer 1 consists of shell units juxtaposed to each other and averages 39.36 pm based on three measurements. The base of the shell units are rounded but not inflated and display proportionally long bladed calcite crystals (Fig. 11. ID). Although layer 1 is very thin and resembling by its proportion those of tinamous, the shape of the calcite crystals forming the base of each eggshell unit is reminiscent of those of Struthio and Rhynchotus rufescens eggshells (Figs. 11.IE, F). Layer 2 by far the thickest of all averages 307 pm based on three measurements and its crystallographic orientation neatly contrasts from layers 1 and 3. Ratio of layers 3 to 2 and 1 to 2 respectively are 0.08 and 0.128. The observations of L. celetius are also restricted to a single eggshell. The particular sample (YPM PU 16961) was chosen for its preservation among other eggshell fragments housed at the YPM. Museum entries indicates that all these eggshell fragments originated from the Fort Union Formation (a site in the Gallatin National Forest, Park County, Montana) and as for L. vulturinus were not associated with any embryonic remains in ovo. These YPM PU specimens were collected at the Bangtail Quarry where the entire skeleton of this species is known from a composite series of bones (Floude, 1988). In contrast to the L. vulturinus specimen, YPM PU 16961 has been diagenetically recrystallized and thus the internal eggshell structural layers are only visible in small sections at one time. These specimens were likely found in hard nodules at that same site as described by Houde (1988) thus explaining the diagenetic alteration. The conditions are such that the thickness and structure of the third layer might observable at one place whereas those of the second layer are offset in respect to the other two structural layers. Although observation of the entire eggshell thickness at one time in complete section was problematic, the observed features when taken separately were preserved enough to allow a detail description. In tangential view, the smooth outer surface of this specimen displays a 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 11.1. SEM of shells of Lithornis vulturinus and Rhynchotus rufescens (Tinamidae; AMNH 13376). A. Circular pore orifice on the eggshell outer surface of L. vulturinus. B. Three aprismatic structural layers compose the entire thickness of the eggshell of L. vulturinus. Layers, 1 and 3 are relatively thin comparatively to layer 2 in L. vulturinus. C. Detail of layer 3 of L. vulturinus. D. Bladed calcite crystals at the base of eggshell units in L. vulturinus. E. Layer 3 of the eggshell of extant R. rufescens. F. Structural layers in R. rufescens with bladed calcite crystals at the base of eggshell units. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 11.1 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. couple of circular pore orifices with diameters visible in the micro digital photo of the specimen (Fig. 11,2A) but not on the eggshell fragment used for SEM observation. In that respect, no pore canals were observed at the SEM as for L vulturinus. Cross sectional view clearly reveals the presence of three structural aprismatic layers (Fig. 11.2B) a key character synapomorphic to the Paleognathae (Grellet-Tinner, 2001). Layer 2 is by far the most conspicuous of all averaging 485.68 pm (Figs. 11.2B, C). Layer 3 is amorphous and similar by its morphology to that of L. vulturinus and Ttinamus averaging in size 38.07 pm (Figs. 11.2D, E). Layer 1 as that of the English specimen is proportionally thin in respect to layer 2 averaging only 51.68 pm (Table 11.1). Flowever, it strikingly differs from L vulturinus by having an acicular crystallization at the base of the eggshell units like that of T . tinamus (Fig. 11,2F) rather than long bladed crystals. Ratios of layers 3 to 2 and 1 to 2 respectively are 0.078 and 0.106. DISCUSSION Two initial unweighted analyses of the eggshell data set were conducted alternatively using Rhynchotus rufescens (AMNH 13376) and Tinamus sp. (AMNH 15478). This permutation is justified because the few examined taxa among the 47 living species of Tinamidae (Sibley and Ahlquist, 1990) reveal a dichotomy in the shape of calcite crystals at the base of the eggshell units exhibiting either acicular or bladed morphs. Presently it is impossible to determine the primitive state of this character among tinamous. The ambiguity among extant species is mirrored by osteological characters and as such the internal topology of Tinamidae to date remains unresolved (S. Bertelli, and A. L Porzecanski, pers. comm. 2002). Two analyses each yielded 39 MPTs each of 316 steps in length. With respect to the outgroups, Troodontidae and Titanosauridae, the resultant strict consensus representation of these trees shows little resolution within the ingroup (not shown); the two Lithornis 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 11.1: Lithornitidae egg and eggshell parameters egg shape Lithornis celetius ? Lithornis vulturinus ? egg greater diameter ? ? egg smaller diameter ? ? egg diameter ratio ? ? ornamentation morphology smooth smooth ornamentation size NA NA pore canal morphology ? ? pore canal size ? ? pore aperture morphology round round internodular space na na pore aperture size ? 73-80 microns eggshell thickness 566 microns 371 microns number of layers 3 3 eggshell character state aprismatic aprismatic layer 1 morphology acicular long bladed layer 2 morphology continuous at 90 degrees continuous at 90 degrees layer 3 morphology amorphous with vertical amorphous with vertical layer 1 size orientation 51.68 microns orientation 39.36 microns layer 2 size 485.68 microns 307 microns layer 3 size 38.67 microns 24.68 microns ratio total thickness/Ll 10.95 9.42 ratio L1/L2 0.10 0.12 ratio L3/L2 0.08 0.08 core location ? ? core size ? ? eggshell units morphology bowl shaped at the base then bowl shaped at the base then eggshell units-thin section elongated elongated monoautochronic ? ? paired eggs ? ? egg spatial position in clutch ? ? air cell yes yes clutch ? ? Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 11.2. SEM of L. celetius (YPM PU 16961) and Tinamus sp. (Tinamidae; AMNH 15478). A. Circular pore orifice on the eggshell outer surface of L. celetius (arrow). B. Although hard to observe, the entire thickness of the eggshell of in YPM PU 16961 consists of three aprismatic structural layers. Layers, 1 and 3 are also relatively thin comparatively to layer 2 in L. celetius. C. Detail of layers 1 and 2 of L. celetius. Note the acicular crystals at the base of the eggshell units. D. Detail of layer 3 of L. celetius. E. Layer 3 of the eggshell of extant T . tinamus. F. Structural layers 1 in T . tinamus with acicular calcite crystals at the base of eggshell units. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 11.2 Jt, . ■ * * -*.....«*■■ * ' " j T ^ g ? j'<v i 5 % S ^ V f ■ ... flih M .S .J h ; a » r " k ■ $ 4 \ - V ‘J y *, ■ » _ J A N A *-*- ‘ * „ ~ /‘\V > i ' • - »«*;/*■,'■ ;r» v » ^.w^.:.-------------- -»™ at™ w Mm — i i »------------- 1 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. specimens cluster towards the base of an unresolved neomithine topology. Re weighting the characters by use of the rescaled consistency (Carpenter, 1988) index generated 4 MPTs each of 32 steps in length. The strict consensus representation of these trees indicates somewhat more resolution within the ingroup; the two Lithornis specimens are placed within Palaeognathae, L. vulturinus as the sister- group of Struthio and L celetius sharing a most recent common ancestor with either of the extant Tinamidae used in this analysis. In line with recent morphological and molecular analyses, palaeognaths are recovered as a distinct clade with respect to neognaths, here represented by Galliformes and Anseriformes. The relationships of palaeognaths recovered by this analysis (to the exclusion of Lithornis) are in concordance with those reported by Grellet-Tinner (2000). Trimming down the number of taxa by eliminating the extinct species produces a tree where the internal topology of the Australasian clades is congruent with that of Lee et al., (1997), but differs in the unresolved positions of Rhea and Tinamidae. Possibly, this may represent an analytical artifact resulting from of our exclusion of extinct taxa from the analysis (Gauthier et ah, 1988). Parsimony analysis of the morphological data set including a composite “London Clay Lithornis’ ’1 resulted in the production of a single MPT of 92 steps in length (see Dyke, 2003 for details). In this single tree, Lithornis is hypothesized to be the sister-taxon to the monophyletic ratites. Interestingly, inclusion of Lithornis to the matrix effects no changes to the topology for extant taxa proposed by Lee et al. (1997). The position of the composite “London Clay Lithornis” in a more derived position than tinamous, and as the sister-taxon to the other included taxa is supported in this analysis on the basis of the following derived characters (numbered as listed by Lee et al., 1997 and as preserved in the specimens considered): 8 (scapula and coracoid fused; seen in BMNH PAL A 5303 and A 5425); 12 (internal tuberosity of humerus knob-like, having a degree of medial protrusion; seen in BMNH PAL A 5204, 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BMNH PAL A 5303, BMNH PAL A33138, BMNH PAL A 38934, BMNH PAL A 5425); 13 (deltoid crest raised from the base of the external tuberosity seen in BMNH PAL A 5204, BMNH PAL A 5303); 18 (pronounced external epicondyle of humerus; seen in BMNH PAL A 5425). Combining the taxon-character data sets for both eggshell morphology and osteology was done in two ways: first, an analysis of all available data including a number of taxa that have missing codings for one or other of the character sets; second, a combined data only for taxa where both oological and osteological data were available. In the first analysis, for example, the ootaxon Struthiolithus (Lydekker, 1883; Lowe, 1931) and the elephant bird Aepyornis lacked osteological codings—these taxa were removed for the second part of the combined analysis. Codings for Lithornis were combined into a single composite as was done for the osteological matrix discussed above. Because of obvious problems regarding the homology of characters between extant palaeognaths and non avian dinosaurs, taxa of Galliformes and Anseriformes were used as outgroups for the combined analyses. As used for the analysis of eggshell characters alone investigation of the effects of re weighting characters in the combined analysis by their rescaled consistency index was performed. Unweighted analysis of the combined data including all taxa and characters resulted in the production of 3 MPTs of 109 steps in length. The strict consensus of these trees places Lithornis within a polytomy basal within palaeognaths but derived with respect to the tinamou. This placement is consistent with our results reported above as well as with the suggestion of Houde (1988) based only on morphological evidence. Combined analysis following the removal of taxa incurring missing character codings resulted in the production of a single MPT of 106 steps in length. This single tree exhibits additional resolution within the ratites in line with the 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. previous suggestions of Lee et al. (1997), placing Lithornis as more derived with respect to tinamou. Characters supporting these placements are discussed above. CONCLUSION The a priori working hypothesis to identify the loosely associated eggshells is supported by this phylogenetic results by placing both specimens within Paleognathae. Consistent to the investigation of the phylogenetic placement of Lithornis is the position of this fossil bird within palaeognaths, most often basal with respect to the ratites. Although never considered previously by use of numerical cladistic analysis, this placement is consistent with the suggestion of Houde (1988) who depicted a number of possible alternatives for the phylogenetic position of these birds. Although limited by the available data, especially in the case of the eggshells, these analyses highlight the usefulness of combining both oology and osteology in resolving the placement of Lithornis. At the base of the tree, both kinds of data are independently able to distinguish between the major clades of Neomithes, Palaeognathae and Neognathae. However, at higher levels within palaeognaths it appears that osteology is required to achieve additional clear resolution within the ratites (Bledsoe, 1988; Lee etal., 1997). Although Lithornis has been known for over 100 years and have been postulated as representing a basal palaeognath since the 1980s, little phylogenetic character data has yet been proposed to support this contention. Some osteological evidence has been presented elsewhere (Houde, 1988) but this has only recently been subject to test on the basis of numerical cladistic analysis (Dyke, 2003; this paper). More data, especially in the form of better preserved fossil material (in particular eggshells certainly associated with embryonic remains in ovo as well as three dimensionally preserved cranial material), will be require to assess the significance of Lithornis and related birds to the general question of palaeognath and basal neognath 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. evolution. The phylogenetic relationships presented in this chapter are consistent with the hypothesis of a single loss of flightlessness with palaeognaths (within the stem lineage leading to extant ratites) as well as with recent proposals that the initial divergence of the group occurred in the southern hemisphere (Gondwana; Cracraft, 2001). The presence of phylogenetically constrained fossil specimens in the earliest Tertiary of the northern hemisphere implies that the evolutionary divergence of all palaeognaths must have occurred at an even earlier time (perhaps even prior to the K- T boundary) but better preserved material from the Cretaceous is required to test this hypothesis. 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER XII Paleobiology, oological trends, and Phylogenetic analysis Paleobiology and oological trends As a model of archosaurian reproduction, the two extant groups, crocodilians and birds, differ in a number of features (Fig. 12.1A). Crocodilians produce large clutches of symmetrical eggs that are laid in masse from two functioning ovaries in contrast to birds that produce a limited amount of asymmetrical eggs that are laid in a monoautochronic mode from a single ovary (Proctor and Lynch, 1993). Crocodilians eggshell is composed of one structural, calcitic layer (Schleich and Kastle, 1988) and modern birds display at least three layers (Mikhailov, 1997; Grellet-Tinner and Chiappe, 2004). Parental involvement for crocodilians is generally limited in assisting the hatchlings during and after hatching, in contrast birds invest a considerable amount of time and energy to incubate their eggs and rear their brood (Proctor and Lynch, 1993). Nests are simple for the former consisting of holes dug in the sand or vegetal mounds, but generally turn into elaborated structures for the latter. Although the studied taxa in this work represent only still-frames of the entire picture of the saurischian evolution, they clearly reveal the reproductive and oological trends uniting all these groups; Namely, a reproductive evolutionary cline from crocodilians to modern birds, and a noticeable pattern of coeval development between the acretion of eggshell layers, presence of larger air cells, the reproductive organs, and brooding/ incubating behaviours. Here, a review of the significant evolutionary steps observed in the studied samples is presented. As crocodilians, the titanosaurs from Auca Mahuevo laid their egg “en mass” in nearly analogous nesting environments. However, titanosaurs developed innovative adaptations observed in their eggshell structure that are not shared with 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. non-dinosaurian reptiles to cope with these environments: Their eggshell thickness is perforated by numerous straight and “Y” shaped pore canals originating in between surficial nodes a condition interpreted as a response to a high nesting moisture content; The (nodular) eggshell ornamentation-a dinosaurian synapomorphy when considering other studied dinosaurs- allows a greater air flow for eggs in the nests covered by vegetation; Titanosaurid dinosaurs from Auca Mahuevo laid numerous, symmetrical but mostly subspherical eggs in rimed nests (Chiappe at al., 2004); whether the acicular nature of the calcitic eggshell is an environmental adaptive strategy or reflects a different chemical equilibrium in the reproductive tracts of the titanosaurs remains untested but markedly differs from the eggshell microstructure of crocodilians. Interestingly, the acicular calcitic crystallization structure is also present in layer 1 of D. antirrhopus (Makovicky and Grellet-Tinner 2000) and C. osmolka eggs but absent in this structural layer of troodontid and extant bird eggshell. However, developmental studies have revealed that the bladed-calcite crystals in modem birds have an acicular state that transforms into the bladed morph during oogenesis (Board and Sparks, 1995). These combined observations advocate that acicular calcite crystals seen in the eggshell microstructure of titanosaurid eggs is synapomorphic for Saurischia. Although rare, few life assemblages consisting of a brooding parent on egg clutch have been reported in the fossil record (Norell et al., 1994, Clark et al., 2001). Therefore, it would seem plausible that in view of the expanse of the sedimentary egg-bearing layers of Auca Mahuevo and the lagerstatten preservation state in some areas of this site that, if present, such life-assemblages would have been recorded. Consequently, as for basal reptiles, the lack of adult skeletons in the sedimentary egg-bearing layers of Auca Mahuevo is interpreted as a lack of developed parental involvement during the incubation and hatching of the eggs. To sum up, the titanosaurids from Auca Mahuevo still share a number of characters with crocodilians, namely; symmetric eggs, one eggshell structural layer that is calcitic, lack of a defined 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12.1. A. Co-evolution of eggshell structure, egg shape and ornamentation, with nesting structures among the studied taxa. Ontogenetic change of acicular to blade-shaped crystals as observed in modern birds. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Titanosaurid D. antirrhopus Troodontids Enantiornithine Crocodylia C. osmolka M. xixiaensis Phu Phock Paleognathae & $ ! Layer l|ji NEORNITHES ORNITHOTHORACES Neognathae MANIRAPTORA SAURISCHiA C O Figure 1 2 .1 A Figure 12. IB. Cladogram solely based on the optimization of oological and reproductive characters of the studied taxa. Note the coeval appearance of brooding behavior with a change in the architecture of the nest, egg, and eggshell in oviraptorids and a second similar advent at the level of troodontids when incubation developed a. Presence of surficial ornamentation. b. Acicular crystals as building blocks of the eggshell structure. c. Eggs contained within a rimed nest. d. Nodular ornamentation in titanosaurids. e. Presence of two and aprismatic. f. Presence of acicular crystals limited in layer 1 and crystal orientation of layer at 90 from that of layer 1. g. Linearituberculate ornamentation. h. Elongated eggs i. Presence of a proto-air cell. j. Appearance of a monoautochronic ovideposition as indicated by the eggs arranged in pairs k. Eggs are laid on the perimeters of circles that superposed in 2-3 layers and with an empty space in the center of the clutch. 1 . Presence of brooding behavior. m. Differentiation of organic lines within layer 2, creating 2 sub-divisions, n. Presence of blade-shaped crystals in layer 1. o. Presence of a single circle of eggs. 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. p. Presence of a fully-developed air cell. q. No space devoid of eggs in the center of the clutch. r. Reduction from two to one functioning ovary. s. Presence of two and prismatic eggshell structural layers. t Eggs are vertically oriented in the substrate with air cell up. u. Absence of eggshell surficial ornamentation. v. Presence of proto-avian incubation. w. Presence of bi-modal nodular ornamentation x. Presence of three prismatic eggshell structural layers. y. Layer 1 wider than layer 2. z. Presence of aerial nests. s-1. Character reversal for the Paleognath clade: prismatic eggshell boundaries. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12. IB C D C O 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. air cell, in “masse” ovideposition, lack of pre-hatching and hatching parental care. Likely, these reproductive features suggest that these dinosaurs also retained two functional ovaries, and needed environments that supported high moisture content in their nests. However, they already differ from primitive archosaurs by having a pronounced surficial eggshell ornamentation, sub-spherical eggs, acicular calcitic crystals, and numerous pores to facilitate gas exchanges between the growing embryos and the outside, and rimed nests. The oviraptor, C. osmolka laid long oval eggs that averaged 7 by 18 cm with a polar asymmetry not as pronounced as that of extant birds, causing earlier mi s-identifications (e.g. Clark et al., 1999). The elongate aspect of these eggs is a mechanical response to a physiological requirement. With few exceptions, the maximum diameter of avian eggs is limited by the diameter of the pelvic opening of the female (Smart, 1995; Carpenter, 1999). The size of the C. osmolka egg indicates that the pelvic opening of the female approximated 7 cm and that there was a need to pack more food into the eggs. The latter was accomplished by elongating the eggs and, furthermore, suggests that the chicks were precocial. Precociality implies a longer incubation time than altriciality. Aside from IGM 100/979 other fossils of oviraptors have been found sitting on their nests (Dong and Currie, 1996) like modem birds. Emus, for instance, spend long incubation periods with little food intake, an activity that requires a great energetic investment. This in turn, could bring the brooding parent to near lethargic levels. If oviraptorids brooded their clutches for similarly long periods of time, it could explain their burial atop their egg clutches by low energy “mass wasting” of sediments from destabilized dunes (Loope at al., 1998). When recovered in such life assemblages, oviraptors sit atop a clutch with an empty space in the center and composed of multiple rows of paired eggs. At the same time this parental care supports a brooding activity, it denies the possibility of any aspect of incubation by having multiple rows of eggs laying in a sub-horizontal 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. position on the top of each other and also no sign of egg manipulation. Nevertheless, this brooding behavior is interpreted here as a precursor to avian incubation and, interestingly, C. osmolka burial atop their egg clutch advocates that the behavior of long brooding periods predates the most recent common ancestor of modern birds. Furthermore, brooding appears in the same taxon that laid eggs with a faint asymmetry, here correlated to the presence of an air cell (Figs. 12.1A, B), a feature that is absent in reptilian eggs (Iverson and Evert, 1995). Flowever, the small size of the air cell size argues that in contrast to its fully developed avian homologue, it was only a proto-air cell. Another feature typical of C. osmolka is the pairing of its eggs (Fig. 12.IB). Such a spatial arrangement is indicative that this oviraptor had two functioning ovaries as sauropods and likely all primitive saurischians, but in addition the paired arrangement advocates for a monoautochronic ovideposition as previously suggested (Clark et al., 1999; Norell et al., 2001) and no post partum manipulation of the eggs by the parent. In addition to this mosaic of characters supporting an avian trend, the eggshell of C. osmolka is structurally divided in two distinct layers (Fig. 12.1A). In that respect it is worth noting that although the volume C. osmolka eggs are at least equal to that of titanosaurs from Auca Mahuevo, their eggshell thickness is twice as thin and their eggs support the weight of a brooding adult and other eggs. Kamat et al. (2000) argued that mollusks achieve a great mechanical equilibrium by increasing the number of structural layers of their shell and emphasis that differences in crystallographic orientation among layers enhances this trend. C. osmolka eggshell fulfill this prerequisite by having two structural layers that are aprismatic. This eggshell structure fulfills two mechanical constraints: A need for resistance to greater external forces (load from the other eggs and the brooding adult) and an eggshell thin enough to be broken by the hatching embryo. Moreover, the surficial ornamentation of the eggshell was previously interpreted as a feature that solely favors air circulation among the eggs in a clutch and that keeps the pore apertures free of 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. any obstruction. In addition to these properties linked to airflow, I propose that, the presence of longitudinal grooves and ridges in the linearituberculate ornamentation of C. osmolka eggshell increases its mechanical resistance to external stress. To sum up, the mechanical properties (the eggshell ornamentation, the egg sub-horizontal spatial arrangement, and the double aprismatic eggshell structure) are congruent with the brooding behavior observed in IGM 100/979 (Norell et al., 1994, 2001; Clark et al., 1999) and other similar life assemblages (Dong and Currie, 1996). Numerous features advocate that this oviraptor had already developed a mosaic of characters that are regarded as evolutionary precursors to a full avian reproductive style (Figs 12.1A, B). Although limited, the oological material recovered with the dromaeosaur AMNH 3015 (D. antirrhopus) offers interesting interpretations. The apposed egg against the outer surface of preserved gastralia suggests that D. antirrhopus advocates a brooding behavior at least identical to that of oviraptors (Fig. 12. IB). This occurrence implies that this form of parental care should have been present in the most common ancestor to this group and Oviraptorosauridae. Because of their similar bi-laminated and aprismatic eggshell structure, acicular crystals in layer 1, surficial ornamentation, and body size, D. antirrhopus and C. osmolka likely laid eggs of similar shape and size. Although the shared characters between these two theropods could imply comparable nest organization and structure, it is impossible to predict yet whether D. antirrhopus laid its eggs in multiple and superimposed rows with an empty space in the center as C. osmolka or adopted a nesting strategy more resembling that of M xixiaensis with one single row of eggs. As in M xixiaensis, the eggshell structural layer 2 of D. antirrhopus is divided in two subsections according to the degree of compaction of the organic lines, a feature however, not observed in C. osmolka (Fig. 12. IB). Although “Baby Louie” has been informally identified as a therizinosaur or a gigantic oviraptor, it is only considered as a theropod in this research. Nonetheless, a few biological and phylogenetic inferences can still be proposed based on the type of 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. eggs and eggshell structure while awaiting a precise taxonomic identification based on skeletal characters. Baby Louie egg morphology and eggshell structure links its eggs to Macroelongatoolithus xixiaensis oospecies. These M. xixiaensis eggs approximate 40 by 16 cm and mostly recovered in pairs (Carpenter, 1999). The paired eggs are either parallel to each other or lay at a 13 degree angle (Fig. 12.1 A) (Youxing et al., 1995). These two patterns are consistent and are observed in a complete clutch containing 13 pairs of eggs housed at the Nanyang Museum. Comparisons with oviraptor eggs are necessary to better understand the reproductive characters of the theropod linage that laid the M. xixiaensis eggs. As C. osmolka, M. xixiaensis eggs are elongate with a proto-air cell, they are arranged in pairs forming a circle with an empty space in the center, and their eggshell is also bi-laminated (Fig. 12.1 A). However they differ from the former by their gigantic size, the presence of elongated blade-shaped crystals in layer 1 (Fig. 12.1 A), and an aprismatic condition that is not as well defines as that of C. osmolka. The synapomorphies shared between these two theropods groups indicate that they have the same general reproductive behaviors, but the autapomorphies of M. xixiaensis suggest a few variations. The gigantic size of M. xixiaensis eggs is indicative of a large theropod. Likewise, the meter and a half wide empty space in the clutch center and the lack of superposed egg layers in the clutch (a mechanical constraint related to weight bearing) support this size inference. However, the difference of crystallography (elongated, blade-shaped/ acicular for oviraptor) cannot yet be attributed to a difference in egg size. Although, elongated blade-shaped crystals are closer to acicular than the other types of blade-shaped morphs (in the calcite crystallographic spectrum), they still express a variation that is genetically controlled (Board and Sparks, 1995) that have no known biological relevance yet. However, elongated blade-shaped crystals could indicate that, although closely related to oviraptors, M xixiaensis theropods are not quite on the same phylogenetic branch (Fig. 12.IB), or that blade-shaped crystals is plesiomorphic to Theropoda 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and acicular crystallization is a homoplasy, or that blade-shaped crystallization is a synapomorphy of Mxixiaensis, troodontids, and avians (Fig. 12.IB). In conclusion, the theropod lineage that laid M xixiaensis eggs shared with oviraptorids similar proto-avian characters, namely: brooding, monoautochronic ovideposition, empty space in the center of the clutch, egg laid on the perimeter of a circle, ornamentation composed of ridges, a bi-laminated and aprismatic eggshell structure. However, two characters depart from oviraptorids: the presence of a single row of eggs that could easily be correlated with weight bearing constraints of the brooding dinosaur, and the crystallization in eggshell layer 1 that is likely indicative of a phylogenetic closer to troodontids and avians. In many respects, troodontids (T. formosus in North America and B. Jaffei in Mongolia) offer intriguing perspectives on the avian evolution of non-avian theropods that were absent in C. osmolka, M. xixiaensis, and D. antirrhopus. First, their eggs are asymmetrical and conical (Fig. 12.1A). This asymmetry is far more pronounced than that of C. osmolka and M. xixiaensis and closely resembles what is commonly observed in modem birds. Thus, troodontids would possess a fully developed air cell instead of a proto-air cell as inferred for oviraptorids and elongatoolithid eggs. Second, in contrast to other known theropods and in resemblance to extant birds, troodontids eggshell lacks surficial ornamentation (Figs. 12.1A, B). Third, by its short and blade-shaped calcite crystals, the eggshell layer 1 is similar to the condition observed in modem neomithines (Fig. 12.1A). Fourth, the eggshell structure brings up another similarity with neomithines by the prismatic condition between layers 1 and 2 (Figs. 12.1A, B). Lastly, both troodontid species displays a novel spatial arrangement of their eggs that mimics that of modem pluvianids and, although, B. jajfei egg clutch was not recovered with an adult atop as T . formosus (Varricchio et al., 1997, 1999, 2002), several interpretations could be made about the parental care behavior and reproductive physiology of troodontids in general. The most important 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. predictions are that, aside a monoautochronic ovideposition, these two troodontids had only one functional ovary as hypothesized in chapters 6 and 7, and also already developed a true proto-avian incubation in contrast to the brooding behavior observed in other theropods lineages. The proto-incubation is best described as sporadic parental involvement to avoid a drastic cooling at night of the eggs that are vertically embedded in the stratum with their large poles (with air cell) up and barely surfacing from the sediments (Fig. 1 A). In addition, because of the hot environments where these troodontid species lived, the incubation by day was restricted to cooling the eggs, possibly like extant pluvianids, by wetting the clutch (Grant, 1982; Howel, 1979). To sum up, the presence of unquestionable egg asymmetry, eggs vertically embedded “heads up”, prismatic eggshell layer with bladed-shape crystals in layer 1, and an adult sitting atop the semi buried egg clutch, advocate that troodontids adopted an avian reproduction mode and incubation style before the rise of birds. Furthermore, all these avian characters seem to originate concomitantly within the same species, possibly indicating an obligatory coeval development between egg shape and eggshell structure with reproductive physiology and nesting behaviors (Zhao, 2000). The main missing character to fully transform the reproduction of troodontids into an avian reproduction is a tri-laminated eggshell structure. Although not truly identified as enantiomithine, the Neuquen eggs have been bracketed between Omithothoraces and Omithuromorpha based on the few skeletal characters observed from an embryo in ovo (Schweitzer et al., 2002). However considering that N. volans and P . deferrarrisi are the only two avians recovered in that same locality and stratum, basic oological morphometry dictates that the Neuquen specimens should be associated with the enantiomithine N. volans. These eggs are crucially important due to their taxonomic association with enantiomithines. As predicted by their phylogenetic proximity to extant birds these eggs are asymmetric indicative of the presence of a fully developed air cell, shaped as those of a modem 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. galliform bird, and their eggshell has no surficial ornamentation, which contrasts with the condition of most of their non-avian ancestors (Figs. 12.1A, B). Most importantly, the eggshell structure is tri-laminated with a prismatic condition and with short blade-shaped calcite crystals in layer 1 (Fig. 12.1A). Although no nest was recovered (Schweitzer et al., 2002), the previous observations and interpretations on oviraptorids and troodontids coupled with the tri-laminated eggshell of these birds advocate that enantiomithines would have had a avian reproductive system and likely incubate their eggs as modern birds do. However, one ubiquitous feature differentiates the eggshell structure of enantiomithines from that of modem birds. The thickness of layer 1 exceeds that of layer 2, so much that the ratio of layers 1 to 2 is around 1.4 (Fig. 12.IB). Interestingly, this character state has not been observed yet in non-avian theropods. As the Neuquen specimens, the Phu Phock eggs were discovered in proximity of each other in a small area but without any preserved nesting stmcture. Likewise, an embryo is preserved in ovo, but its identification is limited to that of a theropod due to the poor preservation of the material. However, egg and eggshell features are more diagnostic than the skeletal remains and by comparison with other identified oologic material offer a novel perspective of the transition between non-avian and avian theropods. As the enantiomithine material, the better-preserved Phu Phock egg display the oval and asymmetric shape characteristic of modem birds. Likewise, its eggshell stmcture is tri-laminated with prismatic contacts in between layers. Moreover, its shares with enantiomithine the unique character state of eggshell layer 1 being thicker than layer 2 with a ratio around 1.4 (Fig. 12. IB). This feature alone indicates a closer phylogenetic proximity between the Thai and the Patagonian eggs than with either non-avian theropods or modem birds. In contrast to enantiomithines eggs, the Thai eggs still display a pronounced nodular ornamentation (Fig. 12.1A). This feature is presently only known in non-avian saurischians and its presence alone 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. would associate these eggs with non-avian theropod lineages. However, all known non-avian maniraptors laid eggs that were at least 200 times more voluminous than the Neuquen eggs (Schweitzer et al., 2002) and 400 times bigger than the Phu Phock eggs. The extremely small size of the Phu Phock eggs can only be associated with a proportionally small-bodied parent. Body size reduction at the transition of non- avian to avian theropods has been hypothesized and verified on numerous accounts with the recent discoveries of small-bodied theropods covered with feathers from the Jehol biota in the Cretaceous layers of Northeastern China. To sum up, the Phu Phock eggs by their mosaic of characters are likely the product of either a small non-avian theropod, or have been laid by a bird that is more primitive than enantiomithines and phylogenetically bracketed between the Aves and Omithothoraces nodes (Fig. 12. IB). The studied suite of oological characters and the coeval evolution of the reproductive and nesting behaviours are optimized on a cladogram (Fig. 12. IB) that only includes oological and associated characters, and disregards skeletal-based character optimization. A couple of patterns are ubiquitous. Regardless, of the position of the Phu Phock eggs in respect to the Aves, most of the oological characters and reproductive behaviours associated with modem birds are rooted among non- avian theropods. Although possibly attributed to a taxonomic bias, most of these pre adaptations are grouped in two main nodes: One at the level of Oviraptorosauridae, and the other at Troodontidae. Although that undeniably these two theropods taxa represent important steps in the evolution of avian reproduction, the phylogenetic distance between Oviraptorosauridae and Titanosauria, for instance, cannot be ignored and likely indicates that the reproductive features that appeared in block in oviraptors might have evolved independently across more basal theropods clades. Likewise, although troodontids are in this analysis the obvious precursors to modem avian reproduction, the importance of small-bodied theropods such as those who laid the Phu Phock eggs cannot be dismissed and the eggs of such dinosaurs advocates their 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. closer phylogenetic ties to Aves than troodonts. Dinosaur systematic, oological and total evidence phylogenetic analyses The main precept of cladistic analysis rests on the concept that features shared by organisms provide a valuable evidence of their relationships (Farris, 1983; Kluge, 1997). Cladistic analyses are related to number and the validity of characters entered for each operational taxonomic unit (OTU), the obtained results (relationships among OTUs) should be regarded as working hypotheses that evolved as new data or data subsets are added. Until recently, cladistic analyses of extinct vertebrates were mostly based on observations of skeletal morphology. They did not include other data subsets that were used in analyses of extant animals and that prove invaluable as they brought significant amount of new data to the phylogenetic debate at large. For instance, a few paleornithologists have already demonstrated the usefulness of nest structure (Proctor and Lynch, 1998), plumage (Sara bertelli, submitted), and recently oology in a broad sense that includes reproductive systems (Grellet-Tinner, 2000; 2004) to recover relationships among certain groups of birds. However, before analizing the phylogenetic relationships of the studied taxa in this research a succinct summary of the dinosaur evolutionary relationships is presented here. Regardless of the adopted phylogeny, the general consensus is that Dinosauria is divided into two major groups: Omithischia and Saurischia. Because the here- studied taxa are saurischian dinosaurs a closer review of Saurischia is hereby offered. Saurischians gave rise to the herbivore sauropodomorphs and the carnivore theropod dinosaurs (Sereno, 1999). Sauropodomorpha regroups more inclusive groups, such as sauropods, eusauropods, neosauropods, diplodocoids, and titanosauriformes to which belong the titanosaurs from Auca Mahuevo. The cosmopolitan titanosauriformes were characterized by their sheer size, a quadrupedal posture, a vegetarian diet, and a gregarious life style. The bipedal theropod predators are divided into ceratosaurians, 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12.2 A. 50% majority rule consensus tree of 40 MPTs with a Cl of 57. A matrix of 14 taxa and 23 characters with 7 ordered characters (7, 12, 15, 16, 17, 18, 19, and 20) was analyzed using the heuristic search algorithm in PAUP 4.0b4a (Swofford, 2001) after defining Crocodylia as the outgroup. Crocodylia Neonithine r 6. Gallus R, americana Tinamidae Causaridae Lithornitidae Phu Phock Neuquen T . formosus B. jaffei D. antirrhopus C. osmolka M. xixiaensis Titanosauridae m 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dinosaurs with no positively associated oologic material yet, and tetanurians. A suite of more inclusive groups in Tetanurae, namely Coelurosauria, Maniraptora, Paraves, and Deinonychosauria are of special interest by their still debated phylogenetic position and because some of those are hypothesized to be closely related to the basal avialans (Sereno, 1999; Clark et al, 2002). The oological data used in this research encompasses some of the clades included within Coelurosauria. Coelurosaurs are generally considered small to medium size predators with an increase stiffening of the caudal vertebrae and an increase of sacral vertebrae (Sereno, 1999). Typical and well-known dinosaurs of this group are the compsognathids from European Lagerstatten and their feathered counterparts from the Chinese Jehol biota (Clark et al., 2002). Maniraptorans are equated to sickle-toed raptors generally including oviraptorids, troodontids, dromaeosaurids, and avialans. The deep-snouted oviraptorosaurs, and therizinosaurs are recovered by most analyses as sister taxa they have been alternatively included and excluded from basal avialans (Chapter 5) by few paleontologists and ornithologists because of some of their features that were erroneously interpreted as avian. Dromaeosaurids, represented in this analysis by D. antirrhopus, are diagnosed by a suite of synapomorphies, the most notorious being a retractable raptorial second pedal digit (Foster et al., 1998). These raptors body size ranged from big to small and some of them were recently discovered in the Chinese lagerstatten of the Jehol biota covered with feathers (Xu et al., 2000; Ji at al., 2001). Their phylogenetic placement is still greatly debated (Foster et al., 1998; Clark et al., 2002) as that of the swift and large-brain troodontids that were alternatively considered sister group to Aves (Foster et al., 1998) or unified with Dromaeosuridae within Deinonychosauria (Clark et al., 2002, Hwang et al., 2002) as a sister taxon to the clade named alternatively Avialans (Gauthier, 1986) or Aves (Foster et al., 1998). An analysis was performed considering only the oological and reproductive characters of the saurischians described in chapters 3 to 11. This analysis included 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12.2B: 50% majority rule consensus tree of 5 MPTs with a Cl of 32 steps after re-weighting the 23 characters by use of the rescaled consistency (Carpenter, 1988). Note the congruence between this tree topology and the character optimization of figure 1 B. tj m Crocodylia G. Gallus R. americana Tinamidae Lithornitidae Causaridae Phu Phock Neuquen T . formosus B. jaffei D. antirrhopus C. osmolka M. xixiaensis Titanosauridae 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 12.3: Consensus tree of 432 MPTs from an analysis of 50 taxa and 242 characters. The original skeletal data is from Clark et al., 2002 and Hwang et al., 2002. 23 oological characters for 14 taxa (13 and 1 out group) were added to the skeletal data subset. Only 4 taxa (D. antirrhopus, C. osmolka, B. jaffei, and T . formosus) are shared by both data subsets. As a result of this bias and also because these shared taxa are derived and terminal species in each clade, no change in tree topology is observed when comparison is made between skeletal and total evidence evolutionary hypotheses. 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. f#«<riirirns. pdyodon Figure 12.3 h S i Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. one outgroup composed of characters synthesized from the extant crocodilians to determine the polarity of character transformations. In addition, a neognath, here a galliform (chickens, pheasants and relatives) considered to be a basal within neornithine phylogeny (Cracraft, 1988) and three paleognath birds were added to the analysis. The majority of the oological characters for the added taxa were taken from Grellet-Tinner’s (submitted) phylogenetic analysis of paleognaths. In total, a matrix of 14 taxa and 23 characters with seven ordered characters (7, 12, 15, 16, 17, 18, 19, and 20) was analyzed using the heuristic search algorithm in PAUP 4.0b4a (Swofford, 2001). Following initial searches, characters were weighed according to the rescaled consistency index (Carpenter, 1988) to determine whether this method would refine the obtained phylogenetic hypotheses by avoiding a priori choices therefore, minimizing researcher bias and choice between alternative and equally parsimonious cladograms based on the most consistent characters (Carpenter, 1988). After choosing Crocodylia as the outgroup, a heuristic search was performed. Forty MPTs with a Cl of 57 were obtained (Fig. 12.2A). Strict and 50% majority rule consensus computing shows different results. Strict consensus offers little resolution within the ingroup (not shown), but majority rule neatly discriminates diverse sub groups within the studied ingroup; the Lithomitidae is placed at the base and within Paleognathae as expected. In turn Paleognathae shares a most recent common ancestor with galliform here representing Neognathae, a topology congruent with our current phylogenetic knowledge of Neornithine (Lee et al., 1997). Phu Phock and Neuquen specimens are grouped together as a sister clade to Neornithine, and the two troodontids are recovered as a distinct clade sister-taxon to the previous groups. C. osmolka, D. antirrhopus, and M. xixiaensis, form a polytomic sister-group at the base of all the other clades beside Titanosaurids that stand outside as a basal saurischian. Re-weighting the characters by use of the rescaled consistency (Carpenter, 1988) index generated 5 MPTs each of 32 steps in length. The strict consensus 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. representation of these trees indicates somewhat more resolution within the ingroup (Fig. 12.2B); Lithornitidae and tinamidae share a common ancestor, and are placed within Palaeognathae as the sister-group of R. Americana, sharing a most recent common ancestor with Casuaridae in this analysis. Although, the rest of the topological configuration of this consensus tree is nearly identical in much respect to the previous one, D. antirrhopus now becomes a sister-taxon to the C, osmolka and M. xixiaensis group, thus offering a greater resolution within Maniraptora. This tree topology conforms to that predicted by character optimization, thus supporting the evolutionary cline and trends of the oological and reproductive characters of the studied taxa. This evolutionary hypothesis compares favorably with the existing phylogeny based upon skeletal data by Foster at al. (1998) but differs in one aspect from the latest hypothesis (Hwang et al., 2002) by the placement of troodontids closer to Avialans than dromaeosaurids (Fig. 12.3). Under the concept of total evidence, the following cladistic analyses incorporate the oological characters for each taxon examined in this dissertation with the already published skeletal characters in Clark et al., (2002) based on the successive studies ofNorell et al. (2001) and Hwang et al. (2004). Four taxa and 32 characters, 23 of those are oological/reproductive characters, were added to this early work. As the number of available associated and identifiable specimens is still limited, the oological observation in this dissertation applies only to a fraction of the 50 studied taxa. The character states of the OTUs of which the oological material was not available were coded with a question mark “?”. The analyses were conducted in the computer program NONA (Goloboff, 1999) using heuristic search methods. One thousand replicates of tree bisection and regraphing algorithm were implemented retaining only the ten shortest trees for each replicate. These retained trees were then subjected to branch swapping extended to cladograms up to 10% longer. A strict consensus search was performed for the many equally parsimonious cladograms resulted from this 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analysis. No substancial topoligcal changes stemed out of the combined analysis. As previously noted in Hwang et al (2002) troodontids share a recent common ancestor with dromaeosaurids forming a clade that is in turn sister taxon to Avialae. The lack of congruence between the result of the oological and the modified Hwang et al. (2002) analyses at the level of Troodontidae and Dromaeosauridae could stem from a few reasons, namely homoplastic characters in either or both data subsets, an overwhelming number of taxa with missing oological characters, and oological characters applied to terminal and very derived species in each group. Although each bias could contibute and add to to topological incongruence, the main issue is likely that the oological/reproductive characters are associated with derived taxa of each clade coupled with the fact that the analysis includes basal taxa within these clades. D. antirrhopus is part of a terminal big polytomy in Dromaeosauridae and B. jaffei as well as T . formosus are derived troodontids. Considering the overall phylogenetic congruence between data subsets including the minor topologic difference in Paraves attributed to a sampling bias in regard to terminal and derived taxa, there is an undeniable body of evidence that support the integration of reproductive and oological characters as data subset or total evidence in phylogenetic analyses. CONCLUSION The questions initially raised in this research were whether detailed oological observations could bring to the phylogenetic debate a better understanding of the relationship of saurischian dinosaurs, to test whether a particular group of non-avian theropods is closely related to birds, and whether oological data could, by itself or in complement to established knowledge, bring novel information about the reproductive behaviors of these dinosaurs, possibly their physiology, and perhaps their metabolism. This research based on eight well-identified non-avian and primitive avialans demonstrated that, as the feathers of modem birds, most of the oological characters^ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and reproductive behaviours associated with modern birds are rooted among non-avian theropods. Although the studied taxa in this work represent only a fraction of the saurischian dinosaur, they clearly reveal reproductive and oological trends common to all these groups; Namely, a reproductive evolutionary cline from crocodilians to modem birds, and a noticeable pattern of coeval development between the acretion of eggshell layers, presence of larger air cells, the reproductive organs, and brooding/ incubating behaviours. Most of these pre-adaptations are grouped in two main nodes: One at the level of Oviraptorosauridae, and the other at Troodontidae. Although that undeniably these two theropods taxa represent important steps in the evolution of avian reproduction, the phylogenetic distance between Oviraptorosauridae and Titanosauria, for instance, cannot be ignored and likely indicates that the reproductive features that appeared in block in oviraptors might have evolved independently across more basal theropods clades. Likewise, although troodontids seem in this analysis the obvious precursors of modem avian reproduction, the importance of small-bodied theropods such as those who laid the Phu Phock eggs cannot be dismissed and the eggs of such dinosaurs advocates their closer phylogenetic ties to Aves than troodonts. A heuristic search was performed to analyze the distribution of the 23 oological and reproductive characters studied in 14 taxa. The strict consensus of five MPTs each of 32 steps in length indicates that Lithornitidae and tinamidae share a common ancestor, D. antirrhopus is a sister-taxon to the C. osmolka and M. xixiaensis group, and the two troodontids are closer to Avialae and in this analysis to Enantiomithine than other non-avian theropods. This evolutionary hypothesis compares favorably with an existing phylogeny based upon skeletal data by Foster at al. (1998) but differs in one aspect from the latest saurischian evolutionary hypothesis (Hwang et al., 2002) by the placement of troodontids closer to Avialans than dromaeosaurids (Fig. 3). Under the concept of total evidence, a new cladistic analysis was run including the 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. oological characters for each taxon examined in this dissertation with 46 taxa and 208 skeletal characters. No substantial topological changes resulted from the combined analysis. Troodontids share a recent common ancestor with dromaeosaurids forming a clade that is in turn sister taxon to Avialae as in Hwang et al. (2002). The only disparity between the result of the oological and the total evidence analyses (at the level of Troodontidae and Dromaeosauridae) stems from the fact that the oological/ reproductive characters are associated in this research with derived, terminal taxa of each clade coupled with the fact that the total evidence analysis includes basal taxa within these clades. Nevertheless, considering the overall phylogenetic congruence between skeletal and oological data subsets including the minor topologic difference in Paraves attributed to a sampling bias in regard to terminal and derived taxa, there is an undeniable body of evidence that supports the integration of reproductive and oological characters as data subset or in total evidence in phylogenetic analyses to better understand the saurischian evolution and the origin of birds. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES CITED Baicich, P. J., and C. J. O. Harrison. 1997. Pp. 20-21 in A Guide to the Nests, Eggs, and Nestlings of North American Birds. Second Edition. Academic Press. San Diego. Bakker, R. T., M. Williams, and P. J. Currie. 1988. Nanotyranus, a new genus of pygmy tyrannosaur, from the latest Cretaceous of Montana. Hunteria 1: 1-30. Bledsoe, A. H. 1988. A phylogenetic analysis of postcranial skeletal characters of the ratite birds. Annals of the Carnegie Museum 57: 73-90. Board, G. B., andN. H. C. Sparks. 1995. Avian Eggshell; pp 71-86 in Egg Incubation: its Effects on Embryonic Development in Birds and Reptiles (C. D., Deeming, and M. W. J. Ferguson, eds.). Cambridge University Press, New York. Bonaparte, J. F., and M. Vince. 1979. El hallazgo del primer nido de dinosaurio triasicos, (Saurischia, Prosauropoda), Triasico Superior de Patagonia, Argentina. Ameghiniana 16: 173-182 Bottjer, D. J., Etter, W., Hagadom, J. W., and C. M. Tang (eds.). 2002. Fossil- Lagerstatten: Jewels of the Fossil Record; in Exceptional Fossil Preservation. Cambridge University Press, New York. Bravo, A. M., J. J. Moratalla, J. V . Santafe, and C. de Santisteban. 2000. Faide 1 1 a , a new Upper Cretaceous nesting site from the Tremp basin (Lerida Province, Spain): pp 15-22 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Breton, G., R. Foumiet, and J-P. Watte. 1986. Le lieu de ponte de dinosaures de Rennes-le-Chateau (Aude): Premiers resultats de la campagne de fouilles 1984. Annales du Museum du Havre 32: 1-13. 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brinkman, D. L., R. L. Cifelli, and N. J. Czaplewski. 1998. First Occurence of Deinonychus antirrhopus (Dinosaurua: Theropoda) from the Antlers Formation (Lower Cretaceous: Aptian-Albian) of Oklahoma. Oklahoma Geological Survey 144: 1-27. Buffetaut, E. and, R. Ingavat. 1984. A very small theropod dinosaur from the Upper Jurassic of northeastern Thailand. Comptes rendus de l’Academie desScience de Paris 298: 915-918. Buffetaut, E., and J. Le Loeuff. 1994. The discovery of dinosaur eggshells in nieteenth-century France: pp. 31-34 in Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Homer, eds.). Cambridge University Press. New York. Buffetaut, E., Suteethom, V., Tong, H., Cuny, G. and, L. Cavin. 2003. A pterodactyloid tooth from the Sao Khua Formation (Early Cretaceous) of Thailand. Mahasarakha University Journal 22 (special issue): 92-98. Calvo, J. O., S. Engelland, S. E. Heredia, and L. Salgado. 1997. First record of dinosaur eggshells (? Sauropoda-Megaloolithidae) from Neuquem, Patagonia, Argentina. GAIA 14: 23-32. Carpenter, J. M. 1988. Choosing among multiple equally parsimonious cladograms. Cladistics, 4: 291-296 Carpenter, K. 1999. Eggs, Nests, and Baby Dinosaurs (K. Carpenter ed.). Indiana University Press. 336 p. Chen, P-J, Z-M. Dong, and S-N. Zhen. 1998. An exceptionally well-preserved theropod dinosaur from the Yixian Formation of China. Nature 391: 147-152. Chiappe, L. M., Coria, R. A., Dingus, L., Jackson, F., Chinsamy, A., and M. Fox. 1998. Sauropod dinosaur embryos from the late Cretaceous of Patagonia. Nature 396: 258-261. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chiappe, L. M., Dingus, L., Jackson, F., Grellet-Tinner, G., Aspinal, R., Clark, J., Coria, R., Garrido, A., and D. Loope. 2000. Sauropod eggs and embryos from the Late Cretaceous of Patagonia; pp 23-29 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia, Spain. Chiappe, L. M, L. Dingus, F. Jackson, G. Grellet-Tinner, and R. Coria. 1999. Auca Mahuevo, an extraordinary dinosaur nesting ground from the Late Cretaceous of Patagonia. Journal of Vertebrate Paleontology 19 (Supplement to Number 3): 37A. Chiappe, L. M., and L. Dingus. 2001. Walking on Eggs: the Astonishing Discovery of Thousands of Dinosaur Eggs in the Badlands of Patagonia. Scribner New York, 219 pp. Chiappe, L. M., Salgado, L., and R. A. Coria. 2001. Embryonic skulls of titanosaur sauropod dinosaurs. Science 293: 2444-2446. Chiappe, L. M, M. Norell, and J. Clark. 2001. Anew skull of Gobipteryx minuta (Aves: Enantiornithines) from the Cretaceous of the Gobi Desert. American Museum Novitates 3346: 1-15. Chiappe, L. M., Schmitt, J. G., Jackson, F. D., Garrido, A., Dingus, L., and G. Grellet- Tinner. 2004. Sedimentary criteria for recognition of dinosaur nesting traces. Palaios 19: 89-95. Clark, J. M., M. A. Norell, and L. M. Chiappe. 1999. An oviraptorid skeleton from the Late Cretaceous of Ukhaa Tolgod, Mongolia, preserved in an avianlike brooding position over an oviraptorid nest. American Museum Novitates 3265: 1-36. Clark, J. M., M. A. Norell, and R. Barsbold. 2001. Two new oviraptorids (Theropoda: Oviraptorosauria), Upper Cretaceous Djadokhta Formation, Ukhaa Tolgod, Mongolia. Journal of Vertebrate Paleontology 21: 209-213. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cousin, R. 1997. Les gisements d’oeufs de dinosauriens des Hautes Corbieres et des Corbieres Orientales (Aude): Ponte, nidification, microstructre des coquilles. Bulletin de la Societe d’Etudes Scientifiques de F Aude XCVII: 29-46. Cousin, R., and G. Breton. 2000. A precise and complete excavation is necessary to demonstrate a dinosaur clutch structure: pp 31-42 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Cracraft, J. 1988. The major clades of birds:p p. 333-355 in The Phylogeny and Classification of the Tetrapods Volume 1: Amphibians, Reptiles, Birds (M. J. Benton ed.). Systematics Association: Oxford. Cracraft, J. 2001. Avian evolution, Gondwana biogeography and the Cretaceous- Tertiary mass extinction event. Proceedings of the Royal Society of London, Series B (Biological Sciences) 268:459-469. Currie, P . J. 1996. The great dinosaur egg hunt. National Geographic Magazine 189: 96-111. Currie, P. J., and J. R. Homer. 1988. Lambeosaurine hadrosaur embryos (Reptilia: Omithischia). Journal of Vertebrate Paleontology 8 (Supplement to Number 3): 13A. Currie, P . J. and, P. Chen. 2001. Anatomy of Sinosauropteryx prima from Liaoning, northeastern China. Canadian Journal of Earth Siences 38: 1705-1727. Czerkas, S. and, C. Yuan. 2002. An arboreal maniraptoran from northeast China. Dinosaur Museum Journal 1: 63-95. Dauphin, Y . 1990. Microstructures et composition chimique des coquilles d’oeufs d’oiseaux et de reptiles. Palaeontographica Abteilung 214 (A): 1-12. Deeming, C. D., and M. W . J. Ferguson, eds. 1995. Egg Incubation: its Effects on Embryonic Development in Birds and Reptiles. Cambridge University Press. New York. 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dequeiroz, A., J. Kim, and M. J. Donoghue. 1995. Separate versus combined analysis of phylogenetic evidence. Annual Review of Ecology and Systematics 26: 657-681. Dingus, L., Clark, J., Scott, G. R., Swisher, C. C., Chiappe, L. M., and R. A. Coria. 2000. Stratigraphy and magnetostratigraphic/faunal constraints for the age of sauropod embryo-bearing rocks in the Neuquen Group (Late Cretaceous, Neuquen Province, Argentina). American Museum Novitates 3290:1-11. Dong Z.-M. and P. J. Currie. 1996. On the discovery of an oviraptorid skeleton over a nest of eggs at Bay an Mandahu, Inner Mongolia, People’s Republic of China. Canadian Journal of Earth Sciences 33: 631-636. Dyke, G. J. 2003. The fossil record and molecular clocks: basal radiations within Neornithes, p. xx-xx. In P. Smith and P. Donoghue (eds.). Molecular Clocks and the Fossil Record. Taylor and Francis: London (in press). Erben, H. 1970. Ultrastrukturen und Mineralisation rezenter und fossiler Eischalen bei Vogeln und Reptilien. Biomineralisation Forschungsberichte 1: 1-66. Faccio, G. 1994. Dinosaurian Eggs from the Upper Cretaceous of Urugay: pp. 47-55 in Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Homer, eds.). Cambridge University Press. New York. Fang, X., L. Lu, Z. Cheng, Y . Zou, Q. Pang, Y . Wang, K. Chen, Z. Yin, X. Wang, J. Liu, H. Xie, and Y. Jin. 1994. On the Cretaceous Fossil Eggs of Xixia County, Henan Province. Geological Publishing House. Beijing. Folk, R.L. 1994. Interaction between bacteria, nannobacteria, and mineral precipitation in hot springs of central Italy. Geographie Physique et Quaternaire 4:233-246. Folk, R.L., and, F. L. Lynch. 2001. Organic matter, putative nannobacteria and the formation of ooids and hardgrounds. Sedimentology 48:215-229. 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Forster, C. A. 1990. The postcranial skeleton of the omithopod dinosaur Tenontosaurus tilletti. J. Vert. Paleontol. 10: 273-294. Forster, C. A, S. D. Sampson, L. M. Chiappe, and D. W. Krause. 1998. The theropod ancestry of birds: new evidence from the Late Cretaceous of Madagascar. Science 279: 1915-1919. Gallenkamp, C. 2001. Dragon Hunter Roy Chapman Andrews and the Central Asiatic Expeditions. Penguin Putnam inc Press. Gauthier, J. 1986. Saurischian monophyly and the origin of birds: in The Origin of Birds and the Evolution of Flight (K. Padian ed.). Memoirs ofthe California Academy of Sciences 8: 1-55. Gauthier, J. A., A. G. Kluge, and T. Rowe. 1988. Amniote phylogeny and the importance of fossils. Cladistics, 4: 105-209. Grant, G. S. 1982. Avian incubation: egg temperature, nest humidity, and behavioral thermoregulation in a hot environment. Ornithological Monographs 30: 1-75. Grellet-Tinner, G. 2000a. Phylogenetic interpretation of eggs and eggshells; pp 61-75 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia, Spain. Grellet-Tinner, G. 2000b. A study of titanosaurid eggshell from Auca Mahuevo. Journal of Vertebrate Paleontology 20 (Supplement to Number 3): 46A. Grellet-Tinner, G. 2001. Phylogenetic interpretation of eggs and eggshells and its implication for Paleognathae phylogeny. Unpublished thesis. Department of Geological Sciences at the University of Texas, Austin. 158pp. Grellet-Tinner, G., and L. M. Chiappe. 2004. Dinosaur eggs and nestings: implications for understanding the origin of birds: pp 185-214 in Feathered dragons: Studies on the Transition from dinosaurs to Birds (J. P. Currie, E. B. Koppelhus, M. A. Shugar, and J. L. Wright eds.). 201 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Grellet-Tinner, G., L. Chiappe and R. Coria. 2004. Eggs of titanosaurid sauropods from the Upper Cretaceous of Auca Mahuevo (Argentina). Canadian Journal of Earth Sciences, 41: 949-960. Grellet-Tinner, G. In press. The membrana testacea of titanosaurid dinosaur eggs from Auca Mahuevo (Argentina): Implications for the exceptional preservation of soft tissue in Lagerstatten. Journal of Vertebrate Paleontology. Grellet-Tinner, G. Submitted. Phylogenetic interpretation of eggs and eggshells and its implications for Paleognathae Phylogeny. Alcheringa. Grellet-Tinner, G., and G. Dyke. Submitted. Re-evaluating the phylogenetic posiiton of the fossil bird Lithomis : combining character data from osteology and oology. Novitates. Grellet-Tinner, G., and Norell, M. 2002. An avian egg from the Campanian of Bayn Dzak, Mongolia. Journal of Vertebrate Paleontology, 17: 431-434 Griffiths, P. J. 1993. The question of Compsognathus eggs. Revue de Palebiology 7: 85-94. Griffiths, P. J. 2000. Compsognathus eggs revisited: pp 77-83 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Harrison, C. J. O., and C. A. Walker. 1977. Birds of the British Lower Eocene. Tertiary Research Special Paper 3: 1-52. Hayward, J. M., C. J. Amlaner, and K. A Young. 1989. Turning eggs to fossils: a natural experiment in taphanomy. Journal of Vertebrate Paleontology 9: 196- 200 . Hayward, J. M., Hirsch, K. F., and T. C. Robertson. 1991. Rapid dissolution of avian eggshells buried by Mount St. Helens ash. Palaios 6: 174-178. 202 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Hayward, J. M., Folsom, S. D., Elmendorf, D. L., Tambrini, A. A., and D. L. Cowles. 1997. Experiments on the taphonomy of amniote eggs in marine environments. Palaios 12: 482-488. Hayward, J. M L, Zelenitsky, D. K., Smith, D. L., Zaft, D. M., and J. K. Claybum. 2000. Eggshell taphonomy at modem gull colonies and a dinosaur clutch site. Palaios 15: 343-355. Hirsch, K. F. 1989. Interpretation of Cretaceous and Pre-Cretaceous eggs and Eggshells: pp. 89-97 in Dinosaur tracks and Traces (D. Gilette, and M. Lockley, eds.). Cambridge University Press. New York. Hirsch, K. F. 1994. Upper Jurassic eggshells from the Western Interior of North America: in K. Carpenter, K. F. Hirsch, and J. R. Homer (eds.). Dinosaur Eggs and Babies, pp. 137-150. Cambridge Univ. Press, Cambridge. Hirsch, K. F. 1996. Parataxonomic classification of fossil chelonian and gecko eggs. Journal of Vertebrate Paleontology 16: 752-762. Hirsch, K. F., and J. Bowles. 1979. Early Eocene crane-like eggs?: pp 211-216 in Proceedings of the 1978 Crane Workshop (J. C. Lewis ed.). Colorado State University Printing Service. Fort Collins. Hirsch, K. F., and L. F. Lopez-Jurado. 1987. Pliocene chelonian fossil eggs from Gran Canaria, Canary islands. Journal of Vertebrate Paleontology 7: 96-99. Hirsch, K. F, and B. Quinn. 1990. Eggs and eggshell fragments from the Upper Cretaceous Two Medicine Formation of Montana. Journal of Vertebrate Paleontology 10: 491-511. Holtz, T. R. Jr. 1994. The phylogenetic position of the Tyrannosauridae: implications for theropod systematics. J. Paleontol. 68(5): 1100-1117. 203 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Horner, J. R. 1987. Ecological and behavioral implications derived from a dinosaur nesting site: pp 50-63 in Dinosaurs of the Past and Present, Vol. II (S. J. Czerkas and E. C Olson eds.). Natural History Museum of Los Angeles County and University of Washington Press. Horner, J. R. 1997. Rare preservation of an incompletely ossified fossil embryo. Journal of Vertebrate Paleontology, 17: 431-434. Homer, J. R., and P. J. Currie. 1994. Embryonic and Neonatal Morphology and Ontogeny of a New Species of Hypacrosaurus (Omithischia, Lambeosauridae) from Montana and Alberta, p. 312-336. In K. Carpenter, K. F. Hirsch, and J. R. Horner (eds.), Dinosaur Eggs and Babies. Cambridge University Press. New York. Horner, J. R., and D. B. Weishampel. 1988. A comparative embryological study of two omithischian dinosaurs. Nature 332: 256-257. Houde, P. 1986. Ostrich ancestors found in the Northern Hemisphere suggest new hypothesis for ratite origins. Nature 324: 563-565. Houde, P. 1988. Paleognathous birds from the Tertiary of the northern hemisphere. Publications of the Nuttall Ornithological Club 22: 1-148. Ji, Q., P . J. Currie, M. Norell, and Ji, S.-A. 1998. Tow feathered theropods from the Upper Jurassic/Lower Cretaceous strata of northeastern China. Nature 393: 753-761. Kamat, S., X. Su, R. Ballarini, and A. H. Heuer. 2000. Structural basis for the fracture toughness of the shell of the conch Strombus gigas. Nature 405: 1036-1040. Kerourio, P. 1981. Nouvelles observations sur le mode de nidification et de ponte chez les dinosauriens du Cretace terminal du Midi de la France. Comptes Rendus Sommaires de la Societe Geologique de France 1: 25-28. 204 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Kerourio, P., and B. Sige. 1984. L’apport des coquilles d’oeufs de dinosaures de Laguna Umayo a l’age de la Formation Vilquechico et a la comprehension de Perutherium altiplanense. Newsletters on Stratigraphy 13: 133-142. Kohring, R. R. 1999. Calcified shell membranes in fossil vertebrates eggshell: Evidence for preburial diagenesis. Journal of Vertebrate Paleontology 19 : 723-727. Kohring, R. R. 1999. Strukturen, Biostratinomie, systematische und phylogenetische Relevanz von Eischalen amnioter Wirbeltiere. Courier Forschunginstitut Senckenberg, 210: 1-307. Kohring, R. R., and O. Sachs. 1997. Erhaltungsbedingungen und diagenese fossiler Vogeleischalen aus dem Nordlinger Ries (Miozan, MN6). Archaeopteryx 15: 73-96. Kolesnikov, C. M., and A. V . Sochava. 1972. A paleobiochemical study of the Cretaceous dinosaur eggshell from the Gobi. Paleontological Journal 2: 235- 245. Lee, K., J. Felsenstein, and J. Cracraft. 1997. The phylogeny of ratite birds: resolving conflicts between molecular and morphological data sets:p p. 173-208 in Avian Molecular Systematics and Evolution (D.P. Mindell ed.). Academic Press: New Flaven. Li, Y., Yin, Z., and Yu Liu. 1995. The discovery of a new genus of dinosaur eggs from Xixia, Henan, China. Journal of Wuhan Institute of Chemical Technology 17: 38-41. Lowe, P . R. 1931. Struthious remains from northern China and Mongolia; with descriptions of Struthio wimani, Struthio anderssoni, and Struthio mongolicus sp. nov. Palaeontologica Sinica Series C, 6 (4): 3-44. Lydekker, B. A. 1883. Siwalik birds. Memoirs of the Geological Survey of India, Palaeontologica Indica, Calcutta, Series 10 3:135-147. 205 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Maddison, W., M. Donoghue, and D. Maddison. 1984. Outgroup analysis and parsimony. Systematic Zoology 33: 83-103. Magalhaes Ribeiro, C. M. 2000. Microstructural anlysis of dinosaur eggshells from Bauru Basin (Late Cretaceous), Minas Gerais, Brasil: pp 117-122 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Makita, T. 1981. X-Ray microanalysis of the avian shell gland and eggshell. Scanning Electron Microscopy 2: 473-480. Makovicky, R J. and H.-D. Sues. 1998. Anatomy and phylogenetic relationships of the theropod dinosaur Microvenator celer from the Lower Cretaceous of Montana. American Museum Novitates 3240: 27pp. Makovicky, P., and G. Grellet-Tinner. 2000. Association between a specimen of Deinonychus antirrhopus and theropod eggshell: pp 123-128 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Maliva, R. G., Knoll, A. H, and R. Siever. 1989. Secular change in chert distribution: a reflection of evolving biological participation in the silica cycle. Palaios 4: 519-532. Manning, T. W., K. A. Joysey, and A. R. I. Cruickshank. 1997. Observations of microstructures within dinosaur eggs from Henan Province, Peoples’ Republic of China : pp 287-290 in Dinofest International (D. L. Wolberg, E. Stump, and G. D. Rosenberg eds.). The Academy of Natural Sciences. Philadelphia. Maryanska, T., H. Osmolka, and M. Wolsan. 2002. Avialan status for Oviraptorosauria. Acta Palaeontolongica Polonica 47: 97-116. Mateus, I., H. Mateus, M. Telles Antunes, O. Mateus, P. Tacquet, V. Ribeiro, G. Manuppella. 1997. Couvee, oeufs et embryons d’un dinosaure theropode du Jurassique superieur de Lourinha. Comptes Rendus de l’Academie des Sciences de Paris, Sciences de la terre et des planetes 325: 71-78. 206 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Mikhailov, K. E. 1987. Some aspects of the structure of the shell of the egg. Palaeontological Journal 3: 54-61. Mikhailov, K. E. 1991. Classification of fossil eggshells of amniotic vertebrates. Acta Palaeontologica Polonica 36: 193 Mikhailov, K. E. 1992. The microstructure of avian and dinosaurian eggshell: phylogenetic implications: pp 4-23 in Papers in avian paleontology honoring Pierce Brodkorb (K. E. Campbell, Jr. ed.). Natural History Museum of Los Angeles County 36. Mikhailov, K. E. 1997. Fossil and recent eggshell in amniotic vertebrates: fine structure,comparative morphology and classification. Special Papers in Palaeontology 56: 1-80. Mohabey, D. M. 2000. Indian Upper Cretaceous ( Maastrichtian) dinosaur eggs: their parataxonomy and implication in understanding the nesting behavior: pp 139- 154 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Moratalla, J. J., and J. E. Powell. 1994. Dinosaur Nesting Patterns: pp. 37-46 in Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Homer, eds.). Cambridge University Press. New York. Mueller-Towe, J. Inken, P.M. Sander, H. Schuller, and D. Thies. 2002. Hatching and infilling of dinosaur eggs as by computed tomography. Schweizerbartlische Verlagsbuchhandlung D-70176: 119-168. Stuttgart Nealson, K. H. 1997. Sediment bacteria: who’s there, what are they doing, and what’s new? Annual Reviews of Earth Planetary Sciences 25: 403-434. Niemow, J. A., and E. I. Friedmann. 1993. Terrestrial Lithophytic (Rock) Communities; pp. 343-412 in Antartic Microbiology. Wiley-Liss, Inc. Press. 207 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Noble, R. C. 1995. Comparative Composition and Utilisation of Yolk Lipid by Embryonic Birds and Reptiles; pp. 17-28 in Egg Incubation: Its Effects on Embryonic Development in Birds and Reptiles (D. C. Deeming, and M. W. J. Ferguson, eds.). Cambridge University Press, New York. Norell, M. A, J. M. Clark, D. Dashzeveg, R. Barsbold, L. M. Chiappe, A. R. Davidson, M. C. McKenna, P. Altangerel, and M. J. Novacek. 1994. A theropod dinosaur embryo and the affinities of the Flaming Cliffs dinosaur eggs. Science 266: 779- 782. Norell, M. A, J. M. Clark, L. M. Chiappe, and D. Dashzeveg. 1995. A nesting dinosaur. Nature 378: 774-776. Norell, M. A, and P. J. Makovicky. 1997. Important features of the dromaeosaur skeleton: information from a new specimen. American Museum Novitates 3215: 28pp. Norell, M. A., J. M. Clark, and L. M. Chiappe. 2001. An embryonic oviraptorid (Dinosauria: Theropoda) from the Upper Cretaceous of Mongolia. American Museum Novitates 3325: 1-17. Osborn, H. F. 1924. Three new Theropoda, Protoceratops zone, central Mongolia. American Museum Novitates 144: 1-12. Ostrom, J. H. 1969. Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bull. Peabody Museum of Natural History 30: 1-1. Ostrom, J. H. 1970. Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn Basin area, Wyoming and Montana. Bull. Peabody Museum of Natural History 35: 234. Ostrom, J. H. 1976. Archaeopteryx and the origin of birds. Biological Journal of the Linnean Society, London 8: 91-182. 208 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Ostrom, J. H. 1976. On a new specimen of the Lower Cretaceous theropod dinosaur Deinonychus antirrhopus. Breviora 439: 1 - 21. Ostrom, J. H. 1976. On a new specimen of the Lower Cretaceous theropod dinosaur Deinonychus antirrhopus. Breviora 439: 1 -21 Owen, R. 1840. Description of the remains of a mammal, a bird, and a serpent from the London Clay. Proceedings ofthe Geological Society of London, 3:162-166. Owen, R. 1841. Description of the fossil remains of a mammal (Hyracotherium leporinum) and of a bird {Lithornis vulturinus) from the London Clay. Transactions of the Geological Society of London, 6:203-208 .vulturinus). Owen, R. 1846. A history of British fossil mammals and birds. London, John Van Voorst. Packard, G. C., Taigen, T. L., Packard, M. J., and R. D. Shuman. 1979. Water-vapor conductance of testudinian and crocodilian eggs (class Reptilia). Respiratory Physiology 38: 1-10. Packard, M. J., and DeMarco, V. G. 1995. Eggshell ultrastructure and formation in eggs of oviparous reptiles; pp. 53-69 in Egg Incubation: Its Effects on Embryonic Development in Birds and Reptiles (D. C. Deeming, and M. W. J. Ferguson, eds.). Cambridge University Press, New York. Padian, K., and L. M Chiappe. 1998. The origin and early evolution of birds. Biological Reviews 73: 1-42. Paganelli, C.V. 1980. The physic of gas exchange across the avian eggshell. American Zoologist 20: 329-338. Palmer, B. D., and Guillette Jr., L. J. 1995. Oviductal Proteins and their Influence on Embryonic Development in Birds and Reptiles; pp. 29-46 in Egg Incubation: Its Effects on Embryonic Development in Birds and Reptiles (D. C. Deeming, and M. W. J. Ferguson, eds.). Cambridge University Press, New York. 20( Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Parker, T., and W. Haswell. 1910. A Text-book of Zoology. McMillan and Co. London. Peitz, C. 2000. Megaloolithid dinosaur eggs from the Maastrichtian of catalunya (NE-Spain)-parataxonomic implications and stratigraphic utility: pp 155-160 in First International Symposium on Dinosaur Eggs and Babies (A.M. Bravo, and T. Reyes, eds). Isona I Conca Della Catalonia. Spain. Pooley, A. S. 1979. Ultrastructural relationships of mineral and organic matter in avian eggshells. Scanning Electron Microscopy 2: 475-482. Powell, J. E. 1992. Hallazgo de huevos asignables a dinosaurios titanosauridos (Saurischia, Sauropoda) de la Provincia de Rio Negro, Argentina. Acta Zoologica. Lilloana 41: 381-389. Proctor, N., and P. Lynch. 1993. Manual of Ornithology, Avian Structure and Function. Yale University Press. New Haven. Racey A., Love M. A., Canham A. C., Goodall J. G. S, Polachan, S. AND, P. D. Jones. (1996). Stratigraphy and reservoir potential of the Mesozoic Khorat Group, NE Thailand, Parti. Stratigraphy and sedimentary evolution. Journal of Petroleum Geology 19: 5-40. Rahn, A., Paganelli, C. V. and, A. Ar. 1975. Relation of avian egg weight to body weight. The Auk 92: 750-765. Rahn, H., and A. Ar. 1980. Gas exchange of the avian egg: time, structure, and function. American Zoologist 20: 477-484. Reid, R. P., Visscher, P. T., Decho, A. W., Stolz, J. F. Bebout, B. M., Dupraz, C., Macyntyre, I. G., Paerl, H. W., Pinckney, J. L., Prufert-Bebout, L., Steppe, T. F., and DesMarais, D. J. 2000. The role of microbes in accretion, lamination and early lithification of modem marine stromatolites. Nature 406: 989-992. 210 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Romanoff, A., and A. Romanoff. 1949. The avian egg (John Wiley and Sons Inc., eds.). New York. Ricqles, A. de, Mateus, O., Telles Antunes, M., and R Taquet. 2001. Histomorphogenesis of embryos of Upper Jurassic theropods from Lourinha (Portugal). Comptes Rendus de l’Academie des Sciences de Paris, Sciences de la terre et des planetes 332: 647-656. Sabath, K. 1991. Upper Cretaceous amniotic eggs from the Gobi Desert. Acta Palaeontologica Polonica 36: 151-192. Sahni, A., Tandon, S. K., Jolly, A., Bajpai, S., Sood, A., and Srinivasan, S. 1994. Upper Cretaceous dinosaurs eggs and nesting sites from the Deccan Volcan Sedimentary of Peninsular India. In Dinosaur Eggs and Babies. Edited by K. Carpenter, K. F. Hirsch, and J. R. Homer. Cambridge University Press, New York, pp. 204-226, Sanchez-Villagr a, M. R., and B. A. Williams. 1998. Levels of homoplasy in the evolution of the mammalian skeleton. Journal of Mammalian Evolution 2: 113-125. Schleich, H., and W. Kastle. 1988. Reptile Egg-Shells SEM Atlas. Gustav Fisher Verlag. Stuttgart. Schweitzer, M., H., F. Jackson, L. M. Chiappe, J. G. Schmitt, J. O. Calvo, and D. E. Rubilar. 2002. Late Cretaceous avian eggs with embryos from Argentina. Journal of Vertebrate Paleontology 22: 191-195. Sereno, P . C. 1997. The origin and evolution of dinosaurs. Annual Revue of Earth and Planetary Sciences. 25: 435-489. Sereno, P. C. 1999. The evolution of dinosaurs. Science 284: 2137-2147. 211 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Sibley, C. G., and J. E. Ahlquist. 1990. Phylogeny and Classification of Birds. A Study in Molecular Evolution. Yale University Press. New Haven. Sige, B. 1968. Dents de micromammiferes et fragments de coquilles d’oeufs de dinosauriens dans la faune de vertebres du Cretace superieure de Laguna Umayo 9Andes peruviennes). Comptes rendus de l’Academie des Sciences, Paris (D) 267: 1495-1498. Sochava, A. V. 1969. Dinosaur eggs from the Upper Cretaceous of the Gobi Desert. Paleontological Journal 3: 517-527. Stal, L. J. 2000. Cyanobacteria mats and stromatolites; pp. 61-120 in The ecology of cyanobacteria. Their diversity in time and space (B. A. Whitton and M. Potts eds.). Kluwer academic Publishers, Netherlands. Stal, L. J., and P. Caumette. 1994. Microbial mats. Structure, Development and Environmental Significance. Springer-Verlag NATO, Germany. Swofford, D. W. 2001. PAUP (Version 4.0b4a)—Phylogenetic Analysis Using Parsimony. Sinaur Associates, Washington D C. Taylor, T. 1970. How an eggshell is made. Scientific American 222: 88-95. Varricchio, D. J., F. Jackson, J. J. Borkowski, and J. R. Homer. 1997. Nest and egg clutches of the dinosaur Troodon formosus and the evolution of avian reproductive traits. Nature 385: 247-250. Varrichio, D. J., F. Jackson, J. J. Borkowski, and C. N. Trueman. 1999. A nesting trace with eggs for the Cretaceous theropod dinosaur Troodon formosus. Journal of Vertebrate Paleontology 19: 91-100. Varricchio, D.J., Homer, J.R.and F. D. Jackson. 2002. Embryos and eggs for the Cretaceous theropod dinosaur Troodon formosus. Journal of Vertebrate Paleontology 22: 564-576. 212 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Vianey-Liaud, M L, Malian, P., Buscail, O., and C. Montgelard. 1994. Review of French Dinosaur Eggshells: Morphology, Structure, Mineral, and Organic Composition; pp. 151-183 in Dinosaur Eggs and Babies (K. Carpenter, K. F . Hirsch, and J. R. Homer, eds.). Cambridge University Press, New York. Vianey-Liaud, M., K. F. Hirsch, A. Sahni, and B. Sige. 1997. Late Cretaceous Peruvian eggshells and their relationships with Laurasian and Eastern Gondwanian material. Geobios 30: 75-90. Vianey-Liaud, M., Khosla, A., and G. Garcia. 2003. Relationships between Europoean and Indian dinsosaur eggs and eggshells of the oofamily Megaloolithidae. Journal of Vertebrate Paleontology 12: 575-585. material. Geobios 30: 75-90. Visscher, P. T., Reid, R. P., and B. M. Bebout. 2000. Microscale observation of sulfate reduction: Correlation of microbial activity with lithified micritic laminae in modern marine stromatolites. Geology 28: 919-922. Walter, M. R., J. Bauld, D. J. Des Marais, and W. J. Schopf. 1992. A General comparison of microbial mats and microbial stromatolites: bridging the gap between the modem and the fossil; pp. 335-338 in The Proterozoic Biosphere: A multidisciplinary approach (W. J. Schopf and C. Klein eds.). Cambridge University Press, New York. Watabe, M., and S. Suzuki. 2000. Results of Hyashibara Museum of Natural Sciences, Mongolian Academy of Sciences, Mongolian Paleontological Center, Joint Paleontological expedition. Hasyashibara Museum of Natural Sciences Research Bulletin 1: 17-25. Wenzel, J. 1992. Behavioral homology and phylogeny. Annual Review of Ecology and Systematics 23: 361-381. White III, H. B. 1995. Maternal Diet, Maternal Proteins and Egg Quality; pp. 1-15 in Egg Incubation: Its Effects on Embryonic Development in Birds and Reptiles (D. C. Deeming, and M. W. J. Ferguson, eds.). Cambridge University Press, New York. 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Williams, D., Seymour, R., and Kerourio, P. 1984. Structure of fossil Dinosaur eggshell from Aix Bassin, France. Palaeogeography, Palaeoclimatology, Palaeoecology 45: 23-37. Winkler, D. W., and F. H. Sheldon. 1993. Evolution of nest construction in swallows (Hirundinidae): A molecular phylogenetic perspective. Proceedings of the National Academy of Sciences 90: 5705-5707. Wilson, J. A., and Sereno. P . C. 1998. Early evolution and higher-level phylogeny of sauropod dinosaurs. Society of Vertebrate Paleontology Memoir 5, 18: 1-67. Witmer, L. M. 1991. Perspectives on avian origins, pp 427-466 in Origins of the higher groups of tetrapods: controversy and consensus (H.-P. Schultze and L. Trueb eds.). Cornell University Press. Ithaca. Xu, X., Zhou, Z. and X. Wang. 2000. The smallest known non-avian theropod dinosaur. Nature 408: 705-708. Xu, X., Zhou, Z., Wang, X., Kuang, X., Zhang, F. and X. Du. 2003. Four-winged dinosaurs from China. Nature 421: 335-340. Zelenitsky, D. and L. V. Hills. 1996. An egg clutch of Prismatoolithus levis oosp. nov. from the Oldman Formation (Upper Cretaceous), Devil’s Coulee, southern Alberta. Canadian Journal of Earth Sciences 33: 1127-1131. Zelenitsky, D., L. V. Hills, and P. J. Currie. 1996. Parataxonomic classification of omithoid eggshell fragments from the Oldman Formation (Judith River Group; Upper Cretaceous), southern Alberta. Canadian Journal of Earth Sciences 33: 1655-1667. Zelenitsky, D., S. P. Modesto, and P. J. Currie. 2002. Bird-like characteristic of troodontid theropod eggshell. Cretaceous research 23: 297-305. 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zelenitsky, D., and M. Vianey-Liaud. 2003. Historical and new perspectives on the classification of fossil eggs. Second International Symposium on Dinosaur Eggs and Babies. Zhang, F., Zhou, Z., Xu, X., and X. Wang. 2002. A juvenile coelurosaurian theropod from China indicates arboreal habits. Naturwissenschaften 89: 394-398. Zhao, Zi-Kui. 1975. The microstructure of dinosaurian eggshells ofNanhsiung, Kwangtung and Guangdong. Vertebrata PalAsiatica 13: 105-117. Zhao, Zi-Kui. 1979. The advancement of researches on the dinosaurian eggs in China. Mesozoic and Cenozoic Red Beds of South China. Beijing Science Publishing: 330-340. Zhao, Zi-Kui 1994. Dinosaur eggs in China: On the structure and evolution of eggshells: pp. 184-203 in Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Homer, eds.). Cambridge University Press. New York. Zhao, Zi-Kui. 2000. Nesting behavior of dinosaurs as interpreted from the Chinese Cretaceous dinosaur eggs. Paleontological Society of Korea Special Publication 4: 115-126. Zhao, Zi-Kui, and S. Ding. 1976. Discovery of the dinosaurian eggshells from Alxa, Ningxia and its strati graphic significance. Vertebrata PalAsiatica 14: 42-44. Zhao, H., and Zi-Kui Zhao. 1998. Dinosaur eggs from Xichuan Basin, Henan Prov ince. Vertebrata PalAsiatica 36: 282-296. Zhao, H, and Zi-Kui Zhao. 1999. A new form of elongatoolithid dinosaur eggs from the Lower Cretaceous Shahai Formation of Heishan, Lioning Province. Verte brata PalAsiatica 37: 278-284. 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix 11.1. Character list. List of characters and character states used in the oological cladistic analysis. The following is a list of 15 characters used in the oological phylogenetic analysis. The primitive state is designated as (0), the derived states as (1, 2, 3, 4) and unknown or missing data as (?). Multitstate characters are unordered and character weighing was made according to the rescaled consistency index (Farris, 1969; Carpenter 1988). 2 General egg shape: (0) totally symmetrical eggs with no broadening at any poles, (1) all other shapes. 3 Egg shape variation: (0) sub elliptical, (1) long sub elliptical, (2) short sub elliptical, (3) oval, (?) character state (0) of character 1. 4 Surficial ornamentation: (0) present, (1) absent, 5 Surficial ornamentation morphology: (0) nodular, (1) linearituberculate, (2) smooth, (3) rugose to slightly rugose, (4) granulated. 6 Pore canal shape: (0) straight to surface bifurcated, (1) oblique to surface, (2) straight to surface, (3) diverticulate, (4) not reaching the surface. 7 Pore aperture shape: (0) round, (1) slit-like. 8 Spherulite morphology: (0) acicular, (1), short bladed (2) long bladed, 9 Layer 3: (0) absent, (1) present. 10 Layer 4: (0) absent, (1) present. 11 Morphology of layer 3: (0) absent, (1) amorphous, (2) blocky crystals, (3) porous. 12 Delimitation between layer 1 and 2: (0) absent, (1) prismatic, (2) aprismatic. 13 Delimitation between layer 2 and 3: (0) absent, (1) prismatic, (2) aprismatic. 14 Delimitation between layer 3 and 4: (0) absent, (1) prismatic, (2) aprismatic. 15 Ratio of Layer land layer 2: (0) layer 2 absent, (1) >0.90, (2) [0.89-0.50], (3) <0.49 16 Ratio of Layer 3 and layer 2: (0) layer 3 absent, (1) >0.10, (2) <0.10. 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix 11.2. Data matrix. Osteological data set including representative fossils of Lithornis; iT denotes missing character coding. See appendix 2 of Lee et al. (1997, 196) for list of characters and Dyke (2003) for further discussion. 00000000011111111112222222222333333333344444444445555555 12345678901234567890123456789012345678901234567890123456 Outgroup 00000000000000000000000000000000000000000000000000000000 Tinamidae 00000000700000100000000000000001000000000000000000000000 Apteryx 1117111120112? 111200000021000001001110011010001000111111 Casuarius 11010111101121111101001011111111111011011110100101100020 Dromaius 11010111101121111101001011111111111011011110100101100020 Struthio 11000111311110111117110111211121121011111121011011100030 Rhea 11170111311110111117110111211121121011111121011011100070 A5204 0700000???? 110????????????????11 ??????? 111020077070? 107? A5303 ??????? 121711 ??????????????????????????????????????????? A33138 077700777771???????????????????????????????????????????? A38935 077700?????????????????????????????????????????????????? A3 8934 ???????????1710????????????????????????????????????????? A5425 7707770121717100? ] ??????????????????????? 11 ????????????? 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix 12.1. Character list. Key: ? = missing data; - character not applicable; O=plesiomorphic character; 1, 2,3, 4, 5, 6= apomorphic character states. 1. Egg geometry: symmetric (0); asymmetric (1). 2. Egg shape: absence of air cell (0); presence of proto air cell (1); fully developed air cell (3). Ordered character states 3. Ratio of the two egg diameters: At or around 36% (0); at or around 60% (1); at or around 85% (2). 4. Ovideposition: en “ masse” (0); monoautochronic (1). 5. Egg parity: absence (0); presence (1). 6. Number of functioning ovaries: two (0); one (1). 7. Clutch or nest morphology (1): none (0); one layer of eggs (1); two or more layers of eggs (1). Ordered character states. 8. Clutch morphology (2): absence of empty space in the center (0); presence of empty space in the center (1). 9. Spatial position of the long axis of the egg to the substrate: none (0); 45 degrees or less (1); 46 to 90 degrees (1). 10. Brooding behavior: none (0); protection (1); possible incubation (2). 11. Eggshell ornamentation: none (0); constant nodular (1); various nodular (2); linearituberculate (3); reticulated (4); mixed linearituberculate and reticulated (5); smooth (6). 12. Number of eggshell layers: One (0); two (1); three (2). Ordered character states. 13. Type of boundary between layers 1 and 2: Prismatic (0); aprismatic (1), aprismatic but intermediate (2). 14. Type of boundary between layers 2 and 3: Prismatic (0); aprismatic (1) 218 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15. Crystal morphology in layer 1: Acicular (0); bladed (1); long bladed, or state intermediate between acicular and bladed (2). Ordered character states. 16. Ratio of total thickness by layer 1: Ratio at or around 3 (0); ratio from 3.3 to 4 (1); ratio from 4.5 to 5 (1); ratio from 1.7 to 2 (2); ratio from 10 and above (3). Ordered character states. 17. Ratio of layers 1 by 2: Ratio at or close to 0.35 (0); ratio at or close to 0.25 (1): ratio at or close to 0.40 (2); ratio at or close to 0.50 (3); ratio at or close to 1.5 (4); ratio at or close to O.l (5). Ordered character states. 18. Ratio of layers 3 by 2: Ratio more than 0.20 but less than 0.6 (0); ratio less than 0.1 (1). Ordered character states. 19. Organic lines in layer 2: uniform repartition (0); one or more subdivisions (1). Ordered character states.. 20. Organic line orientation in layer 2: Horizontal (0); follow eggshell surface (1); opposite to eggshell surface (2). Ordered character states. 21. Pore opening: round (0); elliptical (1); round and elliptical (2). 22. Pore canal orientation: straight (0); oblique (1). 23. Pore canal shape: simple (0); branched (1). Ordered character states. 15. Crystal morphology in layer 1: Acicular (0); bladed (1); long bladed, or state intermediate between acicular and bladed (2). Ordered character states. 16. Ratio of total thickness by layerl: Ratio at or around 3 (0); ratio from 3.3 to 4 (1); ratio from 4.5 to 5 (1); ratio from 1.7 to 2 (2); ratio from 10 and above (3). Ordered character states. 17. Ratio of layers 1 by 2: Ratio at or close to 0.35 (0); ratio at or close to 0.25 (1): ratio at or close to 0.40 (2); ratio at or close to 0.50 (3); ratio at or close to 1.5 (4); ratio at or close to O.l (5). Ordered character states. 18. Ratio of layers 3 by 2: Ratio more than 0.20 but less than 0.6 (0); ratio less than 0.1 (1). Ordered character states. 219 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19. Organic lines in layer 2: uniform repartition (0); one or more subdivisions (1). Ordered character states. 20. Organic line orientation in layer 2: Horizontal (0); follow eggshell surface (1); opposite to eggshell surface (2). Ordered character states. 21. Pore opening: round (0); elliptical (1); round and elliptical (2). 22. Pore canal orientation: straight (0); oblique (1). 23. Pore canal shape: simple (0); branched (1). Ordered character states. 220 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix 12.2. Data matrix TAXON T 3 6-10 11-15 16-20 21-23 Crocodylia ?0-0-" .....0000- " 10— 0 -—0 -0 G. gallus 30-10 02110 llool 01003 R. americana 40-10 03110 00101 01003 ??3 Tinamidae 30-10 04110 00101 01003 ?03 Casuaridae 31311 ' 12112 00111 01003 ??3 Titanosauridae 41010/1 0100- — 0 0000O -0 D. antirrhopus ?1410 1110- 0-0-1 .. ‘ y m i ..... 101 C. osmolka 11210 0110- 0-0-1 12111 Oil T.formosus 20-20 02HT - '1 ■ ■ ■ 11022.. 002 B.jajfei '20-77 0210- 141-1 01022 002 M. xixiaensis 11411 0410- 0-0-1 01111 101 Phu Phock 211?? ?2110 11311 ?1??? 113 JNeuquen 20-?? ?2110 .... 10311.... ?1??? ??3 .... Lithornitidae ?0-?? 04110 01001 01003 ??3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Evolutionary paleoecology and taphonomy of the earliest animals: Evidence from the Neoproterozoic and Cambrian of southwest China
PDF
Characterization of geochemical and lithologic variations in Milankovitch cycles: Green River Formation, Wyoming
PDF
Cyclostratigraphy and chronology of the Albian stage (Piobbico core, Italy)
PDF
Biotic recovery from the end-Permian mass extinction: Analysis of biofabric trends in the Lower Triassic Virgin Limestone, southern Nevada
PDF
Early Jurassic reef eclipse: Paleoecology and sclerochronology of the "Lithiotis" facies bivalves
PDF
Holocene sedimentation in the southern Gulf of California and its climatic implications
PDF
Helicoplacoid echinoderms: Paleoecology of Cambrian soft substrate immobile suspension feeders
PDF
Preservation Of Fossil Fish In The Miocene Monterey Formation Of Southern California
PDF
Grain-size and Fourier grain-shape sorting of ooids from the Lee Stocking Island area, Exuma Cays, Bahamas
PDF
Elemental analysis as a first step towards "following the nitrogen" on Mars
PDF
Integrated geochemical and hydrodynamic modeling of San Diego Bay, California
PDF
Investigation into the tectonic significance of the along strike variations of the Peninsular Ranges batholith, southern and Baja California
PDF
A unified methodology for seismic waveform analysis and inversion
PDF
Fractionation of nitrogen isotopes during early diagenesis
PDF
A tectonic model for the formation of the gridded plains on Guinevere Planitia, Venus: Implications for the thickness of the elastic lithosphere
PDF
Geology and structural evolution of the southern Shadow Mountains, San Bernardino County, California
PDF
Educational development aid and the role of language: A case study of AIT -Danida and the Royal University of Agriculture in Cambodia
PDF
Fourier grain-shape analysis of quartz sand from the Santa Monica Bay Littoral Cell, Southern California
PDF
Age and sex differences in levels of proliferation within the zebra finch telencephalic ventricular zone
PDF
Changing characteristics of deformation, sedimentation, and magmatism as a result of island arc -continent collision
Asset Metadata
Creator
Grellet-Tinner, Gerald
(author)
Core Title
A phylogenetic analysis of oological characters: A case study of saurischian dinosaur relationships and avian evolution
School
Graduate School
Degree
Doctor of Philosophy
Degree Program
Geological Sciences
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Geology,OAI-PMH Harvest,paleontology
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Bottjer, David (
committee member
), Bottjer, Sarah (
committee member
), Chiappe, L. (
committee member
), Corsetti, Frank (
committee member
), Fischer, Alfred G. (
committee member
), Norell, M. (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-331907
Unique identifier
UC11340715
Identifier
3180304.pdf (filename),usctheses-c16-331907 (legacy record id)
Legacy Identifier
3180304.pdf
Dmrecord
331907
Document Type
Dissertation
Rights
Grellet-Tinner, Gerald
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
paleontology