sign in sign up
<<
Home , Fernando Novas, Limusaurus, Megaraptor, Protoavis

1 Theropod Diversity and the Refinement of Avian Characteristics

PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Field Museum, Chicago, USA

Bird origins have been debated ever since Darwin biologists such as Cope favored an evolutionary published his “Origin of Species,” and was subse- relationship between and ornithopod dino- quently challenged on the topic by Sir Richard saurs such as hadrosaurs, based again on a three- Owen, who pointed out the lack of transitional toed foot and a retroverted pubic shaft. While a fossil forms in the evolution of the highly derived variety of ancestors or sister taxa were proposed for avian body plan. Indeed, Owen likely carefully birds, a broad consensus that they were related to selected birds to make his point due to their prevailed until the publication of the many unique traits and physiological features English edition of Heilmann’s (1926) “Origin of such as flight, , bipedality, and a remark- Birds.” Heilmann’s book presented a detailed able respiratory system in which the lungs are study of neontological, embryological, and fossil connected to and ventilated by a complex system evidence, all of which pointed to a theropod of air sacs that pneumatize the skeleton. Within ancestry for birds. Nevertheless, based on the two of this debate, the discovery of the first prevailing assumption of the time that dinosaurs specimens of provided conclusive ancestors had lost their clavicles, their reappear- evidence of avian evolution in the fossil record and ance in birds would violate Dollo’s (1893) law of became the focal point for research and delibera- irreversibility. Heilmann therefore concluded tion on the topic for more than a century. While that the similarities between birds and theropods the fossils of Archaeopteryx provided incontro- were due to convergence, and that birds were vertible evidence for a reptilian origin for birds, derived from more that still opinions varied as to which group of birds retain clavicles. may have originated from. Due to the thoroughness of his book, Following the discovery of the small nonavian Heilmann’s (1926) hypothesis held sway for the theropod in the same limestone next four decades until the discovery and descrip- deposits as Archaeopteryx, Huxley (1868) pre- tion of the mid-sized dromaeosaurid theropod sciently proposed an evolutionary relationships Deinonychus by Ostrom (1969). Ostrom’s detailed between birds and nonavianCOPYRIGHTED theropods based on study MATERIAL of the skeletal anatomy of Deinonychus led shared traits such as three principal, weight- him to recognize derived characters shared bearing in the foot (confirmed by foot-prints), between it and the basal Archaeopteryx a tall ascending process of the astragalus, and (Ostrom, 1976), and to the discovery of a misiden- hollow bones. Other contemporary evolutionary tified specimen of Archaeopteryx (Ostrom, 1970).

Living Dinosaurs: The Evolutionary History of Modern Birds, First Edition. Edited by Gareth Dyke and Gary Kaiser. Ó 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. 10 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Among the new traits that Ostrom mustered A ROADMAP TO THE DINOSAURIAN to revive a theropod ancestry for birds are the HERITAGE OF BIRDS presence of a half-moon-shaped wrist bone that allows the hand to adduct against the forearm as in Despite their radically different body plans, the wing-folding mechanism of living birds, and birds inherited a mosaic of anatomical traits a three-fingered hand with characteristic propor- from various stages of history. A tions between the three metacarpals and phalan- host of discoveries over the past five decades ges. Following Ostrom’s work, progressively more provide a detailed road map to how the highly evidence has been amassed to support this hypoth- specialized avian anatomy was assembled over esis of avian origins, as a plethora of new fossil the evolutionary lineage leading to birds and discoveries continue to blur the morphological demonstrates that many of the traits that are distinction between birds and their closest thero- considered uniquely avian among extant pod relatives. actually arose before the origin of Over the past three decades, widespread adop- birds themselves. While most of our understand- tion of cladistic methodology for establishing ing of where birds fit into the tree of life comes and testing proposed evolutionary relationships fromstudyofhardtissues,dramaticnewdis- has provided the conceptual framework for deci- coveries in the past decade and a half are pro- phering the origin and evolutionary history of viding unprecedented insights on the evolution birds. Gauthier (1986) was the first to apply an of soft tissue systems (Schweitzer et al., 1999; explicit cladistic parsimony analysis of theropod Xu et al., 2001) aspects of physiology (Varricchio relationships to the question of bird origins. In et al., 2002; Varricchio & Jackson, 2004; doing so, he provided the first rigorous test of the Organ et al., 2007), and even behavior (Norell hypothesis of theropod ancestry and set the stage et al., 1995; Varricchio et al., 1997; Xu & Norell, for evaluating the evolutionary history of avian 2005). anatomy, physiology, and behavior in a quantita- That birds are archosaurs and the closest liv- tive framework. Subsequent decades have seen ing relatives to crocodilians has long been a remarkable surge in the discovery of new ther- inferred from shared derived traits such as a opods, including fossil stem birds, with each dis- four-chambered heart and thecodont dentition. covery spawning novel systematic analyses (for Virtually all birds possess an external mandibu- reviews see Weishampel et al., 2004; Norell & lar fenestra, a synapomorphy of Archosauria Xu, 2005; Benton, 2008) and further supporting (Benton, 1990) and fossil avians also possess an the hypothesis that birds are theropod dinosaurs. (Figure 1.1), also considered To date, profound advances have been made in a hallmark of this , although in extant teasing out the evolution of the avian body plan as birds the latter opening is lost or merged with well as correlated physiological features and life the external nares through expansion of the history parameters of modern birds. Likewise, premaxillae and concomitant reduction of our knowledge of these traits in modern birds other preorbital bones. Phylogenetic analyses and their anatomical correlates has been used to have identified numerous derived traits shared yield insight into the biology of nonavian theropod by birds and various hierarchically arranged dinosaurs and to infer whether particular avian subsets of diversity (Figure 1.1). A traits originated before or after the origin of the comprehensive review detailing all of these char- avian lineage itself. Here we provide a general acters is beyond the scope of this chapter, so here overview of the stepwise acquisition of the we concentrate on a select subset of traits and avian body-plan throughout the theropod family evolutionary branching points that are most crit- tree and discuss how the physiology of modern ical to understanding the evolution of the unique birds is being used to reconstruct aspects of dino- avian body plan and its derived physiological and saur biology. functional correlates. Theropod Diversity and the Refinement of Avian Characteristics 11

Fig. 1.1 Guide to avian evolutionary history based on a simplified evolutionary tree of Archosauria. Key traits in avian evolution are mapped onto the with their corresponding position indicated on the skeleton of Archaeop- teryx. Note how traits are distributed throughout the skeleton revealing the mosaic assembly of avian anatomy throughout ornithodiran evolution. See text for more detailed discussion of individual traits. Archaeopteryx recon- struction by M. Donnelly. 12 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

The deepest division among archosaurs is acetabula (Sereno & Arcucci, 1993) reveal that the between the lineages leading to the extant acquisition of a fully open acetabulum occurred in Crocodylia and Aves. Each of these branches also a gradual fashion spanning several branching subtends numerous fossil clades, which were points at the root of the dinosaurian evolutionary dominant faunal elements during the . tree (Figure 1.1). Counterintuitively, birds fall The avian total group (¼ extant plus extinct within the , or “lizard-hipped” branch diversity after the split from the lineage leading of diversity, rather than the , to Crocodylia) is known as Ornithodira and the “bird-hipped” branch. Saurischians are united is characterized principally by the possession of by their possession of hyposphen–hypantrum a mesotarsal ankle (Figure 1.1), in which accessory articulations between the neural arches the articulation between the foot and crus occurs of the trunk vertebrae, which are also marked by between the proximal and distal tarsals and lateral excavations on their neural arches presum- approximates a roller joint, restricting foot ably housing diverticula of the respiratory system. motions to fore–aft swinging without a rota- Hypantrum–hyposphen articulations were lost in tional component. Ornithodirans are also char- a later stage of bird evolution, but some fossil birds acterized by having a clearly defined femoral such as Patagopteryx reveal that this trait was head that is distinctly offset from the femoral present ancestrally. shaft. Besides birds, Ornithodira is comprised Birds form part of the theropod radiation within of extinct dinosaurian subclades, pterosaurs, Saurischia. Progressive reduction of the hand and suite of lesser-known forms. toward the tridactyl condition in birds is encoun- Many of these taxa, especially those within the tered near the base of the theropod radiation. When dinosauromorph clade, share the derived trait present in basal theropods, the fourth is of being bipedal, thus freeing the forelimbs to clawless and the fifth digit, known in , evolve new functions including flight. Although is reduced to a metacarpal splint devoid of many large, herbivorous dinosaurs later re- knuckles (Figure 1.2). Coelophysoids and more evolved a quadrupedal stance in concert with derived theropods (Figure 1.1) display a reduction the evolution of large body mass, all derive in the number of carpals to five or less, and the from bipedal ancestors. number of digits to four or less, although a fifth Dinosauromorphs, including birds, are united metacarpal has been tentatively identified in in their possession of three principal weight-bear- (Xu et al., 2009) indicating that ing bones of the hind foot with the first and this process may have occurred in parallel in a especially fifth toes at least partially reduced number of theropod lineages. Other shared derived (Figure 1.1). Recent discoveries of dinosauriforms traits that reflect the deep theropod origins of birds and basal dinosaurs from and New include a dorsally ascending process of the astrag- Mexico, USA provide new insights on the progres- alus, a fifth pedal digit reduced to only the meta- sive of the reduction and loss of weight- tarsal, and the wishbone. As briefly mentioned bearing function in the first and fifth toes of the above, the presence of clavicles (whether fused ornithodiran foot (Nesbitt et al., 2009a). or not) has been a topic of debate in avian and Dinosauria, which comprises the bulk of known theropod systematics for the better part of a cen- ornithodiran diversity, is characterized by a fully tury (Makovicky & Currie, 1998). Because many of perforated acetabulum, in which the head of the the earliest dinosaur discoveries were of advanced fits into a medially open socket formed by ornithopods and sauropodomorphs, taxa that the three hipbones. The head of the femur is angled have lost their clavicles, these elements were almost perpendicular to the shaft allowing the generally held to be absent in all dinosaurs, hindlimbs to be brought under the body for a leading to the subsequent misinterpretation fully upright, parasagittal gait. Triassic dinosaur- of theropod furculae as either interclavicles omorphs such as Lagerpeton with partially open (Makovicky & Currie, 1998) or fused gastralia Theropod Diversity and the Refinement of Avian Characteristics 13

Fig. 1.2 Comparison of theropod mani showing progressive reduction and loss of digits IV and V and changes in the proportions of manus elements. Eoraptor (A), (B), Sinornithosaurus (C), Archaeopteryx (D), and Confuciusornis (E). Abbreviations: DI–V, digits I–V. All specimens shown at the same scale. [This figure appears in color as Plate 1.2.]

(Chure & Madsen, 1996). Unfused clavicles are ing their interiors (Britt, 1997). These pneumatic now known to occur in prosauropods (Yates & features correlate with one of the avian respiratory Vasconcelos, 2005) and some ceratopsians air-sac systems (see below). Coelophysoids and (Brown & Schlaikjer, 1940), and fused clavicles more derived theropods are characterized by have been documented in an evergrowing list of a first in which the metatarsal is reduced theropod taxa, indicating that this trait is likely a and no longer contacts the ankle (Figure 1.1) – in synapomorphy of almost the entire clade (Smith most birds this toe is rotated on to the plantar et al., 2007; Nesbitt et al., 2009b). Almost all face of the foot and allows for perching in theropods with the exception of Herrerasaurus arboreal forms. also exhibit some degree of pneumatization on Theropods exclusive of the basal coelophysoid the sides of the postaxial cervical vertebral centra, radiation that spanned the –Early with basal forms such as Tawa and coelophysoids , can be largely divided into the two exhibiting fossae (Nesbitt et al., 2009a), while major lineages, and . more derived taxa have invasive foramina that Birds belong to the latter lineage, which pierce the vertebral centra invading and excavat- also includes such well-known denizens as 14 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Allosaurus, , and Velociraptor. in most cladistic analyses (Sereno Both clades radiated throughout the Jurassic and et al., 1996; Smith et al., 2007, 2008), through their giving rise to small- and large-bodied shared possession of traits such as perforated max- forms. Ceratosaurs are characterized by a progres- illary fenestra, pneumatic openings on the axial sive reduction of forelimb traits such as muscle centrum, and a reduction of the ischiadic apron to attachment areas and overall robustness of the form an open obturator notch rather than an forelimbs, accompanied by a reduction of their enclosed fenestra. Both spinosauroids and allo- relative size in larger taxa. sauroids were globally distributed during the Large bodied ceratosaurs such as , and Early Cretaceous, but wane in , and are grouped diversity during the latest Cretaceous, with the together in the clade , and are char- coelurosaurian tyrannosaurids filling the domi- acterized by very short, deep adorned with nant niche at least on northern surficial sculpturing and even horn-like structures landmasses. in some species (Bonaparte et al., 1990; Sampson Throughout most of the 20th century, carniv- et al., 1998; Sereno & Brusatte, 2008), as well as orous dinosaurs were taxonomically divided heavy reduction of forelimb elements including according to body size with small to medium- a completely unossified wrist and loss of phalan- sized taxa grouped in Coelurosauria, and large ges on most fingers (Chiappe et al., 1998). These species lumped within . This taxo- taxa have an almost exclusively Gondwanan nomic scheme is artificial (Holtz, 1994) and recent distribution during the Cretaceous, and have work demonstrates that large body sizes evolved been the subject of intense biogeographic debate multiple times within . Coelurosauria (Sampson et al., 2001; Sereno et al., 2004; Sampson has recently been redefined (Gauthier, 1986; & Krause, 2007). Sereno, 1999a) as the clade encompassing birds The Tetanurae have a greater known Mesozoic and all theropods closer to birds than to Allosau- diversity than do the ceratosaurs, and include rus, and encompasses the smallest known thero- a wide range of both large and small taxa. Teta- pods (hummingbirds) as well as some of the largest nuran theropods are characterized by a large suite (T. rex), although all lineages appear to derive from of synapomorphies, including presence of a max- small to medium-sized ancestors. Coelurosaurs illary fenestra (Figure 1.1) rostral to the archosau- are united by having a third opening, the promax- rian antorbital fenestra and a fully horizontally illary fenestra, within the antorbital fossa, directed femoral head. All birds exhibit this last although this is only recognizable in Archaeop- trait, although the maxillary fenestra is only evi- teryx among birds as is the case with the maxil- dent in Archaeopteryx, having been incorporated lary fenestra (Figure 1.1). Other traits uniting this into the naris in all living and most fossil avian grouping include the presence of three tym- species. Most tetanuran taxa are members of three panic pneumatic systems emanating from the principal groups, the Spinosauroidea, Allosauroi- middle ear (more basal taxa only possess one or dea, and Coelurosauria, although membership and two of these), an expanded and pneumatic exact relationships among these three subclades ectopterygoid (lost in neornithine birds), and remain debated (Sereno et al., 1996; Rauhut, 2003; an expanded astragalar ascending process that Smith et al., 2007, 2008). Allosauroids and spino- istwiceastallasitiswideandcoversalmost sauroids are predominantly large-bodied carni- the full width of the ankle. The manus is fully vores with body masses ranging from a few tridactyl in most Coelurosauria (Figures 1.1 hundred kilograms to as much as seven tons in and 1.2C–E), although a few basal taxa such derived end-members of these clades, such as the as the basal tyrannosauroid Guanlong (Xu spinosaurid and the carcharodonto- et al., 2006) retain a splint-like vestige of the saurid allosauroid . Allosauroids fourth metacarpal (Fig. 1B). The large predatory are recovered as the sister clade to the tyrannosauroids, the beaked and herbivorous Theropod Diversity and the Refinement of Avian Characteristics 15

(Kobayashi et al., 1999, Zanno & Makovicky, feathers themselves, thus providing not only evi- 2010) ornithomimosaurs, and small compsog- dence on the of these integumentary nathidsaregenerallyconsideredtobebasal structures, but also on the color and appearance of lineages within the coelurosaur radiation, these in life. Conversely, preservation of whereas birds and their closest taxa form a a body outline composed of frayed and decom- more exclusive clade within Coelurosauria posing dermal layers has never been reported in known as . In the past decade, discov- any of the hundreds of choristodere specimens eries of numerous coelurosaur taxa sporting proto- collected from these shale beds, casting doubt on feathers, fully formed flight feathers (Ji et al., 2001), the conclusion that such decomposition patterns and even specialized feathers convergent on display should be observed in a dinosaur as posited by structures in oscines (Zhang et al., 2008) have been Lingham-Soliar et al. (2007). made in Jurassic and Cretaceous -bed deposits Maniraptorans are characterized by distin- of northeastern China. Together, these discoveries guishing traits including a half-moon shaped indicate that possession of plumage covering most wrist bone that is thought to represent a fusion of the body with the exception of the feet and snout of the first and second distal carpal (Figure 1.1, trait likely characterizes at minimum the coelurosaur- 11). A pulley-like proximal surface on this element ian node. The presence of hollow, fibrous integu- allows the hand to be flexed sideways toward the mentary structures along the dorsal midline of two forearm, and is responsible for the wing-folding ornithischian taxa and on the body of pterosaurs mechanism in birds. As with many other ana- indicates the presence of such structures alone may tomical traits relevant to understanding the be much wider among ornithodirans, though their origins and relationships of birds, the refinement exact homology remains unclear. Doubts have been of this particular synapomorphy accrued over cast regarding the presence of homologues a range of branches in the phylogeny, and in various theropods (Lingham-Soliar et al., 2007; incipient versions of this structure have been Feduccia et al., 2005), and these have instead been recognized in more basal tetanurans such as interpreted as collagen fibers derived from decom- (Sereno, 1999b). A number of novel position of the skin in a specimen of the compsog- evolutionary features in the thoracic skeleton, nathid . These claims are based which are known to play a role in avian respiration only on very selective comparisons between the (O’Connor & Claessens, 2005), further diagnose soft tissues of feathered nonavian theropods and some maniraptorans. These include presence of experimentally manipulated integument on extant enlarged sternal plates with extensive medial reptiles and extant and extinct marine amniotes. contact and distinct facets for ossified sternal More appropriate comparisons to either birds or ribs (Barsbold, 1983; Norell & Makovicky, 1999) reptiles from the same shale deposits that yield (Figure 1.1, trait 16) and uncinate processes span- the feathered theropods, representing equivalent ning the thoracic ribs (Clark et al., 1999). preservational conditions, were not conducted by Maniraptoran fossils that preserve integumen- Lingham-Soliar and colleagues, and indeed dis- tary structures reveal an increased complexity in missed with the tautological argument that ani- both feather types and morphology over the mals with feathers are by definition birds and thus simple filamentous structures observed in basal need not be considered when testing for feather coelurosaurs (Xu et al., 2001; 2010). Xu and preservation (Fedduccia et al., 2005). Two recent colleagues (2001) demonstrated a correlation studies (Zhang et al., 2010; Li et al., 2010) demon- between the order of appearance of progressively strate that the preservation of nonavian coeluro- more complex feather types in maniraptoran evo- saur integumentary structures matches those of lution with their order of development in avian unquestionable stem birds from the same rock , and it is clear that almost all basic units and that they preserve melanosomes feather types known in birds had evolved earlier imbedded within the keratinous matrix of the in theropod evolution. 16 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Several aberrant clades of theropods are included sickle-clawed Deinonychosauria. Birds and deino- within the Maniraptora. These include the herbiv- nychosaurs are united by numerous apomorphic orous , whose theropod affinities features such as possessing retroverted pubes were strongly debated due to their unusual anat- (Figure 1.1, trait 19), an expanded and flexed cora- omy, but which are now known to possess unques- coid that repositions the humeral articulation tionable theropod hallmarks such as pneumatic closer to the vertebral column and imbues the vertebrae, furculae, feathers, and a semilunate car- scapulocoracoid with an L-shaped profile, a prox- pal that fuses in at least adult specimens of some imal articulation subdivided into two distinct taxa (Kirkland et al., 2005). Another group with facets(Figure1.1,trait18),andashortenedtailwith unusual anatomy and debated affinities are the 25 or less vertebrae of which the anterior ones are Alvarezsauridae. Derived, small-bodied members short and box-like and distal ones are elongate and of this group discovered in sedi- cylindrical (Figure 1.1, trait 17). Most paravians, ments of the Gobi Desert exhibit a remarkable including all birds, are also known to possess pri- mosaic of characters including loss of a postorbital mary and secondary feathers with asymmetrically bar, a double-headed quadrate, a keeled , developed vanes on either side of the rachis short but massive arms with an enlarged pollex (Figure 1.1, trait 20), a feature considered to be an but reduction of the other fingers, a splint-like adaptation for aerodynamic function, but the fibula, and diminutive, supernumerary teeth recently described Anchiornis exhibits symmetri- (Perle et al., 1993; Chiappe et al., 1998). A number cal vane distribution on its primaries (Xu et al. of these traits, such as the reduced postorbital, 2009) as in the basal oviraptorosaur Caudipteryx, streptostylic quadrate, keeled sternum, and complicating our understanding of how many reduced fibula, are also encountered in birds times this trait evolved. more derived than Confuciusornis, leading to ini- Deinonychosauria comprises two distinct tial hypotheses that these fossil taxa represent clades, the and , flightless birds more derived than Archaeopteryx. which are united by the presence of a sickle- Subsequent discoveries of more basal alvarezaurids shaped claw on the second digit of the foot (Fig- in Argentina led to the recognition that many of the ures 1.1 and 1.2, trait 21), and a triangular lateral avian-like characteristics of derived alvarezaurids exposure of the splenial along the edge of the lower evolved convergently in birds, and most recent jaw. A close relationship between dromaeosaurs studies agree that they represent basal manirap- and birds was initially recognized by Ostrom torans rather than members of the avian lineage (1976) following his discovery and description of (Norell et al., 2001; Novas & Pol, 2002; Senter, the first relatively complete dromaeosaurid 2007; Zanno et al., 2009). Deinonychus (Ostrom, 1969), but some debate Oviraptorosaurs represent another anatomi- persisted regarding the affinities of Troodontidae, cally bizarre lineage, some members of which derived members of which share characters with exhibit remarkable convergence on avian anatomy other coelurosaur clades and also lack some in parts of their skeleton. When analyzed in paravian synapomorphies such as a retroverted a limited context, such traits have also prompted . Discovery of a number of basal troodontids hypotheses that this clade represents secondarily from the Early Cretaceous Yixian and Jiufotang flightless birds (Maryanska et al., 2002), a conclu- Formation of China reveals that these traits are sion that is not supported in more rigorous and homoplastic in derived troodontids, and that comprehensive phylogenetic studies incorporat- Deinonychosauria is a natural grouping (Xu ing a greater array of both taxa and characters. et al., 2002). Many of the deinonychosaurs Such studies overwhelmingly posit oviraptoro- recently discovered in China and elsewhere are saurs (often, but not always, in combination with also significant because they represent the smal- therizinosaurs) as sister to the clade that lest nonavian dinosaurs yet discovered (Xu encompasses birds and their sister , the et al., 2000) and are comparable in body size to Theropod Diversity and the Refinement of Avian Characteristics 17 basal avian taxa such as Archaeopteryx and Jeho- basal members of the three principle paravian lornis (Turner et al., 2007; Xu et al., 2009). Some of lineages and hence some phylogenetic lability the small deinonychosaurs from these rock units, between them. The earliest instance of this such as Microraptor and Sinornithosaurus, pos- unique body plan is represented by Pedopenna sess vaned feathers on the hindlimb as well as the (Xu & Zhang, 2005; Figure 1.3B) and Anchiornis forelimb (Xu et al., 2001; Xu & Zhang, 2005; Ji (Hu et al., 2009), which are Middle Jurassic in age et al., 2001), along with a frond-like arrangement and thus older than Archaeopteryx. More derived of the rectrices in a pattern like that of Archaeop- deinonychosaurian taxa evolved larger body sizes teryx (Figure 1.3A). This four-winged body plan culminating in the 30 ft long Utahraptor. may represent a transitional step in the evolution The fossil record of Deinonychosauria was of powered flight (Longrich, 2006; Hu et al., 2009), until recently largely restricted to Cretaceous though its optimization on the evolutionary tree is deposits of the northern continents, but a slew complicated by the extreme similarity between of recent discoveries of dromaeosaurids from

Fig. 1.3 (A) Skeleton of the dromaeosaurid Microraptor gui from the Yixian Formation of Liaoning, China, exhibiting vaned, asymmetric feathers on both fore- and hindlimbs. (B) Detail of hindlimb primary feathers of Pedopenna from the Middle Jurassic of Inner Mongolia, China. Pedopenna is the earliest paravian fossil to exhibit vaned feathers and a four- winged body plan. The inset shows a close-up of the aligned and parallel barbs on each vane that indicate the presence of interlocking barbules, as well as the rachis. Note the large sickle claw characteristic of deinonychosaurians (¼ dromaeosaurs and troodontids) on digit II of the foot. Scale bars equal 5 cm. (Photographs: P. Makovicky.) [This figure appears in color as Plate 1.3.] 18 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Argentina (Novas & Puerta, 1997; Makovicky NEW INFERENCES ON SOFT TISSUE, et al., 2005; Novas & Pol, 2005) are evidence for PHYSIOLOGY, AND BEHAVIOR a Gondwanan radiation of these animals. The discovery of the near-complete of the The recent surge in dinosaur discoveries and Gondwanan dromaeosaurid Buitreraptor (Mako- research has not only yielded a better understand- vicky et al., 2005) provided evidence to unite all of ing of the skeletal evolution of theropods and these different taxa into a single basal lineage, the a more nuanced understanding of the stepwise Unenlagiinae, whose split from the better-known assembly of the unique avian body plan, but Laurasian dromaeosaurids may correlate with the also provided insights into the evolution of break up of Pangaea. The discovery of Gondwanan avian physiology, reproductive biology, and even dromaeosaurids also prompted a reinterpretation aspects of their related behaviors. Through inte- of the purported basal bird Rahonavis as member grative research incorporating fossil and neonto- of the Unenlagiinae, demonstrating that the ske- logical data, advances in our understanding of letons of basal deinonychosaurs and the earliest modern birds are being applied back in time birds are almost indistinguishable. Rahonavis is to generate hypotheses regarding aspects of characterized by hyperelongate forelimbs suggest- dinosaurian biology lost in the fossil record ing that such flight-related proportions may have using phylogenetic history as a guide (Witmer, arisen more than once in paravian evolution, with 1995). Here we review some recent advances in the main occurrence being characteristic of the our understanding of dinosaur biology based on avian lineage (Figure 1.1, trait 22). some of the most remarkable and informative Apart from Archaeopteryx and a few other theropod fossil discoveries made to date and species such as , most Cretaceous new methodological approaches to the study of avian fossils exhibit rapid evolution of the avian fossilized remains. body plan. Sapeornis is the most primitive bird to possess a foreshortened tail with the distalmost Metabolism and respiration segments fused into a (Figure 1.1, trait 23). Without a long tail to counterbalance the body A long-standing debate regarding dinosaurian as in typical nonavian theropods, the last common metabolic regimes has persisted for over 30 ancestor of Confuciusornis and more derived birds years (Chinsamy-Turan & Hillenius, 2004; Padian evolved a posture where the knee is permanently & Horner, 2004), since the recognition that birds angled to bring the center of mass above the foot are derived theropods prompted speculations that and offset the loss of a long counterbalanced tail. they inherited their homeothermic physiology An ossified kneecap, which is unknown in non- from dinosaurian ancestors. Debates on physio- avian dinosaurs, is present in Confuciusornis and logical inferences made on evidence such as his- more derived birds and serves to stabilize the bent tological traits, the possible presence or absence of knee. Sapeornis and Confuciusornis are also the turbinals, choanal position, and basic physiologi- basalmost avian taxa to exhibit a fused sternum cal calculations have been inconclusive and with an incipent sternal keel for anchoring marred by attempts to draw wide-ranging conclu- enlarged flight musculature, marking another sions through oversimplified interpretations of key step in the assembly of the modern avian relatively limited (and often inaccurate) data. body plan. Both retain primitive theropod traits, The presence of a plumage of filamentous or however, such as a functional grasping tridactyl downy feather homologues covering the body in hand (Figure 1.2E), and Sapeornis and most Cre- a variety of coelurosaurs, including taxa that taceous birds retain dentition. Though some ther- presage evolution of vaned feathers with aerody- opods convergently lost their teeth, the avian bill namic functions, suggests that feathers evolved appears to have arisen at or very close to the origin in response to selective pressures other than of the avian crown group. adaptation to aerodynamic locomotion (Norell Theropod Diversity and the Refinement of Avian Characteristics 19

& Xu, 2005; Li et al., 2010). Given the insulating common ancestor of ceratosaurs and tetanurans properties of feathers and the small size and possessed abdominal air sacs (O’Connor & Claes- correspondingly high surface area to volume sens, 2005; Sereno et al., 2008), and that most ratio of most of the nonavian theropods dis- tetanurans potentially had a clavicular air sac covered with plumage, many authors have con- (Makovicky et al., 2005; Sereno et al., 2008). It cluded that feather evolution may in part have should be noted that air sacs do not always invade been driven by a need for insulation, which in skeletal elements in living birds and the degree of turn implies an ability to generate metabolic pneumaticity is observed to correlate with life energy. A recent study of the histology of dino- history parameters such as body size and ecologi- saurs has demonstrated that theropods tend to cal habits (O’Connor, 2009), so absence of pneu- have smaller osteocyte lacunae in their bones, matic traces in bones of extinct theropods cannot indicating smaller cell sizes (Organ et al., 2007). be taken as evidence for absence of air sacs, espe- Living birds have relatively smaller cells and cially if such taxa are bracketed phylogenetically markedly lighter cell nuclei with far less redun- by taxa with positive evidence for air sacs. dant DNA compared to other amniotes. Small cell size facilitates increased basic metabolic Reproductive biology rates due to the higher surface to volume ratio of the cells, and correlates with nuclear mass, so Recent discoveries of nesting or gravid manirap- it is thought that birds underwent active selec- toran dinosaurs from Mongolia (Norell et al., 1995; tion for smaller nucleus size. Organ and col- Figure 1.4) and elsewhere (Varricchio et al., 1997; leagues’ (2007) results robustly suggest that Currie & Chen, 2001; Sato et al., 2005; Grellet- this selective process began much earlier in Tinner & Makovicky, 2006) have yielded crucial theropod history. insights into the evolution of avian reproductive The high avian basal metabolic rate is in part biology. Examination of the histology of such sustained by a unique respiratory system, in which specimens (Erickson et al., 2007) has demon- the incompressible lungs are ventilated by a com- strated that many of them are not fully grown. plex system of interconnected air sacs. Phyloge- Assuming their association with nests is demon- netic continuity has been established between the strative of a parental relationship, it suggests that pneumatic openings in the vertebral columns of nonavian theropods and perhaps even the earliest theropod dinosaurs and those of birds (Britt birds reached reproductive maturity before attain- et al., 1998; O’Connor & Claessens, 2005), ing somatic maturity (¼ cessation of growth). This which are formed through ontogeny as the air pattern was reinforced by a paleohistologic study sacs invade adjacent bones. Five main air sac of four dinosaur taxa, in which reproductive matu- systems are connected to the lungs either directly rity was established from the presence of bony or through their connections to one another in tissues interpreted as medullary bone, which in birds. Of these five, the cervical, clavicular, and some living birds serves as a calcium reserve for abdominal air sac systems invade and pneumatize generating eggshell in gravid females (Lee & Wern- vertebral, girdle, and even limb bones in birds. ing, 2007). The concordance between these studies Skeletal pneumatic features such as openings indicates that nonavian theropods, including para- into bones and honeycombed interior architecture vians, retained the primitive reptilian pattern of correlated with these systems have been recog- reproducing before attainment of somatic matu- nized in theropods, with vertebral pneumaticity rity as opposed to the modern avian reproductive related to the cervical air sacs being virtually cycle in which somatic maturity is decoupled ubiquitous in theropods (Britt et al., 1998). Hard from and precedes reproductive maturity (Erick- tissue correlates of the other two systems are less son et al., 2007; Lee & Werning, 2007). Given that a common, but widespread enough throughout the- number of primitive avian taxa including Arche- ropod diversity to suggest that at least the last opteryx (Erickson et al., 2009), Confuciusornis 20 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Fig. 1.4 Partial skeleton of an oviraptorosaur in brooding posture on a nest of its eggs. Egg identity has been independently confirmed through embryonic remains. Specimens such as this reveal that these dinosaurs laid eggs in pairs over protracted periods (diachronous laying), and brooded them with direct contact indicative of synchronous hatching. Such associations of eggs and sexually mature individuals are now known from multiple nonavian maniraptoran taxa. (Photograph M. Ellison/AMNH) [This figure appears in color as Plate 1.4.]

(Chiappe et al., 2008), and Patagopteryx (Chin- clutches comparable to those of crocodilians in samy et al., 1994) reveal cyclical growth patterns terms of egg numbers and individual egg volumes, and multiple age classes like nonavian dinosaurs, at least some maniraptoran taxa have significantly but unlike living birds which grow to maturity larger egg volumes (Varricchio & Jackson, 2004) very rapidly, the decoupling between growth rate indicating a shift toward the derived avian condi- and reproductive maturity likely occurred later in tion in this important parameter. Although clutch avian evolution. sizes for well preserved nests of the nonavian Living birds exhibit a relatively complex set of maniraptorans and Citipati scale accord- reproductive adaptations and behaviors relative to ing to the same equations as in living birds, indi- other reptiles (Varricchio et al., 1997; Varricchio & vidual eggs are about half the volume of an extant Jackson, 2004). Although dinosaur eggs and nests avian egg for an scaled to corresponding have been known for well over a century, and size. The eggs of these taxa are arranged in pairs correctly recognized since 1923, remarkable dis- within the nest, demonstrating that nonavian coveries of dinosaur embryos or adults associated theropods still retained two functional oviducts, with nests or eggs represent intermediate stages in rather than the single oviduct of extant avians the evolution of the uniquely avian mode of repro- (Varricchio et al., 1997; Clark et al., 1999; Sato duction. While most nonavian dinosaurs exhibit et al., 2005). The dimensions of the pubic canal in Theropod Diversity and the Refinement of Avian Characteristics 21 basal birds such as Archaeopteryx and Confuciu- to body-mass scaling into account (Jerison, 1973)), sornis, which retain a fused contact between the with particularly enlarged optical lobes and pubes distally, compares more favorably with cerebellum thought to be adaptations for neural those of nonavian theropods of similar size, rather control of flight. Despite popular misconceptions than to the unfused and expanded pelvic canals of regarding the brain size of nonavian dinosaurs, more derived birds. This suggests that individual theropods show a progressive increase in EQ egg volume was smaller in these ancestral birds throughout their evolutionary history and recon- than in modern taxa, and that they may have structed EQ values from brain of various retained two functional oviducts. extinct maniraptorans approach those of the basal The relatively large volume of nonavian man- lineages of living birds (Dominguez Alonso iraptoran clutches compared to body size (Varric- et al., 2004). Detailed three-dimensional exami- chio & Jackson, 2004) (Figure 1.4) precludes that nation of the brain of Archaeopteryx using com- all eggs were retained within the female and then puted X-ray tomography (Dominguez Alonso deposited during a single laying event. Rather et al., 2004) demonstrates that the brain of the these animals must have laid eggs over a pro- basalmost avian taxon already possesses a bird- tracted period of time as living birds do, a con- like architecture with a pronounced pontine clusion supported by analysis of the orientation flexure that displaces the hindbrain below the of egg pairs within individual nests. Coupled mid brain, and enlargements of features thought with evidence for a brooding posture in indivi- to be adaptations for enhanced neurosensory duals atop nests (Figure 1.4) (Norell et al., 1995; control of active flight in modern birds, such as Varricchio et al., 1997), and for synchronous enlarged optic lobes, a proportionately well devel- stages of embryonic development within one oped cerebellum, and an enlarged inner ear with clutch (Varricchio et al., 2002), this provides expanded semicircular canals set in a modern compelling evidence that extinct maniraptorans avian configuration. Some combination, though exhibited synchronous hatching like their living not all, of these traits have also been observed in relatives, but unlike more basal egg-laying nonavian coelurosaurs such as troodontids reptiles. (Norell et al., 2009), various oviraptorosaurs Taken together, these findings demonstrate (Balanoff et al., 2009), and ornithomimosaurs that birds inherited some components of their (Balanoff et al., 2009). Parallel trends in relative complex reproductive biology such as nest care/ brain size evolution in birds and other manirap- brooding, diachronous laying, and synchronous toran lineages such as oviraptorosaurs have been hatching from nonavian ancestors, whereas noted, but there is little doubt that an elevated EQ other components such as loss of one oviduct and expansion of certain parts of the brain in birds and concomitant increase in egg volume plus was inherited from a more distant maniraptoran decoupling of somatic and sexual maturity ancestor. Thus, with regard to the evolution of the occurred within the avian lineage itself. Much unique avian brain, phylogeny again demonstrates as with any of the other biological systems how highly derived avian traits were acquired in discussed here, avian reproduction is a mosaic of stepwise fashion throughout theropod evolu- inherited traits many of which pre-date bird tionary history. origins combined with others that post-date this event. ARE BIRD ORIGINS STILL CONTROVERSIAL? Brain evolution While widely accepted by biologists and paleon- Among amniotes, birds are characterized by large tologists, the theropod ancestry of birds is not relative brain sizes (measured as an encephaliza- without its critics. The hypothesis has been tion quotient (EQ) that takes the allometry of brain challenged by a vocal, if small, opposition who 22 PETER J. MAKOVICKY AND LINDSAY E. ZANNO have pointed to a number of perceived inconsis- Wellnhofer, 1996) represent separate ossifications tencies in the theropod ancestry of birds. In gen- rather than being part of the dentary (Figure 1.5). eral, their challenges fall into several categories, Others depend on indefensible assumptions that which include disagreements over homology of structures be completely identical to qualify as various traits and structures, the seeming homologues, such as Feduccia & Martin’s (1998) ‘temporal paradox’ in which most nonavian man- claim that variations in interclavicular angle of iraptorans post-date Archaeopteryx, and inconsis- the between some birds and nonavian tency between inferred theropod paleobiology and theropods represent evidence of separate evolu- preferred scenarios of how some aspect of avian tionary origins of these structures. Many such evolution (often involving hypothetical interme- misconceptions have been disproven by the diate forms) must have progressed. Indeed, most of wealth of evidence amassed against them, but these challenges rely on a combination of all three continue to be cited indefinitely by those favoring categories of arguments. A number of traits shared a nontheropod . by birds and various subsets of dinosaurs and Other challenges relating to the homology of theropods, such as a broad ascending process structures such as digit and wrist identity (Burke of the astragalus (Martin, 1991), the presence of & Fedduccia, 1997) and homology of the ascending a furcula (Feduccia &, Martin 1998), homology of process of the astragalus, conflate primary homol- the wrist and digit elements of the forelimb, and ogy statements based on comparisons in fossils even the presence of thecodont dentition in vari- with embryological observations on a limited set ous theropods, have been challenged (Martin & of avian model taxa. For example, Martin & Stewart, 1999). Many of these assertions, such as Stewart’s (1985; see also Martin et al., 1980) whether a furcula is present in nonavian theropods claim that the ascending process of the nonavian or whether theropod teeth are truly thecodont are theropod astragalus is fundamentally different simply based on inaccurate observations such as from the large spur of bone that emanates dorsally that the interdental plates of some theropod from the avian astragalus, because the former is taxa including Archaeopteryx (Elzanowski & termed a ‘process of the astragalus’ and the latter

Fig. 1.5 (A) Lower jaws of the Munich specimen of Archaeopteryx revealing the presence of interdental plates. (B) Cross-section of the dentary of Allosaurus revealing continuous histological ultrastructure between the bone below the alveoli and the interdental plates and demonstrating that the latter are not separate ossifications. Abbreviations: idp, interdental plates; sp, splenial; tg, germinating tooth. Specimens not to scale. (Photographs P. Makovicky.) [This figure appears in color as Plate 1.5.] Theropod Diversity and the Refinement of Avian Characteristics 23 derives from a distinct center of ossification during & Fallon, 2000; Wagner, 2005), and by the recent early and is dubbed the ‘pretibial bone’, discovery of the basal ceratosaur (Xu is based more on the polemics of how these struc- et al., 2009) which exhibits a reduced splint-like tures are named rather than on relevant observa- digit I and unusual phalangeal formula, demon- tions of their topological relationships. Since the strating that theropod hand evolution is not as embryology of nonavian theropods is unknown, it stereotypical as was once believed. With this remains undeniable that for the life stages that can assumption in doubt, the ‘digital mismatch’ can be compared across both living and fossil archo- no longer be invoked to disqualify the numerous saurs, this tall, flat spur of bone that rises from the derived similarities of element shapes and propor- ankle along the front the is present in virtually tions in hands of basal birds such as Archaeopteryx all birds and tetanurans, and is either significantly and other nonavian coelurosaurs. Moreover, these smaller or absent in more distantly related taxa experimental results support novel models of how (contra James & Pourtless, 2009). the embryology of the theropod hand evolved, Without question, the inconsistency between most notably the Frame Shift Hypothesis (Wagner identifying avian digits by comparison to archo- & Gauthier, 1999), which proposes a serial shift in saur fossils versus identifying them through digit identity between the embryonic primordia embryological studies of modern neornithines and chondrified digits over the course of develop- has been the greatest point of contention between ment. Predictions of this hypothesis with respect the two opposing camps. In short, the dilemma is to Hox gene expression patterns have been rooted in the fact that the tridactyl hand of Archae- recently confirmed with chicken digit II exhibit- opteryx exhibits a phalangeal formula and inter- ing the digit I Hox-gene expression pattern of elemental proportions that identify its digits as pentadactyl taxa such as mouse and alligator representing the first three fingers of the primitive (Vargas et al., 2008). Furthermore, criticisms pentadactyl hand, whereas embryologi- that a wholesale frame shift affecting all digits cal studies (Burke & Feduccia, 1997; Feduccia & in a limb is not documented in any other amniote Nowicki, 2002; Larsson & Wagner, 2002) identify taxon have recently been muted by the confir- these digits as arising from the limb bud conden- mation of a parallel case of a frame shift in the sations that develop into digits II–IV in nonavian Italian three-toed skink Chalcides (Wagner, 2005; amniotes. Young et al., 2009). Interpreting the results of these two different Another mainstay of the opposition to the the- methods for establishing digit identity at face ropod ancestry of birds has been to point to a value, detractors of the theropod origin of birds supposed ‘temporal paradox’ (Feduccia, 1996), (Burke & Feduccia, 1997) conclude that the avian namely the later occurrence in the fossil record and nonavian theropod hands cannot be homolo- of the coelurosaurian and maniraptoran sister gous despite the dozens of primary homologies in clades to birds when compared to Archaeopteryx. the shape, proportion, and number of wrist, hand, While the argument as a whole is based on the and finger bones, which they dismiss as conver- mistaken assumption that taxa such as Velocirap- gence. Such a conclusion implicitly assumes tor represent avian ancestors rather than sister a one-to-one correspondence between condensa- taxa, and should therefore occur earlier in the tions in the developing limb bud and ossified adult fossil record, it has also been rendered moot structures, however, something that is untestable by the recent discovery of several paravian taxa in fossils for which corresponding life stages are that pre-date Archaeopteryx (Xu & Zhang, 2005; not preserved. The fundamental assumption of Zhang et al., 2008; Xu et al., 2009). one-to-one correspondence has been challenged A third persistent trend in the polemics sur- by experimental data that demonstrates consider- rounding avian origins has been the construction able latent lability in the expression of chondrified of scenarios circumscribing how a complex func- digits from various primary condensations (Dahn tion such as avian flight or avian respiration 24 PETER J. MAKOVICKY AND LINDSAY E. ZANNO evolved, followed by application of these scenarios support a theropod ancestry for birds. While this as a “test” of the fossil record. The size discrepancy analysis certainly represents a step forward in the between basal birds and much larger paravians debate, their effort is deeply flawed on a number such as Velociraptor and Deinonychus was long of counts. For one, they based their analysis on cited as evidence that flight (and implicitly a now outdated dataset developed to examine birds themselves) could not have evolved from generic-level interrelationships of coelurosaurs such large and earthbound animals (Feduc- (Clark et al., 2002), and thus focused on traits cia, 1996), but discoveries of small maniraptoran uniting various maniraptoran genera to one taxa of comparable size to Archaeopteryx (Turner another, rather than on the traits more broadly et al., 2007) and with possible arboreal traits (Xu nesting birds within Theropoda. Citing many of et al., 2001) have erased this argument. Similar the older challenges to synapomorphies that sup- evidentiary concerns apply to other scenarios port the birds-as-theropods, they eliminate a num- based on the incorrect projection of parameters ber of relevant characters of the wrist, hand, and of the anatomy and physiology of living birds onto ankle while reinterpreting others to favor charac- distant fossil ancestors that have been summoned ter interpretations put forth by proponents of a against a theropod ancestry of birds. For example, nontheropod ancestry of birds, sometimes in illog- Ruben and colleagues (Ruben et al., 1997) ical fashion. For example James & Pourtless (2009) attempted to argue that theropods had a crocody- go to great length to defend Martin and collegues’ lian hepatic piston pump style of breathing based (1998) hypothesis that the hypocleideum of on interpretation of discolorations inside the body enanationithines has a distinct embryological cavity of two exceptionally preserved compsog- identity from that of crown birds and proceed to nathid specimens as defining the limits of a large redefine the relevant character definition, yet liver subdividing the thoracic cavity. However, in they never provide any insight on how such devel- both cases the limits of these discolorations have opmental distinctions are to be made on fully been demonstrated to be preservational artifacts ossified structures in fossil specimens. To this (Currie & Chen, 2001). decimated dataset they add a broad, but skewed, With the notable exception of a recent paper sample of more basal theropods, crocodilians, and by James & Pourtless (2009), none of these chal- the enigmatic and poorly preserved Triassic fossil lenges to the bird-theropod hypothesis have been , without a correspondingly suffi- set within a modern phylogenetic context and all cient increase in character sampling to accurately have relied on selectively picking certain traits, test the relationships of the diversity of added taxa. specimens, and observations while ignoring Critically, they omit any basal crurotarsan taxa or others to construct narrative, scenario-laden other Triassic archosauriforms necessary to prop- attacks on the theropod ancestry of birds. They erly evaluate the phylogenetic affinities of either contribute little to the overall understanding crocodylomorphs or the enigmatic Triassic fossil of how the derived avian body plan evolved Longisquama, despite the fact that taxon sam- and have generally offered few alternatives for pling has been recognized as a key parameter for avian ancestry, usually positing some vague, achieving accuracy in phylogenetic analysis paraphyletic assemblage of small bodied Triassic (Poe, 1998; Graybeal, 1998). Remarkably, despite reptiles as possible avian sister groups (Feddu- such manipulations, inadequate taxon and char- cia, 1996), or arguing for a close relationship acter sampling, and mistakes in data scoring (e.g. between birds and crocodylomorphs based on James & Pourtless’ (2009) statement that inter- select dental and cranial traits that have a homo- dental plates are absent in Archeopteryx; plastic distribution. Figure 1.5A), the primary signal of the original James & Pourtless (2009) recently presented data set examining the position of birds within a detailed phylogenetic analysis to challenge the maniraptoran dinosaurs remains largely intact, premise of whether such analyses unequivocally attesting to its robustness. Theropod Diversity and the Refinement of Avian Characteristics 25

To date, no credible alternative to the theropod much deeper origins near the base of the theropod ancestry of birds enjoys much support from radiation, while others, such as feathers, evolved the fossil record. Although, new fossil discoveries closer to birds but still characterize a more inclu- offer the potential to challenge existing hypothe- sive group of theropods. These discoveries also ses of relationships, when it comes to bird origins, allow inferences on which derived physiological such discoveries have only served to further and behavioral traits of modern birds evolved strengthen the theropod origin hypothesis before the avian lineage itself, and which ones through novel synapomorphies (e.g. wishbones, came later. Paleohistological data suggest that wrist anatomy, feathers), reduction of gaps in theropods had elevated basic metabolic rates the fossil record, and bridging gaps in parameters over those of living ectotherms, an inference cor- such as body size. In contrast, opposing views have roborated by the presence of feather homologues not been able to muster any new fossil taxa in in this clade, and by the growth rates approaching support of alternative hypotheses (vague as these those of metatherian mammals and basal avian have been; Prum, 2002) in the past 25 years, relying lineages (Erickson et al., 2001). Nevertheless, instead on reinterpretations of a handful of extremely high modern avian growth rates and fossils that are either so poorly preserved (e.g. decoupling between somatic and reproductive Longisquama) or whose identity is so contested maturity evolved much later within the avian (e.g. ) that consensus on their anatomy lineage itself. In similar fashion, birds inherited and affinities is lacking. proportionately large brain sizes from their coe- lurosaurian ancestors, but the evolutionary trend toward increased brain sizes continued within the CONCLUSIONS avian lineage such that modern birds generally have larger encephalization quotient (EQ) values Birds represent the most speciose and widespread than nonavian theropods. clade of amniotes and are characterized by The continuing pace of discovery as well unique locomotory and physiological adaptations, as technological advances in paleomolecular biol- which have long fascinated humans and been the ogy, CT scanning, and biogeochemistry hold great focus of intense evolutionary and ecological promise for future research surrounding the origin research. To fully understand these aspects of of birds. While our understanding of this impor- avian biology we need to comprehend the origin tant branching point in the tree of life has made a of birds and their traits in a historical context. quantum leap in recent years, much still remains While avian origins remained unresolved for most to be discovered about the earliest chapters in the of the first century that followed Darwin’s publi- and their biology. cation of “On the Origin of Species”, a string of discoveries of small- to medium-sized theropod dinosaurs since the 1960s has identified an ACKNOWLEDGMENTS inordinate number of derived characters shared with birds. We thank the editors G. Dyke and G. Kaiser for Today there is little debate over the theropod inviting this chapter and for their patience in ancestry of birds. The accelerated rate of discovery receiving it. Colleagues across the globe, including and description of well-preserved Mesozoic ther- Mark Norell, Carl Mehling, , Ricardo opods over the past decade and a half has not only Martinez, Oscar Alcober, Angela Milner, Alejan- strengthened this hypothesis, but has also immea- dro Kramarz, Oliver Rauhut, , surably improved our knowledge of how avian Rodolfo Coria, Diego Pol, and Oliver Hampe pro- anatomy and biology evolved. vided access to specimens and insights that helped We now know that some avian hallmarks such shape this manuscript. M. Donnelly executed the as the wishbone and skeletal pneumaticity have reconstruction in Figure 1.1. Research that 26 PETER J. MAKOVICKY AND LINDSAY E. ZANNO contributed to this chapter was supported by the Chure DJ, Madsen JH. 1996. On the presence of furculae U.S. National Science Foundation grant EAR in some non-maniraptoran theropods. Journal of 0228607. Vertebrate 16: 573–577. Clark JM, Norell MA, Chiappe LM. 1999. An oviraptorid skeleton from the Late Cretaceous of Ukhaa Tolgod, Mongolia, preserved in an avianl-like brooding REFERENCES position over an oviraptorid nest. American Museum Novitates 3265: 1–36. Alonso PD, Milner AC, Ketcham RA, Cookson MJ, Clark JM, Norell M, Makovicky P. 2002. Cladistic Rowe TB. 2004. The avian nature of the brain and approaches to the relationships of birds to other inner ear of Archaeopteryx. Nature 430: 666–669. theropods. In Mesozoic Birds: Above the Heads of Balanoff AM, Xu X, Kobayashi Y, Matsufune Y, Norell Dinosaurs, Chiappe LM, Witmer LD (eds). Berkeley: MA. 2009. Cranial osteology of the theropod dinosaur University of California Press; 31–61. Incisivosaurus gauthieri (Theropoda: Oviraptoro- Currie PJ, Chen PJ. 2001. Anatomy of Sinosauropteryx ). American Museum Novitates 3651: 1–35. prima from Liaoning, northeastern China. Canadian Barsbold R. 1983. Carnivorous dinosaurs from the Journal of Earth Sciences 38: 1705–1727. Cretaceous of Mongolia. Joint Soviet–Mongolian Dahn RD, Fallon JF. 2000. Digital identity is regulated by Paleontological Expedition Transactions 19: 5–120. interdigital bmp signaling. Developmental Dynamics (In Russian) 219:8. Benton MJ. 1990. Phylogeny of the major groups Dollo LAMJ. 1893. Les lois d’Evolution. Bulletin de la - morphological data and divergence dates. Journal of Socie´te´ Belge de Ge´ologie, de Pa´ leontologieet Molecular Evolution 30: 409–424. d’Hydrologie, Me´moire 7: 164–166. Benton MJ. 2008. How to find a dinosaur, the role of Domınguez Alonso P, Milner AC, Ketcham RA, synonymy in biodiversity studies. Paleobiology 34: Cookson MJ, Rowe TB. 2004. The avian nature of 516–533. the brain and inner ear of Archaeopteryx. Nature Bonaparte JF, Novas FE, Coria RA. 1990. Carnotaurus 430, 666–669. doi:10.1038/nature02706 satrei Bonaparte, the horned, lightly built carnosaur Elzanowski A, Wellnhofer P. 1996. Cranial morphology from the Middle Cretaceous of Patagonia. Natural of Archaeopteryx: Evidence from the seventh skele- History Museum of Los Angeles County, Contribu- ton. Journal of 16: 81–94. tions in Science 416: 1–41. Erickson GM, Rogers KC, Yerby SA. 2001. Dinosaurian Britt B. 1997. Postcranial pneumaticity. In The Encyclo- growth patterns and rapid avian growth rates. Nature pedia of Dinosaurs, Currie PJ, Padian K (eds). San 412: 429–433. Diego: Academic Press; 590–598. Erickson GM, Curry-Rogers K, Varricchio DJ, Britt BB, Makovicky PJ, Gauthier J, Bonde N. 1998. Norell MA, Xu X. 2007. Growth patterns in brooding Postcranial pneumatization in Archaeopteryx. Nature dinosaurs reveals the timing of sexual maturity in 395: 374–376. non-avian dinosaurs and genesis of the avian condi- Brown B, Schlaikjer EM. 1940. The structure and tion. Biology Letters 3: 558–561. relationships of Protoceratops. Annals of the New Erickson GM, Rauhut OWM, Zhou Z, Turner AH, York Academy of Sciences 40: 133–266. Inouye BD, Hu D, Norell MA. 2009. Was dinosaurian Burke AC, Feduccia A. 1997. Developmental patterns physiology inherited by birds? Reconciling slow and the identification of homologies in the avian hand. growth in Archaeopteryx. PLoS ONE 4(10): e7390. Science 278: 666–668. doi:10.1371/journal.pone.0007390 Chiappe LM, Norell MA, Clark JM. 1998. The of Feduccia A. 1996. The Origin and Evolution of Birds. a relative of the stem-group bird Mononykus. Nature New Haven: Yale University Press. 392: 275–278. Feduccia A, Martin LD. 1998. Theropod-bird link recon- Chinsamy A, Chiappe LM, Dodson P. 1994. Growth sidered. Nature 391: 754. rings in Mesozoic birds. Nature 368: 196–197. Feduccia A, Nowicki J. 2002. The hand of birds revealed Chinsamy-Turan A, Hillenius WJ. 2004. Physiology of by early ostrich embryos. Naturwissenschaften 89: non-avian dinosaurs. In The Dinosauria, Weishampel 391–393. DB, Dodson P, Osmolska H (eds). Berkeley: University Feduccia A, Lingham-Soliar T, Hinchcliffe JR. 2005. Do of California; 643–659. feathered dinosaurs exist? Testing the hypothesis on Theropod Diversity and the Refinement of Avian Characteristics 27

neontological and paleontological evidence. Journal Society of London Series B (Biological Sciences) 274: of Morphology 266: 125–166. 1823–1829. Gauthier JA. 1986. Saurischian monophyly and the Longrich N. 2006. Structure and function of hindlimb origin of birds. In The Origin of Birds and Evolution feathers in Archaeopteryx lithographica. Paleobiol- of Flight, Padian K (ed.). San Francisco: California ogy 32: 417–431. Academy of Sciences Memoir; 1–55. Makovicky PJ, Currie PJ. 1998. The presence of a furcula Grellet-Tinner G, Makovicky P. 2006. A possible egg in tyrannosaurid theropods, its phylogenetic and func- of the dromaeosaur Deinonychus antirrhopus: phylo- tional implications. Journal of Vertebrate Paleontol- genetic and biological implications: Canadian. Journal ogy 18: 143–149. of Earth Sciences 43: 705–719. Makovicky PJ, Apesteguıa S, Agnolin FL. 2005. The Heilman G. 1926. The Origin of Birds. London: With- earliest dromaeosaurid theropod from South America. erby; 210 pp. Nature 437: 1007–1011. Holtz TR. 1994. The phylogenetic position of the Martin LD. 1991. Mesozoic birds and the origin of birds. – implications for theropod system- In Origins of the Higher Groups of , Con- atics. Journal of Paleontology 68: 1100–1117. troversy and Consensus. Schultze H-P, Trueb L (eds). Hu DY, Hou LH, Zhang LJ, Xu X. 2009. A pre- Ithaca: Cornell University Press; 485–540. Archaeopteryx troodontid theropod from China Martin LD, Stewart JD. 1985. Homologies in the avian with long feathers on the metatarsus. Nature 461: . Nature 315: 159. 640–643. Martin LD, Stewart JD. 1999. Implantation and replace- Huxley TH. 1868. On the animals which are most nearly ment of bird teeth. Smithsonian Contributions to intermediate between birds and reptiles. Geological Paleobiology 89: 295–300. Magazine 5: 357–365. Martin LD, Stewart JD, Whetstone KN. 1980. The origin James FC, Pourtless IV, JA. 2009. Cladistics and the of birds: structure of the tarsus and teeth. Auk 97: origin of birds: a review and two new analyses. 86–93. Ornithological Monographs 66: 1–78. Martin LD, Zhou Z, Hou L, Feduccia A. 1998. Confus- Jerison HJ. 1973. Evolution of the Brain and Intelligence. ciusornis sanctus compared to Archaeopteryx New York, NY: Academic Press. lithographica. Naturwissenschaften 85: 286–289. Ji Q, Norell MA, Gao KQ, Ji SA, Ren D. 2001. The Maryanska T, Osmolska H, Wolsan M. 2002. Avialan distribution of integumentary structures in a feath- status for . Acta Palaeontologica ered dinosaur. Nature 410: 1084–1088. Polonica 47: 97–116. Kirkland JI, Zanno LE, Sampson SD, Clark JM, Nesbitt SJ, Smith ND, Irmis RB, Turner AH, Downs AM, DeBlieux DD. 2005. A primitive therizinosauroid Norell A. 2009a. A complete skeleton of a Late Tri- dinosaur from the Early Cretaceous of Utah. Nature assic saurischian and the early evolution of dinosaurs. 435: 84–87. Science 326: 1530–1533. Kobayashi Y, Lu JC, Dong ZM, Barsbold R, Azuma Y, Nesbitt SJ, Turner AH, Spaulding M, Conrad JL, Norell Tomida Y. 1999. Palaeobiology – herbivorous diet in MA. 2009b. The theropod furcula. Journal of Morphol- an ornithomimid dinosaur: Nature 402: 480–481. ogy 270: 856–879. Larsson HCE, Wagner GP. 2002. Pentadactyl ground Norell MA, Makovicky PJ. 1999. Important features of state of the avian wing. Journal of Experimental the dromaeosaurid skeleton II: Information from Zoology 294: 146–151. newly collected specimens of Velociraptor mongo- Lee AH, Werning S. 2008. Sexual maturity in growing liensis. American Museum Novitates 3282: 1–45. dinosaurs does not fit reptilian model. Proceedings of Norell MA, Xu X. 2005. Feathered dinosaurs. Annual the National Academy of Sciences 105: 582–587. Review of Earth and Planetary Sciences 33: 277–299. Li Q, Gao K-Q, Vinther J, Shawkey MD, Clarke JA, Norell MA, Clark JM, Chiappe LM, Dashzeveg D. 1995. D’Alba L, Meng Q, Briggs EEG, Prum RO. 2010. Plum- A nesting dinosaur. Nature 378: 774–776. age color patterns of an extinct dinosaur. Science 327: Norell MA, Makovicky PJ, Clark JM. 2001. Relation- 1369–1372. ships among Maniraptora: problems and prospects. In Lingham-Soliar T, Feduccia A, Wang XL. 2007. A new New Perspectives on the Origin and Evolution of Chinese specimen indicates that ’protofeathers’ in the Birds: Proceedings of the International Symposium Early Cretaceous theropod dinosaur Sinosauropteryx in Honor of John H. Ostrom, Gauthier JA, Gall LF are degraded collagen fibres. Proceedings of the Royal (eds). New Haven: Yale University Press; 49–68. 28 PETER J. MAKOVICKY AND LINDSAY E. ZANNO

Norell MA, Makovicky PJ, Bever GS, Balanoff AM, Sampson SD, Krause DW. 2007. Majungasaurus crena- Clark JM, Barsbold R, Rowe T. 2009. A review of tissimus (Theropoda: Abelisauridae) from the Late the Mongolian Cretaceous dinosaur Saurornithoides Cretaceous of Madagascar – Preface. Society of Verte- (Troodontidae: Theropoda). American Museun Novi- brate Paleontology Memoir 27: XIII–XIV. tates 3654: 1–63. Sampson SD, Witmer LM, Forster CA, Krause DW, Novas FE, Pol D. 2002. Alvarezsaurid relationships O’Connor PM, Dodson P, Ravoavy F. 1998. Predatory reconsidered. In Mesozoic Birds: Above the Heads dinosaur remains from Madagascar: Implications for of Dinosaurs, Chiappe LM, Witmer LM (eds). Berke- the Cretaceous biogeography of Gondwana. Science ley: University of California Press; 121–124. 280: 1048–1051. Novas FE, Pol D. 2005. New evidence on deinonycho- Sampson SD, Carrano MT, Forster CA. 2001. A bizarre saurian dinosaurs from the Late Cretaceous of Pata- predatory dinosaur from the Late Cretaceous of Mada- gonia. Nature 433: 858–861. gascar. Nature 409: 504–506. Novas FE, Puerta PF. 1997. New evidence concerning Sato T, Cheng YN, Wu XC, Zelenitsky DK, Hsiao YF. avian origins from the Late Cretaceous of Patagonia. 2005. A pair of shelled eggs inside a female dinosaur. Nature 387: 390–392. Science 308: 375–375. O’Connor PM. 2009. Evolution of archosaurian body Schweitzer MH, Watt JA, Avci R, Knapp L, Chiappe LM, plans: skeletal adaptations of an air sac-based breath- Norell MA, Marshall M. 1999. Beta-keratin specific ing apparatus in birds and other archosaurs. Journal of immunological reactivity in feather-like structures of Experimental Zoology, Part A – Ecological Genetics the Cretaceous alvarezsaurid, Shuvuuia deserti. Jour- and Physiology 311A: 629–646. nal of Experimental Zoology 285: 146–157. O’Connor PM, Claessens LP. A. M. 2005. Basic avian Senter P. 2007. A new look at the phylogeny of Coelur- pulmonary design and flow-through ventilation in osauria (Dinosauria: Theropoda). Journal of System- non-avian theropod dinosaurs. Nature 436: 253–256. atic Palaeontology 5: 429–463. Organ CL, Shedlock AM, Meade A, Pagel M, Edwards Sereno PC. 1999a. A rationale for dinosaurian . SV. 2007. Origin of avian genome size and structure in Journal of Vertebrate Paleontology 19: 788–790. non-avian dinosaurs. Nature 446: 180–184. Sereno PC. 1999b. The evolution of dinosaurs. Science Ostrom JH. 1969. Osteology of Deinonychus antirrho- 284: 2137–2147. pus, an unusual theropod from the Lower Cretaceous Sereno PC, Arcucci AR. 1993. Dinosaurian presursors of Montana. Peabody Museum of Natural History, from the Middle Triassic of Argentina: Lagerpeton Yale University Bulletin 30: 1–163. chanarensis. Journal of Vertebrate Paleontology 13: Ostrom JH. 1970. Archaeopteryx – notice of a new spec- 385–399. imen. Science 170: 537. Sereno PC, Brusatte SL. 2008. Basal abelisaurid and Ostrom JH. 1976. Archeopteryx and the origin of birds. carcharodontosaurid theropods from the lower Creta- Biological Journal of the Linnean Society 8: 91–182. ceous Elrhaz Formation of Niger. Acta Palaeontolo- Padian K, Horner JH. 2004. Dinosaur physiology. In The gica Polonica 53: 15–46. Dinosauria, Weishampel DB, Dodson P, Osmolska H Sereno PC, Dutheil DB, Iarochene M, Larsson HCE, Lyon (eds). Berkeley: University of California Press; GH, Magwene PM, Sidor CA, Varricchio DJ, Wilson 660–671. JA. 1996. Predatory dinosaurs from the Sahara and Poe S. 1998. Sensitivity of phylogeny estimation to tax- Late Cretaceous faunal differentiation. Science 272: onomic sampling. Systematic Biology 47: 18–31. 986–991. Perle A, Norell MA, Chiappe LM, Clark JM. 1993. Flight- Sereno PC, Wilson JA, Conrad JL. 2004. New dinosaurs less Bird from the Cretaceous of Mongolia. Nature link southern landmasses in the Mid-Cretaceous. Pro- 362: 623–626. ceedings of the Royal Society of London, Series B – Prum RO. 2002. Why ornithologists should care about Biological Sciences 271: 1325–1330. the theropod origin of birds: Auk 119: 1–17. Sereno PC, Martinez RN, Wilson JA, Varricchio DJ, Rauhut OWM. 2003. The interrelationships and evolu- Alcober OA, Larsson HCE. 2008. Evidence for avian tion of basal theropod dinosaurs. Special Papers in intrathoracic air sacs in a new predatory dinosaur Palaeontology 69: 3–213. from Argentina. PLoS ONE 3 (9): e3303. Ruben JA, Jones TD, Geist NR, Hillenius WJ. 1997. Lung doi:10.1371/journal.pone.0003303 structure and ventilation in theropod dinosaurs and Smith ND, Makovicky PJ, Hammer WR, Currie PJ. 2007. early birds. Science 278: 1267–1270. Osteology of ellioti (Dinosauria: Theropod Diversity and the Refinement of Avian Characteristics 29

Theropoda) from the Early Jurassic of Antarctica and Xu X, Zhang FC. 2005. A new maniraptoran dinosaur implications for early theropod evolution. Zoological from China with long feathers on the metatarsus. Journal of the Linnean Society 151: 377–421. Naturwissenschaften 92: 173–177. Smith ND, Makovicky PJ, Agnolin FL, Ezcurra MD, Pais Xu X, Zhou ZH, Wang XL. 2000. The smallest known DF, Salisbury SW. 2008. A -like theropod non-avian theropod dinosaur. Nature 408: 705–708. (Dinosauria: Tetanurae) in Australia: support for Xu X, Zhou HH, Prum RO. 2001. Branched integumental faunal exchange across eastern and western structures in Sinornithosaurus and the origin of feath- Gondwana in the Mid-Cretaceous. Proceedings of ers. Nature 410: 200–204. the Royal Society of London, Series B – Biological Xu X, Norell MA, Wang XL, Makovicky PJ, Wu XC. 2002. Sciences 275: 2085–2093. A basal troodontid from the Early Cretaceous of Turner AH, Pol D, Clarke JA, Erickson GM, Norell MA. China. Nature 415: 780–784. 2007. A basal dromaeosaurid and size evolution pre- Xu X, Norell MA, Kuang XW, Wang XL, Zhao Q, Jia CK. ceding avian flight. Science 317: 1378–1381. 2004. Basal tyrannosauroids from China and evidence Vargas AO, Kohlsdorf T, Fallon JF, VandenBrooks J, for protofeathers in tyrannosauroids. Nature 431: Wagner GP. 2008. The evolution of HoxD-11 expres- 680–684. sion in the bird wing: insights from Alligator missis- Xu X, Clark JM, Forster CA, Norell MA, Erickson GM, sippiensis. PLoS ONE 3(10): e3325. doi:10.1371/ Eberth DA, Jia CK, Zhao Q. 2006. A basal tyranno- journal.pone.0003325 sauroid dinosaur from the Late Jurassic of China. Varricchio DJ, Jackson FD. 2004. Two eggs sunny-side Nature 439: 715–718. up: reproductive physiology in the dinosaur Troodon Xing X, Qi Z, Norell MA, Sullivan C, Hone DWE, Erick- formosus.InFeathered Dragons: Studies on the Tran- son GM, Wang X-L, Han F-L and Yu G. 2009. A new sition from Dinosaurs to Birds, Currie PJ, Koppelhus feathered maniraptoran dinosaur fossil that fills a EB, Shugar MA, Wright JA (eds). Bloomington: Indiana morphological gap in avian evolution. Chinese Sci- University Press; 215–233. ence Bulletin 54: 430–435. Varricchio DJ, Jackson F, Borkowski JJ, Horner JR. 1997. Xu X, Clark JM, Mo JY, Choiniere J, Forster CA, Erickson Nest and egg clutches of the dinosaur Troodon for- GM, Hone DWE, Sullivan C, Eberth DA, Nesbitt S, mosus and the evolution of avian reproductive traits. Zhao Q, Hernandez R, Jia CK, Han FL, Guo Y. 2009. A Nature 385: 247–250. Jurassic ceratosaur from China helps clarify avian Varricchio DJ, Horner JR, Jackson FD. 2002. Embryos and digital homologies. Nature 459: 940–944. eggs for the Cretaceous theropod dinosaur Troodon Yates AM, Vasconcelos CC. 2005. Furcula-like clavicles formosus. Journal of Vertebrate Paleontology 22: in the prosauropod dinosaur . Journal 564–576. of Vertebrate Paleontology 25: 466–468. Wagner GP. 2005. The developmental evolution of avian Young RL, Caputo V, Giovannotti M, Kohlsdorf T, digit homology: an update. Theory In Biosciences 124: Vargas AO, May GE, Wagner GP. 2009. Evolution of 165–183. digit identity in the three-toed Italian skink Chalcides Wagner GP, Gauthier JA. 1999. 1, 2, 3 ¼ 2, 3, 4: A solution chalcides: a new case of digit identity frame shift. to the problem of the homology of the avian hand. Evolution and Development 11: 647–658. Proceedings of the National Academy of Sciences, Zanno LE, Gillette DD, Albright LB, Titus AL. 2009. A USA 96: 5111–5116. new North American therizinosaurid and the role of Weishampel DB, Dodson P, Osmolska H. (eds). 2004. The herbivory in ‘predatory’ dinosaur evolution. Proceed- Dinosauria. Berkeley: University of California Press. ings of the Royal Society of London, Series B – Bio- Witmer LM. 1995. The extant phylogenetic bracket and logical Sciences 276: 3505–3511. the importance of reconstructing soft tissues in fos- Zhang FC, Zhou ZH, Xu X, Wang XL, Sullivan C. 2008. A sils. In Functional Morphology in Vertebrate Paleon- bizarre Jurassic maniraptoran from China with elon- tology, Thomason JJ (ed.). Cambridge: Cambridge gate ribbon-like feathers. Nature 455: 1105–1108. University Press; 19–33. Zhang F, Kearns S, Orr P, Benton M, Zhou Z, Johnson D, Xu X, Norell MA. 2004. A new troodontid dinosaur from Xu X, Wang X. 2010. Fossilized melanososmes and the China with avian-like sleeping posture. Nature 431: colour of Cretaceous dinosaurus and birds. Nature 838–841. 463: 1075–1078.