Technical Comments

Response to Comment on “A Southern Tyrant Reptile”

See allHide authors and affiliations

Science  27 Aug 2010:
Vol. 329, Issue 5995, pp. 1013
DOI: 10.1126/science.1190195

Abstract

Herne et al. doubt our identification of tyrannosauroid pubes from the Lower Cretaceous of Australia. They suggest that the fossil is broken and can be identified only as an indeterminate neotetanuran (representing the wider clade that includes coelurosaurs such as birds and tyrannosauroids, and also more primitive large theropods, the allosauroids). However, we maintain that unique tyrannosauroid features are clearly preserved or interpretable.

We reported an Australian tyrannosauroid, represented by a pubis from the late Early Cretaceous [National Museum of Victoria (NMV) P186046] (1). Herne et al. (2), however, argue that the pubic tubercle is broken and cannot provide evidence for tyrannosauroid affinities. We described the base of the broken pubic tubercle as indicative of an anterolaterally curving morphology, similar to that of tyrannosaurids (3) [figure 1D in (1)]. In NMV P186046, the broken surface of the tubercle faces laterally, and the bone surfaces on either side of the break curve anterolaterally (Fig. 1). Furthermore, the distal portion of the tubercle is preserved and projects laterally (Fig. 1). Thus, an anterolaterally directed, “flangelike” pubic tubercle was present. Although this has not been included in published phylogenetic analyses, it is absent in all non-tyrannosauroid theropods with pubes similar to NMV P186046 [figure S1 in (1)] and thus indicates tyrannosauroid affinities. A well-defined depression is located posterior to the pubic tubercle (Fig. 1A). Herne et al. (2) state that this depression is smooth, unlike in tyrannosaurids in which it is rugose [figure 1D in (1)] (3). However, rugosity is present proximally (Fig. 1) and is independent of irregular patches of superficial damage. A well-defined, rugose depression adjacent to the tubercle is also present in tyrannosaurids but is absent in other theropods [figure S1 in (1), (3)].

Fig. 1

Proximal end of the right pubis of NMV P186046 in lateral (A) and anterior (B) views. Scale bar, 20 mm.

The transversely narrow pubic boot of NMV P186046 indicates placement within Coelurosauria [e.g., (4, 5)], the wider clade including Tyrannosauroidea. Even the largest coelurosaur, Tyrannosaurus, has a narrow pubic boot, with a length-to-width ratio of 5.3 (3). This is similar to the ratio of 6.6 in NMV P186046, which is narrow compared with ~2.4 in non-coelurosaurian neotetanurans (allosauroids) with boots of comparable size (6). Herne et al. (2) suggest that NMV P186046 is incomplete distally and may have been wide before breakage. However, to be proportionally broad, as in allosauroids, NMV P186046 must have been 58 mm wide, almost three times the preserved width of 21 mm. This requires an eccentric morphology not documented in any other theropod and is therefore unlikely. Furthermore, in basal neotetanurans, the entire pubis is transversely broad [e.g., (6, 7) and figure S1 in (1)], not just the pubic boot. NMV P186046 is slender [figure 1C in (1)], as in other coelurosaurs (1, 35). Among coelurosaurs, only tyrannosaurids have a large and anteriorly expanded pubic boot similar to that of NMV P186046 [e.g., (1, 3, 8, 9)]. Based on this observation and the morphology of the pubic tubercle, we maintain that NMV P186046 can be identified as a tyrannosauroid.

Finally, Herne et al. (2) state that because a long history of collecting has not uncovered any Gondwanan tyrannosauroids, their presence is unlikely. There are three problems with this claim. First, any hypothesis of tyrannosauroid absence should be based on evidence, not expectation. We have made the case, here and previously (1), that the anatomical evidence for an Australian tyrannosauroid is clear. Second, the Gondwanan dinosaur record is poorly sampled in comparison with that from Laurasia. New discoveries and more critical paleobiogeographic analyses have shown that the hypothesis of distinct “Laurasian” and “Gondwanan” dinosaur faunas during the Early Cretaceous (10) is too simplistic. Many “endemic” clades are now known from both hemispheres [e.g., “Gondwanan” abelisauroids (11) in Europe and “Laurasian” dromaeosaurids in South America (e.g., 12)]. Third, of the Australian theropod clades cited by Herne et al. (2), only the carcharodontosaurian Australovenator is represented by substantially complete remains (13). These are of similar age to NMV P186046 and indicate that it is the sister taxon of Fukuiraptor from Japan (14). This demonstrates that closely related theropod taxa can be present in both Laurasian and Gondwanan faunas during the middle Cretaceous. Thus, the presence of Australian tyrannosauroids is not implausible and should not be rejected a priori.

References and Notes

View Abstract

Navigate This Article