The ceratopsid dinosaurs, a group of large herbivorous dinosaurs from the Late Cretaceous of western North America, are known for their unusual cranial ornamentation. Ceratopsid skulls are typically characterized by a caudally projecting bony frill composed of the parietal and squamosal bones, as well as a combination of dorsally projecting horns above the eyes and external nares. The supraorbital horns are derived from the postorbital bones, and the horn over the nares is an outgrowth of the nasal bones, augmented by a separate ossification in chasmosaurine ceratopsids (a clade including Chasmosaurus, Triceratops, and others). Horn size varies both within and between taxa. For instance, Centrosaurus possesses an elongate nasal horn, and brow horns are small or absent. Triceratops has two elongate brow horns, whereas the nasal horn is reduced in size relative to that of Centrosaurus. These varied horn morphologies have invited a tremendous body of speculation over horn use.
Proposed uses for ceratopsid horns include intraspecific display and combat (Farlow and Dodson 1975; Lull 1933; Sampson et al. 1997), defense against predators (Hatcher et al. 1907; Lull 1933), thermoregulation (Barrick et al. 1998), and as aids in knocking down vegetation (Tait and Brown 1928). Images of two male Triceratops with horns locked in struggle over a group of females are particularly compelling, but largely speculative. The use of horns in intraspecific combat has received special attention in both the popular press and the scientific literature (e.g., Farlow and Dodson 1975; Lull 1933). Most cranial abnormalities, such as anomalous fenestrae in the frill, have been attributed to wounds received from intraspecific agonistic behavior (Lull 1933).
Previous workers have reviewed ceratopsian adaptations for combat as well as their similarities to modern horned animals (Farlow and Dodson 1975; Molnar 1977), but none have tested the physical possibility of combat in these animals. Farlow and Dodson (1975) divided ceratopsian combat into three categories, following a model for ungulate horn evolution proposed by Geist (1966). In brief, animals with type 1 combat behavior stood parallel and delivered blows against each other's flanks. Farlow and Dodson (1975) considered the small hornless neoceratopsian Protoceratops and the American mountain goat to exemplify this behavior. In type 2 combat behavior, individuals locked horns and wrestled; Triceratops, some chameleons, and domestic cattle were thought to share this behavior pattern. Animals using type 3 combat behavior, inferred in Centrosaurus and the modern rhinoceros, may also clash horns or use them to inflict injury. All three of these categories included the possible use of horns and frills as display organs. Molnar (1977) summarized cranial structures expected in animals engaging in horn locking behavior. These include firm support for the horncores; projection of the horns beyond the snout; protection of the jaw muscles, ears, and eyes; stiffening of the cervical vertebral series; and evidence of cranial puncture wounds. Because all of these features occur in ceratopsids, Molnar concluded that horn locking behavior was possible in Triceratops and many of its close relatives.
Alexander (1989) illustrated two Triceratops models locking horns; he concluded that their horns could interlock, but they may not have been strong enough for intraspecific combat (based on comparisons of horn cross-sectional area versus body mass in Triceratops and a variety of horned mammals). Farlow (1990) disputed Alexander's (1989) claim about Triceratops horn strength, using comparisons of horn cross-sectional area versus horn reach for Triceratops and horned mammals. He also briefly commented on orientation differences between antelope horns and ceratopsid horns. Yet, neither Alexander (1989) nor Farlow (1990) fully investigated all possible horn locking configurations and their behavioral implications.
The present study uses scale models to test the assumption that the horned dinosaur Triceratops could lock horns with its conspecifics. I seek to test the feasibility of horn locking behavior in this animal, not to prove that such behavior actually occurred. Additionally, this study does not consider or test other possible uses for the horns (such as predator defense or intraspecific recognition).
Triceratops was chosen as a study subject for three reasons. First, it is one of the best-known horned dinosaurs, with a number of well-preserved, uncrushed skulls in museum collections. Second, its horn morphology is representative of many other chasmosaurine ceratopsids. Finally, accurate scale models of Triceratops skulls were readily available for use in this project.
Abbreviations used in the text are as follows: AMNH: American Museum of Natural History, New York, New York. MPM: Milwaukee Public Museum, Wisconsin. ROM: Royal Ontario Museum, Toronto, Ontario. SDSM: South Dakota School of Mines and Technology Museum of Geology, Rapid City. SMM: Science Museum of Minnesota, St. Paul. SMNH: Saskatchewan Museum of Natural History, Regina. TMP: Royal Tyrrell Museum of Paleontology, Drumheller, Alberta. USNM: National Museum of Natural History, Washington, D.C. YPM: Yale Peabody Museum of Natural History, New Haven, Connecticut.