Read an Excerpt

Tyrannosaurid Paleobiology

Indiana University Press

Phylogenetic Revision of Chingkankousaurus fragilis, a Forgotten Tyrannosauroid from the Late Cretaceous of China

Stephen L. Brusatte, David W. E. Hone, and Xu Xing

Introduction

Recent discoveries, especially the feathered theropods of the Jehol Biota, have placed China at the forefront of contemporary dinosaur research (e.g., Chen et al. 1998; Xu et al. 2003; Norell and Xu 2005; Xu and Norell 2006). However, vertebrate paleontology has a long history in China, and the country's rich dinosaur fossil record has been explored for over a century. Much of the pioneering work on China's dinosaurs was led by C. C. Young (Yang Zhongjian), the "father of Chinese vertebrate paleontology." For over 40 years, from the early 1930s until his death in 1979, Young spearheaded expeditions across China and discovered many of the country's most recognizable dinosaurs, such as the colossal sauropod Mamenchisaurus and the prosauropods Lufengosaurus and Yunnanosaurus (Dong 1992).

In 1958, Young described a single fragmentary bone from the Late Cretaceous (Campanian-?Maastrichtian; see Weishampel et al. 2004; Zhao et al. 2008) Wangshi Series of Shandong Province as a new genus and species of giant theropod, Chingkankousaurus fragilis. This specimen, the posterior region of a large right scapula (IVPP V 836), has long been ignored because of its fragmentary condition. However, those authors who have considered this specimen have often disagreed about its phylogenetic affinities. Young himself (1958) noted similarities with Allosaurus, and much later Dong (1992) formally assigned the specimen to Allosauridae. Steel (1970) and Dong (1979) placed the specimen within Megalosauridae, a wastebasket assemblage of large theropods that are now regarded as basal tetanurans (Benson 2010; Benson et al. 2010). Finally, Molnar et al. (1990:199) referred IVPP V 836 to Tyrannosauridae "on the basis of its very slender scapular blade." This referral was taken one step further by Holtz (2004), who synonymized Chingkankousaurus with the common Asian Late Cretaceous tyrannosaurid Tarbosaurus. Unfortunately, most of these referrals have been based on vague criteria and were often simply asserted instead of supported by explicit discussion of characters and measurements. This was often unavoidable at the time, but an influx of new theropod discoveries from Asia and elsewhere over the past two decades now allows a firm basis for comparison.

In this chapter, we reassess IVPP V 836 based on firsthand examination of the specimen, compare it with the scapulae of other theropods, and use this information to comment on the taxonomy and phylogenetic placement of Chingkankousaurus fragilis. Although a systematic revision of a fragmentary specimen may seem trivial, it is important to establish the phylogenetic affinities of IVPP V 836 because this specimen has been referred to many disparate theropod groups and comes from an area (Shandong) where the theropod fauna has been more poorly sampled than in many other regions in China. If it truly does represent an allosauroid or megalosaurid, then this specimen would be among the last surviving members of these groups, would greatly expand their stratigraphic ranges in Asia, and would indicate that more basal theropods persisted alongside tyrannosaurids in the large predator niche of Late Cretaceous Asia (contrary to Brusatte et al. 2009b). However, if IVPP V 836 represents a tyrannosaurid or a closely related form, it is further evidence that that these enormous theropods were the sole large predators during the waning years of the Cretaceous in Laurasia.

Institutional Abbreviations AMNH, American Museum of Natural History, New York; HMB, Humboldt Museum für Naturkunde, Berlin; IVPP, Institute of Vertebrate Paleontology and Paleoanthropology, Beijing; JME, Jura Museum, Eichstatt, Germany; LH, Long Hao Institute of Geology and Paleontology, Hohhot, China; MCNA, Museo de Ciencias Naturales y Antropológicas (J. C. Moyano) de Mendoza, Mendoza, Argentina; MPC, Mongolian Paleontological Center, Ulaanbaatar; UMNH, Utah Museum of Natural History, Salt Lake City.

Phylogenetic Definitions and Phylogenetic Framework

In this chapter we use the phylogenetic definitions of Sereno et al. (2005) for Tyrannosauroidea and Tyrannosauridae. Tyrannosauroidea is defined as the most inclusive clade containing Tyrannosaurus rex but not Ornithomimus edmontonicus, Troodon formosus, or Velociraptor mongoliensis. The more derived Tyrannosauridae is defined as the least inclusive clade containing T. rex, Gorgosaurus libratus, and Albertosaurus sarcophagus. In our discussion of tyrannosauroid phylogeny, we follow the phylogenetic analysis and cladogram presented by Brusatte et al. (2010). This cladogram is depicted in Figure 1.5, and major clades are denoted.

Identification

Although fragmentary, IVPP V 836 (Fig. 1.1) can be identified as a partial right scapula owing to its shape and features of its morphology. This bone was originally described as a scapula by Young (1958), an identification that has been followed by subsequent authors (e.g., Molnar et al. 1990). However, Chure (2000) questioned this identification, noting that the symmetrical cross section figured by Young (1958) is unusual for a scapula. Although Young (1958) describes the cross section as symmetrical, in fact the medial surface is convex, and the lateral surface is flat to slightly concave, as is usual for theropod scapulae (Fig. 1.2). This results in a triangular cross section at mid-shaft and a semi-ovoid cross section anteriorly at the broken edge (Fig. 1.1D). The medial convexity is due to a pronounced ridge, described below, which is a normal feature for tyrannosaurid (e.g., Brochu 2003:fig. 80) and other theropod scapulae (Fig. 1.2A–B). Other features of the bone, such as the slightly concave lateral surface and weakly rugose distal end, are also present in theropod scapulae (Fig. 1.2C–D).

Other possible identifications for the bone, including the possibility that it is part of a dorsal rib or a gastral element, are untenable. The specimen is straight along its entire length, whereas theropod dorsal ribs are strongly curved, and only very small fragments would appear straight if observed in isolation (e.g., Madsen 1976:pl. 40; Brochu 2003:fig. 64). Additionally, the dorsal ribs of large theropods often bear a thick ridge on their anterior surface, which is paralleled by a depressed groove (e.g., Daspletosaurus: AMNH 5468). The posterior surface is often corrugated, with a deep groove corresponding to the ridge on the lateral surface. This morphology is not present in IVPP V 836, which has a single ridge on one surface and a flat to slightly concave opposing surface. Although the distal ends of anterior dorsal ribs are sometimes expanded to articulate with the sternum, these expansions are usually slight and rarely, if ever, more than twice mid-shaft depth, as is the case in IVPP V 836 (e.g., Lambe 1917:figs. 6, 7; Brochu 2003:fig. 64).

Similarly, gastral elements of the largest theropods, such as Tyrannosaurus, are smaller than IVPP V 836, and their detailed morphology differs (e.g., Brochu 2003:fig. 70). In particular, although the medial ends of the medial gastral elements may be expanded relative to the mid-shaft, these expansions are usually irregular in shape (not spatulate as in IVPP V 836), extremely rugose, and often fused to the opposing medial gastral element. Additionally, IVPP V 836 is extremely large for a theropod gastral element.

Description

IVPP V 836 is the posterior end of a right scapula, measuring 520 mm long anteroposteriorly as preserved (Fig. 1.1). It is 47 mm deep dorsoventrally at its broken proximal end, and it maintains a relatively constant depth for most of the length of the shaft. However, it thins slightly to 43 mm in depth before expanding distally into a spatulate shape. As preserved, this expansion is 83 mm deep, but both its dorsal and ventral margins are eroded. When the preserved dorsal and ventral margins of the more proximal shaft are extended distally, filling in some of the missing regions, it appears as if the distal expansion was at least 94 mm deep. It is likely, however, that it was even deeper in life, as both the dorsal and ventral edges of the expansion are still quite thick, whereas they usually taper to a thin crest in most large theropod scapulae.

The lateral surface of the scapula is flat to slightly concave (Fig. 1.1A). The concavity is deepest dorsally, where it is overhung by a thickened ridge that parallels the dorsal margin of the blade. The ridge is thickest at the midpoint of the preserved fragment and thins out both proximally and distally. Ventrally the lateral concavity becomes progressively weaker until the lateral surface flattens out. This flat region, which corresponds to a flat surface on the medial surface of the blade, occupies approximately one half of the blade height.

The medial surface of the scapula is generally convex, due to the presence of a medial ridge (Fig. 1.1B–C). The ridge is strongest proximally: here it is most convex medially and also most extensive dorsoventrally, as it comprises the entire medial surface of the blade. Distally the ridge becomes weaker, as it becomes less convex and offset and thins into a more discrete crest that sweeps dorsally to parallel the dorsal margin of the blade. The ridge eventually funnels out into a broad triangular shape, which smoothly merges with the flat medial surface of the distal expansion.

Both the lateral and medial surfaces of the spatulate distal expansion are rugose (Fig. 1.1). This rugosity is most pronounced on the medial surface and takes the form of a mottled array of pits and raised mounds. A similar pattern of rugosity is seen on well-preserved theropod scapulae and corresponds to a number of muscle attachment sites (Brochu 2003:fig. 81). Additionally, in some tyrannosaurids the medial surface of the scapular expansion is more rugose than the lateral surface (e.g, Albertosaurus: AMNH 5458).

The ventral margin of the scapula, as seen in lateral and medial views, is straight for a short region proximally but describes a broad, concave arch distally. The dorsal margin, in contrast, is straight for most of its length. There is a small region distally that appears to be convex, but this may be an artifact of erosion. However, there is a slightly convex, raised margin in this region in some tyrannosaurid scapulae (Brochu 2003:fig. 80), suggesting that it may be a real feature.

Comparisons and Phylogenetic Affinity

Despite being a fragment of a single bone, IVPP V 836 exhibits a number of features that can be compared with those of other theropods (Fig. 1.3), allowing for a reasonable discussion and determination of its phylogenetic affinities. Importantly, the fact that the minimum shaft depth is preserved allows for the estimation of two important ratios that quantify scapula gracility and the relative size of the distal expansion (Fig. 1.4).

Although complete measurements are not possible, IVPP V 836 is clearly an elongate, gracile, and strap-like scapula. The length of the bone is at least 12 times greater than its minimum dorsoventral height, which is known with certainty (Table 1.1). A blade that is more than 10 times longer than deep has been used as a phylogenetic character in tyrannosauroid cladistic analyses and is optimized as a synapomorphy of Tyrannosauridae or slightly more or less inclusive clades (Sereno et al. 2009:character 69; Brusatte et al. 2010:character 234). As shown in Table 1.1, all tyrannosauroids except Dilong and Guanlong possess this character, although the latter taxon approaches this condition, whereas only a few non-tyrannosauroid theropods exhibit such strap-like scapulae.

Additionally, although complete measurements are again impossible, the distal expansion of IVPP V 836 is extensive compared to depth of the blade itself (Table 1.2). The ratio of the expansion depth to the minimum depth of the blade is at least 2.2 and was probably much greater in life. An expansion that is more than twice the minimum blade depth has been used as a phylogenetic character in tyrannosauroid cladistic analyses and also optimizes as a synapomorphy of Tyrannosauroidea or proximate clades (Holtz 2001:character 82; Holtz 2004:character 386; Sereno et al. 2009:character 70). As shown in Table 1.2, all tyrannosauroids except Guanlong possess this character, whereas the scapulae of other large theropods have relatively less expanded distal ends. There is also phylogenetically informative variation within tyrannosauroids, as all taxa more derived than Raptorex possess an expansion that is more than 2.5 times minimum blade depth (see also Brusatte et al. 2010:character 235).

When these two ratios are plotted against each other in a simple bivariate plot, most tyrannosauroids are seen to occupy the upper right-hand corner of the graph, whereas other theropods fall into the lower left-hand corner (Fig. 1.4). IVPP V 836 falls on the edge of the tyrannosauroid cluster but could only shift deeper within the tyrannosauroid region of the plot if more complete measurements were possible (because, for IVPP V 836, the scapular gracility ratio must have been greater than the plotted 12.23, and the scapular expansion ratio must have been greater than the plotted 2.2). In other words, even though IVPP V 836 is incomplete and likely missing much of its proximal region and distal end, the fragmentary preserved remains are themselves enough to quantitatively document similarities with tyrannosauroids to the exclusion of other theropods. The complete scapula could only be more strap-like, with a relatively larger distal expansion. In short, it could only be more tyrannosauroid-like.

Other features of the scapula are shared with tyrannosauroids as well. The straight dorsal margin and concave ventral margin are seen in Albertosaurus (Parks 1928), Dilong (IVPP V 14242), Eotyrannus (MIWG 1997 550), Gorgosaurus (Lambe 1917),Guanlong (Xu et al. 2006), Tarbosaurus (Maleev 1974), and Tyrannosaurus (Brochu 2003). Other large theropods exhibit different morphologies (Fig. 1.3). For instance, in most allosauroids, both margins are straight (Aerosteon: MCNA-PV3137; Allosaurus: UMNH UUVP 4423; Neovenator: Brusatte et al. 2008; Sinraptor: Gao 1999). In Acrocanthosaurus (Currie and Carpenter 2000), as well as Ceratosaurus (Madsen and Welles 2000) and Piatnitzkysaurus (Bonaparte 1986), the dorsal margin is concave, and the ventral margin is straight or convex. Finally, it is also possible that the pronounced rugosity on the medial surface of the distal end, seen in IVPP V 836 and Albertosaurus (AMNH 5458), may be a synapomorphy of tyrannosauroids or a less inclusive clade, but it is only apparent on well-preserved specimens. Only additional material can clarify this feature.

Systematic and Phylogenetic Placement

As shown, IVPP V 836 shares features with tyrannosauroids that are otherwise unknown, or rare, in other large theropods. Additionally, it comes from a time (Late Cretaceous) and place (Asia) in which tyrannosaurids were common animals and likely the sole apex predators in most terrestrial ecosystems (Currie 2000; Brusatte et al. 2009b). Therefore, we assign IVPP V 836 to Tyrannosauroidea. Within Tyrannosauroidea, IVPP V 836 is more derived than the basal taxa Guanlong and Dilong in both of the scapular ratio characters considered above (Tables 1.1 and 1.2), and therefore it can be assigned to the unnamed tyrannosauroid clade that includes Eotyrannus, Stokesosaurus, Xiongguanl ong, Raptorex, Bistahieversor, Dryptosaurus, Appalachiosaurus, and Tyrannosauridae (see Brusatte et al. 2010). This phylogenetic position is visually shown in the cladogram in Figure 1.5.

It is tempting to assign IVPP V 836 to even less inclusive clades, such as Tyrannosauridae or even Tarbosaurus. Indeed, Holtz (2004) formally assigned IVPP V836 to Tarbosaurus and sunk Chingkankousaurus fragilis, which he considered a nomen dubium, into the genus Tarbosaurus. We agree that C. fragilis is a nomen dubium – there are no clearly autapomorphic features on IVPP V 836, nor a unique combination of characters that can diagnose it relative to other tyrannosauroids. However, we hesitate to refer the specimen to a less inclusive clade than the Eotyrannus + Stokesosaurus + more derived tyrannosauroid clade.

Referring IVPP V 836 to Tarbosaurus is problematic for two reasons. First, Tarbosaurus does not possess any clearly autapomorphic features of the scapula, and we prefer synapomorphy-based assessments (sensu Nesbitt and Stocker 2008) when referring fragmentary fossils to established taxa. Second, there are at least two other large tyrannosauroids that lived during the Late Cretaceous of Asia, Alioramus (Kurzanov 1976; Brusatte et al. 2009a) and Alectrosaurus (Gilmore 1933; Mader and Bradley 1989). Scapulae are unknown for both of these genera, thus precluding any comparison with IVPP V 836.

---

Excerpted from Tyrannosaurid Paleobiology by J. Michael Parrish, Ralph E. Molnar, Philip J. Currie, Eva B. Koppelhus. Copyright © 2013 The Burpee Museum of Natural History. Excerpted by permission of Indiana University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---