Triceratops, Marsh, 1889

Erickson, Gregory M., Van Kirk, Samuel D., Su, Jinntung, Levenston, Marc E., Caler, William E. & Carter, Dennis R., 1996, Bite-force estimation for Tyrannosaurus rex from tooth-marked bones, Nature 382 (6593), pp. 706-708 : 706-707

A recently unearthed Triceratops sp. pelvis from the Hell Creek Formation of Montana (USA) bears dozens of large bite marks ( Fig. 1 )8.

Casts ofsome of the deeper punctures show that an adult T. rex produced the marks using its longer anterior caniniform teeth8. The bitten bones are predominantly composed of cancellous bone tissue, capped only by a thin layer of dense cortical bone8. On the basis of these marks, it is difficult to gauge whether the teeth that produced the bite marks were particularly robust. We attempted to quantify the forces that the tyrannosaur dentition absorbed when biting the Triceratops ilium, by using laboratory simulations. W e contrasted the results with those for extant taxa to place them in a comparative context, and assessed the functional and behavioural implications of these comparisons.

Using histological examination we determined that extant bovine ilia exhibit comparable microstructure to Triceratops ilia. Consequently, bovine ilia were used to model the bitten Triceratops bones. Sections of ili a with cortices ofvarying thickness were penetrated with a T rex tooth replica to a depth of 11.5 mm (the depth of the deepest ilium bite mark8) using a servohydraulic mechanical loading frame. The forces produced throughout these simulations were recorded. When indented, the bovine ilia exhibited localized crushing as the only mode of failure, and the punctures produced were comparable in morphology to the T rex bite marks. The forces during testing increased with increasing penetration depths ( Fig. 2 ). Peak forces ranged from 1,900 to 15,100 N ( Fig. 3 ). A positive correlation between peak penetration force and cortical thickness was found ( Fig. 3 ).

A bone sample removed from the bitten Triceratops ilium within 2cm of the deepest bite mark (11.5mm) revealed a cortical thickness of 2.5 mm. From a linear regression of our data ( Fig. 3 ), we determined that roughly 6,410N of force was required to produce the bite mark. E stimates as great as 13,400 N for posterior teeth were obtained when biting velocity, energy absorptio n by flesh, and the mechanical advantage of poste rior teeth relative to more anterior teeth were taken into consideration ( Fig. 3 ).

These bite-force estimates make it possible to evaluate speculations on tyrannosaur tooth strength and potential behaviours using comparisons with extant taxa. The largest m aximum bite force measureme nts or estimates for extant vertebrates at posterior tooth positions are: 550 N for labrador dogs9, 749 N for humans10, 1,412 N for wolves11, 1,446 N for dusky sharks (location of force measurement within jaw not given)n, 1,712 N for orangutans1.1, 4,168 for lions11, and 13,300 N for American alligators14 (K. A Vliet, personal communication). Using bite force as a relative indicator of dental strength, the results suggest that T rex teeth were as strong as, or in most cases substantially stronger than, those of any extant taxa tested to date. Consequently speculations that their dentition was mechanically weak were not supported.

Peak bite-force estimates for large American alligators ( Alligator mississipiensis ) are within the range we calculated for a feeding T. rex . This taxon shares many dental attributes with T. rex (including thecodont implantation15,16, stout semi-sharp caniniform teeth that are transversely rounded6,7, 16, and nearly identical histological structures17, 18). These morphological similarities imply similarity in function19 Alligators use their teeth to procure large prey and to engage conspecifics during confrontations20 Both activities demand teeth that can sustain large compression and bending forces, particularly because impacts with bones are frequent20 The bite-force estimates and tooth mark evidence show that T. rex teeth could similarly withstand large bite forces and sustain repetitive bone impacts. Therefore, it is not unreasonable to suspect that the T. rex dentition could be used in behaviours similar to those of alligators, and with some mechanical safety21 Physical evidence supports this reasoning. Bony calli on adult tyrannosaur crania attest to biting injuries during intraspccific aggression22, 23, and a healed hadrosaur tail injury has been attributed to biting by a T. rex during a failed predation attempt5.

Although our data suggest that T. rex could produce enormous bite forces and possessed a dentition that could endure stresses associated with prey struggles, they by no means prove that T. rex was predacious. Indeed it could be argued that these characteristics enhanced their utilization of scavenged carcasses. Nevertheless, these results refute assertions that T. rex was mechanically limited by its dentition to scavenging carrion. We contend that if T. rex could consistently manoeuvre into a position to engage prey with its dentition, it could have exploited a predatory niche.

It has been shown recently that theropod bite marks are much more common in the fossil record that was once suspected8 24 25 Consequently, the methods used here could be used to assess biteforce estimates for other tyrannosaur individuals, as well as for many theropod species. Such data would greatly augment our understanding of dinosaur tooth form and function, the physical capacities of their teeth and jaws (ontogenetically and interspecifically), and provide new insight into the musculoskeletal biomechanics of dinosaur crania.

Figures

FIG. 1 Triceratops sp. pelvis inventrolateralview bearing bite marks from an adult Tyrannosaurus rex. The sacrumand left ilium (Museumofthe Rockies specimen MOR 799, Montana State University, Bozeman, Mn have 58 definitive bite marks attributable to 'puncture and pull' biting behaviour by the feeding tyrannosaur(s)8 • Arrows denote some of the more conspicuous bite marks. Brackets bound a region where the tyrannosaur(s) removed approximately one-sixth of the anterior portion of the ilium by means of repetitive biting. Scale bar, 25cm.

FIG. 2 Typical force against penetration curves produced during the penetration of bovine ilia by an adult Tyrannosaurus rex tooth replica. The two curves represent simulations conducted on samples with 2.5-mm thick cortices.

FIG. 3 Maximum penetration force values for an adult Tyrannosaurus rex tooth replica impacted through bovine ilia with varying cortical thickness. Peak penetration forces increased with increasing cortical thickness (y = 2305.402x + 646.634, r2 = 0.91). METHODS Because of the 7-fold range of peak force values in the simulations, it was necessary to obtain precise measurements of the cortical depths penetrated by the T. rex teeth. A bone sample taken adjacent to the deepest bite mark revealed a 2.5 mm cortical thickness, From the regression equation, approximately 6,410 N offorce was required to produce the bite mark. This bite mark was made by one of the tyrannosaur's longer caniniform teeth8, probably a tooth between the fourth and seventh maxillary positions (based upon American Museum of Natural History specimen AMNH 5027, New York). To account for the relative mechanical advantage of more posteriorly positioned teeth27, moment calculations were used to calculate the simultaneous forces produced at the most posterior tooth positions. Values ranging from 7,870-10,300 N were projected, assuming 6,410 N of force were produced simultaneously by teeth from the fourth to seventh tooth positions. Because bone strength increases with strain rate28 and the penetration rate of the tooth replica was just 1 mm s- 1, it is likely that the simulation force values underestimated actual forces. A tooth-impact velocity of 10 mm s-1 for a biting tyrannosaur (based on extant large reptile feeding; G.M.E., personal observations) would have required ~20% more bite force28 • Adhering flesh8 may have absorbed another 10% (or more) of the initial bite force29. These considerations suggest that bite forces as high as 13,400 N could have been produced by an adult T. rex during feeding. Greater forces may have been possible during snapping bites or those involving bodily inertia to augment tooth penetration. Such biting is characteristically used when prey are seized initially (G.M.E., personal observations of reptilian feeding). Taphonomic interpretations suggest that the bite marks on the Triceratops ilium were not the result of this behaviour8. Additionally, if the tyrannosaur's contra lateral teeth were used when the deepest bite mark was made, greater bite forces may have been generated than those we estimated13·30. This is indeterminable from MOR 799.