Recent discoveries of more than ten new species of tyrannosauroid theropods are helping to understand the origin and evolution of colossal body size and other characteristic features of Tyrannosaurus rex and its terminal Cretaceous relatives. Particularly important has been the discovery and reinterpretation of Late Jurassic tyrannosauroids from Europe and North America, which are intermediate in size and phylogenetic position between small basal tyrannosauroids and the largest Late Cretaceous species. The fragmentary nature of these Jurassic specimens, however, has frustrated attempts to understand their systematics and phylogeny. A new specimen from the Late Jurassic of England was recently named as a new species (Stokesosaurus langhami) of the genus Stokesosaurus, which is known from several fragmentary fossils from North America. We review the systematics and phylogeny of these European and North American specimens and show that there are no unequivocal synapomorphies uniting them. Furthermore, a revised phylogenetic analysis does not recover them as sister taxa. This necessitates a taxonomic revision of this material, and we name a new genus (Juratyrant) for the British specimen.
Introduction
Over the past decade, a wealth of new fossil discoveries, phylogenetic analyses, and explicit biomechanical studies have shed new light on the anatomy, evolution, and biology of tyrannosauroid theropod dinosaurs (see review in Brusatte et al. 2010). A major driver of this resurgence has been the discovery of more than 10 new tyrannosauroid species, many of which are small-bodied animals, not much larger than a human, that lived up to 100 million years before the iconic terminal Cretaceous Tyrannosaurus rex (e.g., Hutt et al. 2001; Rauhut 2003a; Xu et al. 2004, 2006; Carr et al. 2005, 2011; Benson 2008; Brusatte et al. 2009; Ji et al. 2009; Sereno et al. 2009; Averianov et al. 2010; Carr and Williamson 2010; Li et al. 2010). These new species have begun to unveil the sequence of character and body size changes during the transition from small-bodied basal tyrannosauroids to the colossal tyrannosaurids (T. rex and close relatives) that lived during the final 20 million years of the Cretaceous and exceeded 1 tonne in mass (Erickson et al. 2004).
Among this influx of new tyrannosauroid discoveries is a specimen from the Late Jurassic of England that was described by Benson (2008) as the holotype of a new species, Stokesosaurus langhami (OUMNH J.3311-1–J.3311-30). This specimen is particularly important because it is one of the most complete tyrannosauroid fossils from the Jurassic, a time that tyrannosauroids appear to have had a wide (and perhaps cosmopolitan) distribution, but during which they lived in the shadow of other giant predatory dinosaurs (basal tetanurans; Benson 2010; Benson et al. 2010; Carrano et al. 2012). In his initial description, Benson (2008) noted several similarities between the new specimen and the fragmentary remains of another species, Stokesosaurus clevelandi Madsen, 1974 (the type species of Stokesosaurus), from the Late Jurassic of North America. As a result, Benson (2008) referred the new species to the genus Stokesosaurus. Over the past few years, however, the discovery of several new tyrannosauroids suggests that many of the characters used by Benson (2008) to unite S. clevelandi and S. langhami may actually be more widely distributed among basal tyrannosauroids. This leaves open the question of whether the two species are sister taxa, and therefore, whether S. langhami can be retained in the genus Stokesosaurus or requires a new generic name.
We here provide a systematic reassessment of the genus Stokesosaurus and its two constituent species. We show that the characters used by Benson (2008) to unite both species are problematic, either because they indeed are more widely distributed among basal tyrannosauroids, or because they cannot confidently be assessed in many other tyrannosauroids. Thus, strong evidence for a clade comprising S. clevelandi and “S.” langhami may be lacking. We then provide a new phylogenetic analysis, based on the recent comprehensive dataset of Brusatte et al. (2010), which does not find support for a sister grouping of S. clevelandi and S. langhami. This necessitates the removal of “S.” langhami from Stokesosaurus, and requires the erection of a new generic name for the British material.
Institutional abbreviations.—BMR, Burpee Museum of Natural History, Rockford, Illinois, USA; CMN, Canadian Museum of Nature, Ottawa, Canada; FRDC, Fossil Research and Development Center, Gansu Bureau of Geology and Mineral Resources Exploration, China; IVPP, Institute of Vertebrate Paleontology and Paleoanthropology, Beijing, China; LH, Long Hao Institute of Geology and Paleontology, Hohhot, Nei Mongol Autonomous Region, China; MIWG, Museum of Isle of Wight Geology, Isle of Wight County Museum Service, Sandown, UK; OUMNH, Oxford University Museum of Natural History, Oxford, UK; UMNH, Utah Museum of Natural History, Salt Lake City, Utah, USA.
Review of putative Stokesosaurus synapomorphies
Benson (2008) considered four characters to be unique synapomorphies shared only by Stokesosaurus clevelandi and “Stokesosaurus” langhami among tyrannosauroids. Where comparisons were possible, these characters were demonstrated to be absent in other tyrannosauroids, including non-tyrannosaurid taxa such as Aviatyrannis, Dilong, and Guanlong. They were regarded, therefore, as unequivocal autapomorphies of a monophyletic genus Stokesosaurus. Recent discoveries have revealed that all four of these characters are problematic because most are more widely distributed among tyrannosauroids, especially an array of basal taxa that have recently come to light (e.g., Xu et al. 2006; Ji et al. 2009; Li et al. 2010). We review each of these characters here.
Iliac blade with semi-oval outline in lateral view.—Benson (2008) did not specifically define this character, but we quantify it here as an iliac blade whose posterior margin is less than half as tall dorsoventrally as the region above the acetabulum, which results in a curved dorsal margin and overall “semioval” appearance. This condition contrasts with the more general theropod morphology in which the ilium is sub-rectangular in shape and nearly as tall posteriorly as above the acetabulum (e.g., Madsen 1976; Colbert 1989; Norell and Makovicky 1999; Peyer 2006; Carrano 2007). This general morphology is present in most tyrannosauroids, including Aviatyrannis (Rauhut 2003a), Guanlong (Xu et al. 2006), and tyrannosaurids (e.g., Brochu 2003). The semi-oval morphology, on the other hand, is indeed present in both S. clevelandi and “S.” langhami (Fig. 1A, B). However, in the course of the present study it was also observed in the basal tyrannosauroids Dilong (IVPP V14243; Xu et al. 2004: fig. 1k), Sinotyrannus (Ji et al. 2009: fig. 3), and Xiongguanlong (Li et al. 2010: fig. 6a in supplementary material). Most notably, Xiongguanlong has an ilium whose proportions and overall shape are remarkably similar to those of both putative species of Stokesosaurus. Therefore, a semi-oval ilium is not a unique synapomorphy of S. clevelandi and “S.” langhami, but is rather a more widely distributed character among basal tyrannosauroids.
Narrow preacetabular notch.—The open space between the preacetabular process and pubic peduncle of the ilium is narrow in both S. clevelandi and “S.” langhami, and it remains narrow across its entire length when seen in lateral view (Fig. 1A, B). In contrast, the notch of other tyrannosauroids is wider, and expands in width as it continues anteriorly. This wide condition is present in Guanlong (Xu et al. 2006: fig. 2e), Aviatyrannis (Rauhut 2003a: fig. 1), Raptorex (Sereno et al. 2009: fig. 2e), and tyrannosaurids (e.g., Gorgosaurus, Lambe 1917: fig. 6; Tyrannosaurus, Brochu 2003: fig. 90), as well as close tyrannosauroid outgroups (e.g., Currie and Chen 2001; Ji et al. 2003; Peyer 2006). Therefore, the narrow morphology is shared only by S. clevelandi and “S.” langhami among tyrannosauroids. We hesitate to consider this a robust synapomorphy, however, as the preacetabular notch is broken in other basal tyrannosauroids sharing overall morphological similarity (e.g., a semi-oval iliac blade, above) with putative Stokesosaurus species: Dilong (IVPP V14243), Eotyrannus (MIWG 1997.550), Sinotyrannus (Ji et al. 2009), and Xiongguanlong (Li et al. 2010). Any of these taxa may have a narrow preacetabular notch, resulting in a wider distribution of this feature. However, the absence of data renders current phylogenetic optimizations preliminary.
Posterodorsally inclined ridge on lateral surface of ilium. —Tyrannosauroids are unusual among theropods in possessing a discrete, linear ridge extending dorsal to the acetabulum on the lateral surface of the iliac blade (e.g., Holtz 2004), which probably served to separate major hindlimb muscles (Carrano and Hutchinson 2002). The ridge projects straight dorsally, or nearly so, on most known tyrannosauroid specimens (e.g., tyrannosaurids, Lambe 1917; Brochu 2003; Aviatyrannis, Rauhut 2003a), but Benson (2008) noted that S. clevelandi and “S.” langhami shared an atypical condition in which the ridge is oriented posterodorsally at a strong angle from the vertical (Fig. 1A, B). Recent re-examination of the holotype specimen of the basal tyrannosauroid Eotyrannus by one of us (SLB) revealed the presence of a fragmentary left ilium that was not described in the initial publication naming this genus (Hutt et al. 2001) and has yet to be mentioned in the literature (MIWG 1997.550) (Fig. 1C). Although fragmentary, the specimen is identified as a left ilium because the base of one of the ventral peduncles is preserved, and it is mediolaterally narrow like the pubic peduncle of tyrannosauroids and other coelurosaurs but unlike the thicker and more conical ischial peduncle (e.g., Brochu 2003; Rauhut 2003b). A linear ridge is present, well preserved, and well developed on the lateral surface of the blade, and it extends strongly posterodorsally at approximately the same angle as seen in Stokesosaurus clevelandi and “S.” langhami. Therefore, a posterodorsally-inclined iliac ridge can no longer be considered as a unique synapomorphy of a monophyletic Stokesosaurus.
Fig. 1.
Ilia of basal non-tyrannosaurid tyrannosauroids with a posterodorsally inclined ridge on the lateral surface of the ilium. A. Right ilium (reversed) of Juratyrant langhami Benson, 2008 (OUMNH J.3311-21), Kimmeridge Clay, Dorset England, Late Jurassic (early Tithonian). B. Left ilium of Stokesosaurus clevelandi, Madsen 1974 (UMNH VP 7473), Morrison Formation, Utah, USA, Late Jurassic (early Tithonian). C. Left ilium of Eotyrannus lengi Hutt, Naish, Martill, Barker, and Newberry, 2001 (MIWG 1997.550), Wessex Formation, Isle of Wight, England, Early Cretaceous (Barremian). All in lateral view. Arrows denote the lateral ridge.
Even if our identification of the posterodorsal ridge is incorrect in Eotyrannus (we could be misidentifying the one partially preserved peduncle as the pubic peduncle; if it is the ischial peduncle then the ridge would be directed strongly anterodorsally), we argue that the orientation of the ridge may not be a systematically robust character. Brusatte et al. (2009) identified an anterodorsally oriented ridge as an autapomorphy of Alioramus altai, but subsequent reexamination of specimens show that this feature is also seen in some, but not all, specimens of Daspletosaurus (CMN 8506), Gorgosaurus (CMN 2120), and Tyrannosaurus (BMR 2002.4.1) (Brusatte et al. 2012), genera which usually have a dorsally directed ridge. Therefore, there is wide variation in the orientation of the ridge in some tyrannosaurid taxa, which suggests that this feature may be too variable to confidently use in taxon diagnoses.
Corrugated structure on the medial surface of the iliac blade, opposite of the ridge on the lateral surface.— Benson (2008) noted that both S. clevelandi and “S.” langhami possess a furrow on the medial surface of the iliac blade, directly corresponding to the shape and position of the linear ridge on the lateral surface. This morphology is not seen in other tyrannosauroids represented by ilia whose medial surfaces are well preserved and visible (i.e, not articulated with the sacrum), including the basal taxon Aviatyrannis (Rauhut 2003a: fig. 1c) and the derived tyrannosaurid Tyrannosaurus (Brochu 2003: fig. 92B). Unfortunately, the medial surface of the ilium cannot be observed, or is not well enough preserved to assess the presence of a corrugated furrow, in any known specimen of the basal tyrannosauroids Dilong, Eotyrannus, Guanlong, Sinotyrannus, and Xiongguanlong. Therefore, we hesitate to consider this a robust synapomorphy of a monophyletic Stokesosaurus.
Revised phylogenetic analysis of tyrannosauroid interrelationships
The above discussion makes clear that the four characters considered by Benson (2008) to be unique synapomorphies of a monophyletic Stokesosaurus are problematic, either because they are now known to be more widely distributed among basal tyrannosauroids or because they cannot confidently be scored in many of the closest relatives of S. clevelandi and “S.” langhami. The strongest arbiter of whether these two species form a monophyletic genus, however, is a numerical cladistic analysis that takes into account as many characters as possible. Recent phylogenetic analyses have included Stokesosaurus as a terminal, but it has usually been scored either as a composite or based solely on “S.” langhami (Benson 2008; Brusatte et al. 2010). S. clevelandi and “S.” langhami were included as separate terminals in the analysis of Choiniere et al. (2010), but this did not include a full array of basal tyrannosauroid taxa or characters relevant to tyrannosauroid ingroup relationships. Therefore, the monophyly of Stokesosaurus has yet to be comprehensively and explicitly tested.
Here, we analyze the phylogenetic relationships of tyrannosauroids and assess the monophyly of Stokesosaurus by including both S. clevelandi and “S.” langhami as terminals in a revised version of the cladistic dataset of Brusatte et al. (2010). This dataset includes every well known tyrannosauroid taxon and a thorough sample of over 300 characters specific to tyrannosauroid ingroup relationships. The original version of the analysis, which represented Stokesosaurus with the character scores of “S.” langhami, resulted in a single most parsimonious tree, with most clades well supported. Thus, it is the most appropriate dataset for testing the monophyly of Stokesosaurus.
We made several small modifications to the original dataset. In addition to separating S. clevelandi and “S.” langhami into distinct terminals, we also added Aviatyrannis as a new terminal, with all character scores based on the description of Rauhut (2003a). Although Aviatyrannis is a fragmentary taxon represented solely by an ilium (we do not score characters based on the isolated ischium referred to the taxon by Rauhut 2003a), it is critical to consider because it shares several characters with one or both putative species of Stokesosaurus, and may be a closely related (or perhaps congeneric) taxon. We also added seven new characters relating to the ilium and dorsal vertebrae, including some of the characters considered by Benson (2008) to be synapomorphies of a monophyletic Stokesosaurus, and modified two character scores for Dryptosaurus based on the recent redescription of this taxon by Brusatte et al. (2011). Full details of the additions and changes to the dataset see Supplementary Online Material (SOM) available at http://app.pan.pl/SOM/app58-Brusatte_Benson_SOM.pdf. The end result is a 25-taxon, 314-character dataset.
The dataset was subjected to a parsimony analysis in TNT v. 1.1 (Goloboff et al. 2008), with the ingroup constrained as monophyletic. First, we analyzed the matrix under the “New Technology search” option, using sectorial search, ratchet, tree drift, and tree fuse options with default parameters. The minimum length tree was recovered in 10 replicates, a process that aimed to sample as many tree islands as possible. This resulted in eight most parsimonious trees (MPTs) of 573 steps (consistency index [CI] = 0.637, retention index [RI] = 0.833). These eight trees were then analyzed under traditional TBR branch swapping to more fully explore each tree island. This resulted in one additional most parsimonious tree. The strict consensus of the nine MPTs is moderately well resolved: the relationships of taxa more derived than Xiongguanlong are fully resolved, but many basal taxa fall into a polytomy at the base of Tyrannosauroidea. One of the clades falling into this polytomy is a trichotomy of S. clevelandi, “S.” langhami, and Eotyrannus.
Examination of the individual MPTs showed that the basal polytomy was due solely to the fragmentary taxon Aviatyrannis, which acts as a wildcard because it can equally parsimoniously occupy several positions among basal Tyrannosauroidea. Therefore, we ran a second analysis in which Aviatyrannis was deleted, using the same search strategy outlined above. This analysis resulted in a single most parsimonious tree (Fig. 2) of 570 steps, with a CI of 0.640 and RI of 0.835. “S.” langhami and Eotyrannus are recovered as sister taxa, with S. clevelandi as their closest outgroup. Therefore, a monophyletic Stokesosaurus was not recovered. The “S.” langhami and Eotyrannus clade, however, is poorly supported, with a Bremer support of 1 and a bootstrap percentage of less than 50%. This is almost certainly due to the fragmentary nature of S. clevelandi, which is only scored for characters relating to the holotype ilium (see Benson [2008] for an explanation of why other material previously referred to S. clevelandi cannot confidently be assigned to this taxon).
To test the robustness of our results, we also ran a modified analysis (excluding the wildcard Aviatyrannis) in which the orientation of the linear ridge on the lateral surface of the ilium is scored as uncertain for Eotyrannus (character 258). This is a conservative score that addresses the possibility that we have misidentified the ridge as extending posterodorsally, which is the score in the original Brusatte et al. (2010) analysis (see above). This modified analysis results in three most parsimonious trees (also of 570 steps, with a CI of 0.640 and RI of 0.835). The strict consensus topology is identical to of the original analysis (Fig. 2), except that Eotyrannus, S. clevelandi, and “S.” langhami fall into a basal polytomy with the clade comprised of Xionggualong and more derived tyrannosauroids. Therefore, while this sensitivity analysis does not indicate a positive grouping of Eotyrannus and “S.” langhami, it does not provide clear evidence for a monophyletic Stokesosaurus either.
Fig. 2.
The phylogenetic relationships of tyrannosauroids, based on a revised analysis of the Brusatte et al. (2010) dataset. Details of the analysis are described in the text and the dataset is presented in SOM. The cladogram shown here is the single most parsimonious tree recovered by the analysis, with the wildcard taxon Aviatyrannis excluded (570 steps, CI = 0.640, RI = 0.835). Numbers next to nodes denote bootstrap percentages (based on 1000 replicates) and Bremer support. Note that Stokesosaurus clevelandi and “S.” langhami (here referred to by its new genus name, Juratyrant) are not found as sister taxa, and therefore a monophyletic Stokesosaurus is not recovered. When Aviatyrannis is included in the analysis, the strict consensus of nine most parsimonious trees (not figured) shows identical and fully resolved relationships among Xiongguanlong and all more derived taxa. However, Stokesosaurus clevelandi, “S.” langhami, and Eotyrannus, form a polytomy. This clade, in turn, is part of a large basal polytomy that also includes the Xiongguanlong + more derived clade, Dilong, Aviatyrannis, Guanlong, Kileskus, Proceratosaurus, and Sinotyrannus. On the figured cladogram, the following unambiguous synapomorphies support major clades, with character numbering following that in the character list of Brusatte et al. (2010) and SOM: all tyrannosauroids more derived than Dilong (33, 41, 49, 80, 180, 181, 196, 198, 221, 239, 241, 244, 257, 274, 281, 289, 290); the clade of S. clevelandi, Juratyrant, and Eotyrannus (258, 310, 311, 313); the clade of Juratyrant and Eotyrannus (no unambiguous synapomorphies).
Systematic revisions
In summary, because (i) there are no autapomorphies or unique combination of characters that unite Stokesosaurus clevelandi and “Stokesosaurus” langhami relative to other basal tyrannosauroids (especially phylogenetically proximal taxa such as Eotyrannus, Xiongguanlong, and Aviatyrannis); and (ii) a phylogenetic analysis does not recover S. clevelandi and “S.” langhami forming a clade exclusive of other taxa, then a new generic name must be erected for “S.” langhami. Here, we provide a systematic revision of S. clevelandi and “S.” langhami, present updated diagnoses, and name a new genus for “S.” langhami.
Systematic palaeontology
Theropoda Marsh, 1881
Tetanurae Gauthier, 1986
Tyrannosauroidea Osborn, 1905 (sensu Sereno et al. 2005)
Genus Stokesosaurus Madsen, 1974
Type species: Stokesosaurus clevelandi Madsen, 1974; see below.
Diagnosis.—Same as for the type and only known species.
Stokesosaurus clevelandi Madsen, 1974
Holotype: UMNH 2938 (formerly UUVP 2938), a left ilium.
Type horizon: Brushy Basin Member of the Morrison Formation, lower Tithonian, Upper Jurassic.
Type locality: Cleveland-Lloyd Dinosaur Quarry, Utah, USA.
Emended diagnosis.—Tyrannosauroid theropod with a single autapomorphy: a swollen rim around the articular surface of the pubic peduncle, which is especially prominent on the medial surface (Benson 2008). Furthermore, S. clevelandi can be differentiated from other phylogenetically proximal tyrannosauroids by a unique combination of characters: an anteroposteriorly thick ridge on the lateral surface of the ilium which projects posterodorsally and extends to the dorsal margin of the iliac blade.
Remarks.—The sole autapomorphy of Stokesosaurus clevelandi is absent in “Stokesosaurus” langhami (Benson 2008), Aviatyrannis (Rauhut 2003a), Guanlong (Xu et al. 2006; IVPP V14531), Raptorex (Sereno et al. 2009; LH PV18), and tyrannosaurids (e.g., Gorgosaurus, Lambe 1917; Tyrannosaurus, Brochu 2003). It appears to be absent in the type and only known specimen of Xiongguanlong (FRDC-GS JB16-2-1), but there is some breakage in this region. This character cannot be assessed in Dilong (IVPP V14243) and Eotyrannus (MIWG 1997.550) because the pubic peduncle is damaged in all known specimens.
The unique combination of characters differentiates S. clevelandi from “S.” langhami and all other tyrannosauroids. A thickened lateral ridge, defined here (and used in the phylogenetic analysis) as a ridge with an anteroposterior width greater than 20% of its dorsoventral height, is also present in Guanlong (Xu et al. 2006), Sinotyrannus (Ji et al. 2009), and tyrannosaurids (e.g., Brochu 2003). A thin ridge, however, is present in “S.” langhami (Benson 2008), Dilong (IVPP V14243), Eotyrannus (MIWG 1997.550), Aviatyrannis (Rauhut 2003a), and Xiongguanlong (Li et al. 2010). The lateral ridge extends to the dorsal margin of the ilium in Guanlong, Sinotyrannus, and Aviatyrannis, whereas it stops short of the dorsal margin in “S.” langhami, Eotyrannus, Xiongguanlong, and tyrannosaurids. Finally, the posterodorsally oriented ridge is present in “S.” langhami and Eotyrannus (see above). S. clevelandi, therefore, is the only tyrannosauroid that possesses a combination of a thick, posterodorsally-trending ridge that extends to the dorsal margin of the ilium.
Fig. 3.
Autapomorphies of tyrannosauroids Juratyrant langhami Benson, 2008, Kimmeridge Clay, Dorset England, Late Jurassic (early Tithonian) (A, B) and Stokesosaurus clevelandi Madsen, 1974, Morrison Formation, Utah, USA, Late Jurassic (early Tithonian) (C). A. Right pubis (OUMNH J.3311-22) in lateral view, with the autapomorphic lateral fossa denoted. B. Left ischium (OUMNH J.3311-25) in lateral (B1) and anterior (B2) views, with an inset close up (2.5× magnification) of the autapomorphic folded proximal region of the ischial apron (B3). The autapomorphic convex ischial tubercle is also denoted. C. Pubic peduncle of the left ilium (UMNH 2938) in medial view, with the autapomorphic swollen rim indicated.
Genus Juratyrant nov.
Type species: Juratyrant langhami Benson, 2008; see below. Etymology: “Jura” refers to the Jurassic age of the taxon and “tyrant” is an Anglicized version of the Greek “tyrannos” and Latin “tyrannus,” in reference to the vernacular characterization of tyrannosauroids as “tyrant dinosaurs” (based on the original etymology of Tyrannosaurus rex).
Diagnosis.—Same as for the type and only known species.
Juratyrant langhami Benson, 2008 Figs. 1A, 3A, B.
Holotype: OUMNH J.3311-1–J.3311-30, an associated partial skeleton from a mature individual (see Benson [2008] for details). Individual bones include: one cervical vertebra (OUMNH J.3311-1); five dorsal vertebrae (OUMNH J.3311-2–J.3311-5 and J.3311-30); a complete sacrum (OUMNH J.3311-6–J.3311-9); five caudal vertebrae (OUMNH J.3311-10–J.3311-14); four isolated vertebral transverse processes (OUMNH J.3311-16–J.3311-19); the complete pelvic girdle (left ilium, J.3311-20; right ilium, J.3311-21; right pubis, J.3311-22; left pubis, J.3311-23; right ischium, J.3311-24; left ischium, J.3311-25); both femora (left femur, J.3311-26; right femur, J.3311-27); both tibiae (right tibia, J.3311-28; left tibia, J.3311-29); and an unidentified bone fragment (OUMNH J.3311-15).
Type horizon: The Pectinatites pectinatus Ammonite Zone, P. eatlecottensis Subzone, Kimmeridge Clay, Upper Jurassic: lower Tithonian.
Type locality: Dorset, England, United Kingdom. The specimen was recovered 6 miles west of Swanage between Rope Lake Head and Freshwater Steps (marked as Kimmeridge Ledges on Ordnance Survey maps, Ordinance Survey, 1979).
Emended diagnosis.—Tyrannosauroid theropod with four autapomorphies: ischial apron with a “folded” appearance (Benson 2008); a fibular flange that continues as a distinct low ridge to the proximal end of the tibia (Benson 2008); ischial tubercle of the ischium expressed as a convex bulge (Brusatte et al. 2010); deep fossa on the lateral surface of the pubis ventral to the acetabulum (new character). Furthermore, Juratyrant langhami possesses two probable autapomorphies, which are difficult to assess in other taxa because of damage or non-preservation of the bone in question: a prominent hyposphene that extends posteriorly as a thin sheet on the fifth sacral vertebra (Benson 2008) and an extensor groove of the femur expressed as a broad, concave outline in distal view (Brusatte et al. 2010).
Remarks.—Benson (2008) considered a “folded” ischial apron (figured here in Fig. 3B) and a proximally extensive fibular flange as autapomorphies of Juratyrant langhami, and we confirm that these still remain unique to this taxon among all known tyrannosauroids. Benson (2008) further regarded a prominent hyposphene on the fifth sacral vertebra to be unusual to J. langhami, and we tentatively consider this an autapomorphy here but note that it is difficult to assess in most other tyrannosauroids, especially basal taxa phylogenetically proximal to J. langhami (e.g., Dilong, Eotyrannus, Raptorex, Sinotyrannus, Xiongguanlong). Guanlong also possesses a prominent hyposphene on the fifth sacral, and it projects even further posteriorly relative to the centrum face than in J. langhami, but it is not sheet-like as in J. langhami (IVPP V14531). Tyrannosaurids, on the other hand, do not possess a prominent hyposphene that projects far posterior to the centrum face (e..g., Alioramus IGM 100/1844, Brusatte et al. 2012; Tyrannosaurus, Brochu 2003).
We note three additional autapomorphies, two of which are definitive and one of which is probable. First, the ischial tubercle of J. langhami is present as a convex bulge on the posterior surface of the ischium. In Guanlong (IVPP V14531) and outgroup taxa it is expressed as a groove, whereas in more derived taxa such as Dryptosaurus (Brusatte et al. 2011), Raptorex (Sereno et al. 2009), Appalachiosaurus (Carr et al. 2005), and tyrannosaurids (e.g., Lambe 1917; Brochu 2003) it is present as a discrete, either ovoid or triangular, flange whose rugose lateral surface is depressed relative to the remainder of the ischium. J. langhami is the only taxon with a tubercle expressed as a bulge, which is not depressed or discretely offset from the posterior margin of the ischium. This unusual condition was noted by Brusatte et al. (2010) and treated as an intermediate morphology between the groove-like and flange-like states in an ordered character statement (character 278).
Second, we note that there is a deep fossa on the lateral surface of the pubis ventral to the acetabulum, which is bordered anteriorly by a stout ridge (Fig. 3A). This ridge separates the fossa from the rugose pubic tubercle on the anterior surface of the pubis. The fossa and corresponding ridge are absent in all other tyrannosauroids known from well preserved pubes, including Guanlong, Raptorex, and tyrannosaurids (e.g., Brochu 2003).
Third, J. langhami possesses a uniquely-shaped extensor groove on the anterior surface of the distal femur, in which the groove is present, but shallow, and expressed as a broad concave margin in distal view (Benson 2008: fig. 11F). In more basal taxa such as Guanlong (IVPP V14531) and Dilong (IVPP V14243) the extensor groove is absent and the anterior surface of the femur is flat, and in more derived taxa (Xiongguanlong, Dryptosaurus, Raptorex, tyrannosaurids) the groove is present and expressed as a deep, U-shaped cleft in distal view. J. langhami, therefore, is unique in possessing a shallow and broad extensor groove, and this was noted by Brusatte et al. (2010) who scored J. langhami for its own intermediate character in an ordered character statement related to the presence and depth of the groove (character 290). We acknowledge, however, that the femora of OUMNH J.3311 (the holotype and only specimen of J. langhami) are deformed; the right femur is crushed mediolaterally and does not exhibit an extensor groove whereas the left is crushed anteroposteriorly and shows a broadly curved groove (Benson 2008). Thus, this feature is only proposed hesitantly as an autapomorphy.
Finally, Benson (2008) described Juratyrant langhami as possessing an autapomorphic condition of the posterior dorsal vertebrae, in which the postzygapophyses are reduced and raised dorsally relative to the prezygapophyses (Benson 2008: fig. 3). This condition is not present in the basal tyrannosauroid Guanlong (IVPP V14531), but is present in Xiongguanlong (Li et al. 2010; FRDC-GS JB16-2-1), Raptorex (Sereno et al. 2009; LH PV18), and tyrannosaurids (e.g., Alioramus, IGM 100/1844; Tarbosaurus, Maleev 1974; Tyrannosaurus, Brochu 2003). We include this character in our revised phylogenetic analysis and recover the derived state (dorsally elevated postzygapophyses) as a synapomorphy of the clade of all tyrannosauroids more derived than, and including, J. langhami.
Conclusions
Specimens such as OUMNH J.3311-1–J.3311-30 and the various fossils of Stokesosaurus clevelandi are important, as they represent Late Jurassic tyrannosauroids that are intermediate in phylogenetic position and body size between small basal tyrannosauroids (e.g., Dilong and Guanlong) and the largest and latest-surviving tyrannosaurids (e.g., Albertosaurus and Tyrannosaurus). Understanding the systematics of these specimens, however, is challenging because of their fragmentary nature. We here show that OUMNH J.3311-1–J.3311-30, which Benson (2008) described as the holotype of a new species of Stokesosaurus (S. langhami), does not share any unequivocal synapomorphies with the type species of Stokesosaurus (S. clevelandi). Furthermore, the two do not group together in a phylogenetic analysis. Therefore, we erect a new genus name for the British material, Juratyrant. Hopefully, as more complete specimens of S. clevelandi and other Jurassic tyrannosauroids are found, the systematics and phylogeny of these so-called “intermediate tyrannosauroids” will become better understood.
Acknowledgements
We thank Thomas Carr (Carthage College, Kenosha, Wisconsin, USA) and Xu Xing (IVPP) for their helpful reviews and several curators and collections managers for access to specimens in their care, including: Magdalena Borsuk-Białynicka (Institute of Paleobiology, Polish Academy of Sciences, Warsaw, Poland), Philip Currie (University of Alberta, Edmonton, Canada), John Horner (Museum of the Rockies, Bozeman, Montana, USA), Steve Hutt (Dinosaur Isle Museum, Sandown, UK), Carl Mehling (AMNH, New York, NY, USA), Paul Sereno (University of Chicago, Chicago, IL, USA), and Xu Xing. Conversations about tyrannosauroids with Thomas Carr, Jonah Choiniere (Bernard Price Institute, Johannesburg, South Africa), Philip Currie, Thomas Holtz (University of Maryland, College Park, MD, USA), Mark Loewen (Natural History Museum of Utah, Salt Lake City, UT, USA), Pete Makovicky (Field Museum of Natural History, Chicago, IL, USA), Mark Norell (AMNH, New York, NY, USA), Paul Sereno, Tom Williamson (New Mexico Museum of Natural History and Science, Albuquerque, NM, USA), and Xu Xing were helpful and insightful. RBJB's research is supported by a fellowship at Trinity College, Cambridge. SLB is supported by a National Science Foundation Graduate Research Fellowship and Doctoral Dissertation Improvement Grant (NSF DEB 1110357), and his visits to England and Utah was supported by the American Museum of Natural History, Division of Paleontology (administered by Mark Norell).
