Dinosaurian and Mammalian Predators Compared Provided by ZENODO

Home , Aublysodon, Borogovia, Caenagnathus, Ceratosaurus, Chirostenotes, Coelurus

Paleobiology, 28(4), 2002, pp. 527–543

View metadata, citation and similar papers at core.ac.uk brought to you by CORE Dinosaurian and mammalian predators compared provided by ZENODO

Blaire Van Valkenburgh and Ralph E. Molnar

Abstract.—Theropod dinosaurs were, and mammalian carnivores are, the top predators within their respective communities. Beyond that, they seem distinct, differing markedly in body form and an- cestry. Nevertheless, some of the same processes that shape mammalian predators and their communities likely were important to dinosaurian predators as well. To explore this, we compared the predatory adaptations of theropod dinosaurs and mammalian carnivores, focusing primarily on aspects of their feeding morphology (skulls, jaws, and teeth). We also examined suites of sympatric species (i.e., ecological guilds) of predatory theropods and mammals, emphasizing species richness and the distribution of body sizes within guilds. The morphological comparisons indicate reduced trophic diversity among theropods relative to carnivorans, as most or all theropods with teeth appear to have been hypercarnivorous. There are no clear analogs of felids, canids, and hyaenids among theropods. Interestingly, theropods parallel canids more so than felids in cranial proportions, and all theropods appear to have had weaker jaws than carnivorans. Given the apparent trophic similarity of theropods and their large body sizes, it was surprising to ﬁnd that species richness of theropod guilds was as great as or exceeded that observed among mammalian carnivore guilds. Separation by body size appears to be slightly greater among sympatric theropods than carnivorans, but the magnitude of size difference between species is not constant in either group. We suggest that, as in modern carnivoran guilds, smaller theropod species might have adapted to the threats posed by much larger species (e.g., tyrannosaurs) by hunting in groups, feeding rapidly, and avoiding encounters whenever possible. This would have favored improved hunting skills and associated adaptations such as agility, speed, intelligence, and increased sensory awareness.

Blaire Van Valkenburgh. Department of Organismic Biology, Ecology, and Evolution, University of Cali- fornia, Los Angeles, California 90095-1606. E-mail: [email protected] Ralph E. Molnar. Museum of Northern Arizona, 3101 North Fort Valley Road, Flagstaff, Arizona 86001

Accepted: 4 June 2002

Introduction plained as adaptations to minimize competition. Thus sympatric large carnivores usually Tyrannosaurus rex and the lion, Panthera leo, would seem to have little in common other differ in body size, dental morphology, or than both being the largest carnivores in their skeletal anatomy (see Eaton 1979; Van Valken- respective ecosystems. Among their many dif- burgh 1985, 1988; Dayan et al. 1990, 1992). It ferences, T. rex was a five-ton biped with re- might be expected that sympatric theropods duced forelimbs and multiple serrated teeth; diverged in a similar fashion if they faced the other, a 100-kg quadruped with muscular, comparable pressures from interspecific com- clawed forelimbs and teeth specialized for petition. distinct functions, such as killing or slicing. In this paper, we first review the relevant lit- Nevertheless, there is reason to believe that erature on mammalian carnivores to establish some of the same processes that shape mam- the background against which theropods will malian predators and their communities were be viewed. We then compare some of the pred- important to dinosaurian predators as well. atory adaptations of theropod dinosaurs and Interspecific competition, for example, ap- mammalian carnivores, focusing primarily on pears to play a significant role in determining aspects of their feeding morphology (skulls, the abundance and distribution of many large jaws, and teeth). This is followed by an ex- extant carnivores. For instance, by choosing to amination of suites of sympatric species (i.e., hunt at a different place or time, coyotes avoid ecological guilds) of predatory theropods and wolves, cheetahs avoid lions, and leopards mammals, emphasizing species richness and avoid tigers (Van Valkenburgh 2001). Associ- the distribution of body sizes within guilds. ated with these behavioral differences are The results presented here are not meant to be morphological differences that can be ex- conclusive, as they are based on a limited sam-

᭧ 2002 The Paleontological Society. All rights reserved. 0094-8373/02/2804-0007/$1.00 528 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR ple of mammals and theropods. Moreover, the ture on interspecific interactions among ex- gulf between carnivoran morphology and the- tant large carnivores found ample evidence of ropod morphology is large and this makes carcass theft, intraguild predation, and inter- quantitative comparisons difficult. For exam- specific avoidance in tropical and temperate ple, theropods differ from carnivorans in limb ecosystems (Palomares and Caro 1999; Van stance (biped vs. quadruped), masticatory Valkenburgh 2001). Four key points are sum- musculature, tooth replacement, and skull marized here. First, most interspecific inter- construction. Nevertheless, similar biome- actions between predators occur as contests chanical and ecological constraints are ex- for the possession of a kill and these contests pected to have been important to both kinds can be frequent. For example, in areas of high of predators. This paper is intended as no spotted hyena density, African wild dogs lost more than a preliminary exploration of tro- 60–86% of their kills to hyenas (Kruuk 1972; phic divergence among coexisting theropods, Fanshawe and Fitzgibbon 1993). As is true of and it is our hope that it will inspire further, carcass theft, the motivation for intraguild more-detailed studies. predation appears to be hunger in many in- stances. However, equally or more often, the The Guild of Large, Predatory Mammals victim is not eaten and the likely motivation is The ecological concept of guild was first de- to remove a competitor who might also prey fined by Root (1967: p. 335) as ‘‘a group of spe- on young. Second, body size is the usual de- cies that exploit the same class of environmen- terminant of rank within the guild; larger spe- tal resources in a similar way.’’ As such, species tend to dominate smaller ones (e.g., lion cies within a guild are expected to compete over hyena in Africa or wolf over coyote in more intensely with each other than with North America). Third, the body size rule can those outside the guild. In theory, guilds are be overturned by the smaller species acting as not limited by taxonomic boundaries; an ex- a group (e.g., hyenas vs. lions; African wild tant predator guild could include reptiles, dogs vs. hyenas). Fourth, intraguild predation birds, and mammals, for example (see Jaksic and kleptoparasitism occur in both forested et al. 1981). However, in practice, they often and open environments. Although these types are so limited for various reasons, such as to of interaction have been observed less fre- maximize the potential role of competition or quently in forested environments, the behav- to facilitate the study (for review of the guild ior of some species strongly suggests that they concept, see Simberloff and Dayan 1991). have had a significant impact. For example, Large, predatory mammals form a guild in leopards avoid areas where tigers are com- which competition is expected to be relatively mon (Seidensticker 1976; Seidensticker et al. intense. Their shared food resources, prey, are 1990), and do not cache their kills in trees often difficult and dangerous to kill, and con- where tigers and dholes are absent (Mucken- sequently carcass theft (kleptoparasitism) is a hirn and Eisenberg 1973). worthwhile alternative to hunting. Carcass Previous work on morphological diver- theft involves confrontation between individ- gence among sympatric large carnivores sub- uals, which can be dangerous given that the stantiates the importance of interspecific com- participants are armed with teeth and jaws, petition over evolutionary timescales. In a se- and in some cases, claws, designed to kill. ries of papers, Van Valkenburgh (1985, 1988, Moreover, given the species’ predatory abili- 1991, 1994) explored morphological separa- ties, interspecific competition can be manifest tion among large carnivores in past and pre- as intraguild predation (Polis and Holt 1992). sent communities. Comparisons of dental and If competition among carnivores is as intense inferred dietary differences among sympatric as predicted, then adaptations to minimize predators from extinct and living communi- dangerous encounters and escape predation ties revealed repeated patterns of divergence are expected in less dominant members of the within guilds into meat specialists (e.g., felids, guild. extinct nimravids), bone crackers (e.g., hyaen- Two recent reviews of the published litera- ids, borophagine canids), and omnivores (e.g., DINOSAURIAN AND MAMMALIAN PREDATORS 529 coyotes, early amphicyonids). Within dietary theropods likely preyed on smaller ones and types, coexisting species were likely to differ may have even eaten juveniles of their own in body size or locomotor adaptations. Larger species. Consequently, we might expect to predators are capable of taking larger prey, find predatory adaptations, as well as adap- thus effecting some dietary separation. Differ- tations to interspecific competition and intra- ences in locomotory abilities (such as climb- guild predation, among sympatric carnivo- ing, endurance running) can also reduce over- rous theropods that are similar to those of lap in prey choice through differences in hab- sympatric mammalian carnivores. itat preference. Perhaps more importantly, they can allow one predator to escape from Materials and Methods another, such as when leopards climb to avoid The Sample more-terrestrial lions, hyenas, and wild dogs (Van Valkenburgh 1985). Our guilds of predatory dinosaurs were de- Further evidence of the potential role of in- fined to include all known theropods from a terspecific competition in shaping the mor- well-sampled geologic formation that could phology of sympatric carnivores comes from reasonably be assumed to have been sympat- the work of Dayan and colleagues (Dayan et ric in time and space. We excluded from the al. 1989a,b, 1990, 1992; Dayan and Simberloff guild edentulous theropods (ornithomimo- 1994). In studies of several suites of sympatric saurs, oviraptors), small putatively edentu- species, including mustelids, canids, and fe- lous theropods (elmisaurids), and those with lids, they found evidence of remarkably even non-sectorial teeth (therizinosaurids) on the size separation between species in various fea- assumption that these taxa were not feeding tures, such as canine tooth length (felids) or regularly on large prey. Although we are un- lower molar length (canids, mustelids). They certain of their diets, the differences in their argue that such even size separation does not morphology from that of the presumably hy- occur by chance and is better explained as di- percarnivorous theropods do not suggest that vergence due to competition for food. Simi- they were exploiting ‘‘the same class of envi- larly, Kiltie (1988) found that differences in ronmental resources in a similar way’’ (Root jaw length among sympatric neotropical cats 1967). We will, however, mention their occur- were fairly constant and suggested that this rences. We also did not include crocodilian or reflected differences in maximum jaw gape avian species, but we do not consider them to and thus prey size. have been major competitors of the large ter- This discussion of ecological separation and restrial theropods that form the focus of this morphological divergence among sympatric analysis. Furthermore, although the teeth of mammalian carnivores was intended to en- many theropods, carnivorous ‘‘thecodonts,’’ gage the reader in considering how theropods ziphodont crocodilians, and varanoids, and might have coexisted. Among living carnivor- even the canines of some sabercats are basi- ans, almost all species larger than about 21 kg cally similar in form (Farlow et al. 1991), the take prey much larger than themselves, teeth of other crocodilians—such as those whereas smaller species feed mostly on prey found in the faunas we examine—seem to be that is 45% or less of their body weight (Car- different. Specifically, in this instance the croc- bone et al. 1999). Using an energetic model, odile teeth are unserrated and rounded in sec- Carbone et al. (1999) argued that it becomes tion, not compressed as are most theropod increasingly difficult to subsist on small prey teeth (Molnar personal observation). Vara- items as predator body mass increases be- noids are not included, both because of the cause of limitations on intake rate and forag- scarcity of their fossils and because the ple- ing time. Thus because of their large body siomorphic forms appear not to have been size, most theropods probably hunted prey as predators on relatively large prey (Losos and large or larger than themselves. Given this, we Greene 1988). expect that carcass theft was worthwhile and Because we wished to compare species rich- battles over kills occurred. Moreover, larger ness and ecological diversity in dinosaurian 530 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

and mammalian guilds, we chose paleofaunas TABLE 1. Theropod species included in this paper list- that included both small and large theropods, ed by paleofauna. Body mass estimates are not shown for edentulous theropods (indicated by *) because they suggesting that many of the theropods that are presumed not to have been predators of large prey coexisted might be represented. This limited and thus were excluded from the allometric and guild our choices to three well-sampled paleofaun- analyses. The Hell Creek species were excluded from the guild analyses because theropod species richness in this as, the Morrison (Late Jurassic, western United paleofauna appears depauperate. Faunal composition States and Canada), the Judith River (Late data are from Weishampel (1990). Sources for body mass Cretaceous, Canada), and the Nemegt (Late estimates: (1) Paul 1988 (estimated by immersion of scale models, details in Paul 1988:234–235); (2) Colbert Cretaceous, Mongolia) (Table 1). Given that it 1962 (estimated by immersion of scale models); (3) Chure is unlikely that all theropods that existed have 1995 (using the formula published in Anderson et al. been preserved or discovered for each of these 1985); (4) original estimate (based on comparison with similar-sized, better-known taxa); (5) original calcula- faunas, we consider the guild lists to represent tion (after Anderson et al. 1985). minimum numbers of coexisting species. It is possible that because of time and space aver- Estimat- ed body aging, the guilds include species that would mass have existed in different habitats or at differ- Species (kg) Source ent times, thus inflating actual guild size. Morrison (Late Jurassic) However, the species are of large body size Allosaurus fragilis 2090 1, 2 and presumably large home range (as is typ- Marshosaurus bicentesimus 225 1 ical of extant large carnivorous mammals [see Saurophaganax maximus 2720 3 Torvosaurus tanneri 1950 1 Gittleman and Harvey 1982; Farlow 1993]), Ceratosaurus nasicornis 980 1 and so we assume that there was spatial over- Elaphrosaurus sp. 210 1 lap among them. The question of temporal Stokesosaurus clevelandi 80 1 Coelurus fragilis 20 1 overlap is discussed below in each of the pa- Ornitholestes hermanni 13 1 leoguild descriptions. Koparion douglassi — — The Morrison Guild. The Morrison fauna Judith River (Late Cretaceous) derives from the Morrison Formation of the Ricardoestesia gilmorei — — western United States, from Montana to New Dromaeosaurus albertensis 15 1 Saurornitholestes langstoni 5 1 Mexico and from Oklahoma to Utah, which Troodon formosus 50 1 was deposited in Late Jurassic (Kimmeridgian Albertosaurus libratus 2500 1 and Tithonian) times (Turner and Peterson Aublysodon sp. 200 1 1999). The theropod fauna includes three large Daspletosaurus torosus 2300 1 *Elmisaurus elegans (estimated weight greater than 1000 kg), one *Struthiomimus altus moderately large (estimated 500–1000 kg), *Dromiceiomimus samueli two moderately small (100–500 kg), and four *Ornithomimus edmontonicus *Chirostenotes pergracilis small (less than 100 kg) taxa. There is no rea- *Caenagnathus collinsi son to infer that any one of these was eden- *cf. Erlikosaurus sp. tulous. The large taxa include Allosaurus fra- Nemegt (Late Cretaceous) gilis, Torvosaurus tanneri,andSaurophaganx Bagaraatan ostromi 25 5 maximus: the first two are relatively well Deinocheirus mirificus 9000 1 Therizinosaurus cheloniformis — — known, and estimated at approximately the Borogovia gracilicrus 30 4 same weight. The larger S. maximus is less well Saurornithoides junior 27 1 known, and cranial remains have not yet been Tochisaurus nemegtensis 55 4 found. It appears to have been less abundant Maleevosaurus novojilovi 350 5 Tyrannosaurus bataar 5000 1, 2 than the others. The moderately large thero- *Gallimimus bullatus pod is Ceratosaurus, known from several spec- *Elmisaurus rarus imens assigned to three species (C. nasicornis, *Oviraptor mongoliensis C. magnicornis, C. dentisulcatus) (Gilmore 1920; Hell Creek (Late Cretaceous) Madsen and Welles 2000). Fossils of Ceratosau- cf. Albertosaurus sarcophagus 2400 1 Nanotyrannus lancencis 1100 1 rus are clearly less abundant than those of Al- Tyrannosaurus rex 5000 1 losaurus, suggesting that this may have been DINOSAURIAN AND MAMMALIAN PREDATORS 531 true of the living animals, although it could we recognize seven carnivorous theropods reflect taphonomic bias. Allosaurus is the most that probably took large prey and seven eden- common of the theropods, making up more tulous theropods that were unlikely to have than 60% of all theropod specimens (Foster regularly hunted large prey (Table 1). Of the and Chure 1998). The smaller forms are also carnivorous forms, Albertosaurus libratus and rare as fossils. Of the moderately small forms, Daspletosaurus torosus are represented by sub- Elaphrosaurus sp. is represented by only very stantial or complete skeletons. Dromaeosaurus rare elements and Marshosaurus bicentesimus is albertensis, Saurornitholestes langstoni, Troodon also uncommon, although not as rare as Ela- formosus,andRicardoestesia gilmorei are known phrosaurus. Likewise, the small forms are un- from more limited material. Aublysodon sp. is common. Koparion douglassi,believedtobea represented in the Judith River by isolated primitive troodontid (Chure 1994), is known teeth, although known from fragmentary only from teeth, whereas Ornitholestes herman- skeletal remains from other regions. Among ni is represented by a single skeleton (and an the edentulous forms, Struthiomimus altus, additional manus), Coelurus fragilis by a single Dromiceiomimus samueli,andOrnithomimus ed- partial skeleton (and rare additional pieces), montonicus are represented by substantial or and Stokesaurus clevelandi by rare isolated piec- complete skeletons. Caenagnathus collinsi, El- es. To some extent this rarity is probably due misaurus elegans, Chirostenotes pergracilis (pos- to taphonomic bias, the more fragile bones of sibly synonymous with Caenagnathus [Currie the smaller theropods being less likely to be and Russell 1988]), and cf. Erlikosaurus sp. are preserved, but collecting bias and bias toward known only from incomplete specimens. The studying large specimens may also have been 14 species divided into three distinct groups involved in the late nineteenth and early twen- by estimated weight. The large forms weighed tieth centuries. over 2000 kg, and include Albertosaurus and Turner and Peterson (1999) divided the Daspletosaurus. The intermediate forms (Stru- Morrison into four zones deposited over about thiomimus, Dromicieomimus,andOrnithomi- 8 million years, from the Kimmeridgian into mus) weighed 100–200 kg, and the small forms the basal Tithonian. Koparion is found only in weighed less than 100 kg. These included zone 4 (the youngest) and all of the other spe- Caenagnathus, Chirostenotes, Elmisaurus, Troo- cies are found in zones 2 and 3 (zone 1 yields don, Dromaeosaurus, Saurornitholestes,andcf. only Allosaurus sp.). Allosaurus fragilis, Cerato- Erlicosaurus. saurus nasicornis,andCoelurus fragilis have Information on the stratigraphic occurrenc- been found in both zones; Marshosaurus bicen- es of these taxa is less easily available than for tesimus and Stokesosaurus clevelandi are from the Morrison, but Dodson (1971) provided in- zone 3; and Torvosaurus tanneri, Ornitholestes formation for the outcrop of the Dinosaur Pro- hermanni, Elaphrosaurus sp., and Marshosaurus vincial Park in Alberta. Dromicieomimus was sp. are found in zone 2. There are three radio- found higher in the section than Struthiomi- metric dates for the Morrison, 154 Ma for the mus, but as only a single specimen of each was base of zone 1, 150 Ma for the base of zone 3, recorded, this may not be significant. Caenag- and 149 Ma for the base of zone 4 (Turner and nathus occurred throughout the section, as did Peterson 1999). Our focus is on zones 2, 3, and Albertosaurus libratus, but the single specimen 4, and thus the time span represented is esti- of Daspletosaurus torosus was found near the mated to be 2–4 Myr. bottom of the beds. Dodson suggested, be- The Judith River Guild. The Campanian Ju- cause of their rarity, that the small theropods dith River Formation of the Montana-Alberta ‘‘habitually inhabited other environments in region has yielded a number of theropod taxa, other areas’’ (p. 68), unlike Albertosaurus and some rather better known than others (see the ornithomimosaurs. Dodson 1983). This unit dates to about 75 mil- The Nemegt Guild. Similar information for lion years ago, and seems to have been depos- the Nemegt Formation of Mongolia appears ited over a rather shorter period of time than not to be readily available. The age is not ac- the Morrison (Eberth et al. 1992). In the fauna, curately known but is generally taken to be 532 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

Campanian or Maastrichtian. We recognize taxonomic composition and environmental eight carnivorous forms and three noncarni- characteristics, are given by Van Valkenburgh vorous theropods from this paleofauna. The (1985, 1988, 1989, 1991) and Van Valkenburgh distribution by estimated weight is similar to and Hertel (1998). The guilds were defined to that for the Judith River. Large forms (over include all species larger than 7 kg that were 2000 kg) include Tyrannosaurus bataar and predatory and potentially competed for food. probably Deinocheirus mirificus and Therizino- saurus cheloniformis, both known only from Data Analysis quite incomplete material. Both Deinocheirus Comparisons of predatory adaptations in and Therizinosaurus are presumed carnivorous theropods and mammalian carnivores were in this paper, but we recognize that this di- made using least-squares regression of log10 agnosis could change with the discovery of transformed measurements, including skull craniodental remains. The two intermediate length, snout width, tooth length, jaw depth, species are estimated to weigh between 200 and jaw length. The data for the sampled and 500 kg and include Maleevosaurus novoji- mammals and theropods are derived from lovi, a carnivore, and Gallimimus bullatus,an previous publications as noted in the text and edentulous putative omnivore. The small relevant figure captions. forms probably weighed 100 kg or less and in- Species richness was estimated as the total clude an edentulous, presumably non- or number of species within each guild. Body weakly carnivorous species (Oviraptor mongo- size distributions within guilds were com- liensis), and highly carnivorous taxa (Sauror- pared by using Barton-David (B-D) statistics nithoides junior, Borogovia gracilicrus, Tochisau- (Barton and David 1956; Simberloff and rus nemegtensis,andBagaraatan ostromi). An Boecklen 1981). B-D statistics test whether size additional small species, Elmisaurus rarus, differences between successively sized species whose skull and teeth are unknown, is here are more similar than expected by chance. If assumed to have been edentulous based on its they are, and if they approach some constant, possible sister-taxon relationship to the Ovi- the result is usually assumed to reflect char- raptosauria (Currie 1990). Only T. bataar, M. acter displacement (see Dayan et al. 1989a,b). novojilovi,andG. bullatus are known from rea- To test for size ratio constancy, body mass es- sonably complete or complete skeletons; the timates for each species within a guild were others are represented by less or much less ordered from smallest to largest and log10 complete rare material. Data from the Polish- transformed. The differences in log10 trans- Mongolian expeditions (Gradzinski et al. formed mass values between successively 1968; Gradzinski and Jerzykiewicz 1972) in- sized species within a guild are the size ratio dicate that Tyrannosaurus bataar and ornithom- data. The B-D statistic used here, G, is the ratio imid (presumably Gallimimus bullatus) re- of the smallest to the largest size ratio within mains have a broad stratigraphic range each guild. Probability estimates for G-values through the beds. are estimated as described in Barton and Da- Mammalian Guilds. The three theropod vid (1956). guilds are compared with seven previously studied mammalian predator guilds. Four of Results and Discussion these are extant: Serengeti (East Africa), Ma- Comparative Morphology laysia, Chitawan (Nepal), and Yellowstone (North America). The remaining three are ex- We begin our comparison of dinosaurian tinct North American guilds: (1) Orellan, 34– and mammalian predatory guilds with an ex- 32 Ma; (2) Irvingtonian, 1.7–0.7 Ma; and (3) amination of the array of predatory types ap- Rancholabrean, 0.7–0.01 Ma. The fossil guilds parent in each group. As noted above, guilds were included because it is clear that extant of mammalian carnivorans (members of the predator guilds are depauperate as a result of order Carnivora) typically contain highly car- the late Pleistocene extinction event. More de- nivorous species (e.g., felids, some canids), tailed descriptions of these guilds, including more omnivorous species (e.g., most canids, DINOSAURIAN AND MAMMALIAN PREDATORS 533 some ursids), and perhaps bone crackers (e.g., specialized in different foods, but they cer- hyaenids). Can we find these same ecomorphs tainly do not appear to have been highly car- among dinosaurs in general and theropods in nivorous. particular? Apparently not, in that the bone- In theropods, the absence of strongly het- crackers and probably the more omnivorous erodont dentitions such as are typical of om- forms seem to be missing from dinosaur com- nivorous mammals might reflect the limited munities. This claim is based on the relative resources that were available. Omnivorous lack of dental diversity among theropods. carnivorans rely fairly heavily on fruits for With the exception of theriznosaurids and part or most of the year (Ewer 1973; Van Val- edentulous species, all theropods considered kenburgh 1989), and fruits of significant size here with known cranial material are relative- (Ͼ5 cm diameter) that likely depended on bi- ly similar in tooth form. Although the teeth otic dispersal were not common until the lat- may vary in size along the theropod tooth est Mesozoic (Tiffney 1984; Wing and Tiffney row, they all are shaped somewhat like the ca- 1987). Among living carnivorans, the ability nine teeth of carnivorans and would have to crack bones allows access to marrow, in- been of limited use for grinding plant matter cluding fat deposits (see Haynes 1982). It is or cracking large bones regularly (Farlow et not clear that dinosaur bones, which are often al. 1991). It is clear that theropod teeth occa- pneumatic in saurischians, contained a highly sionally contacted bone in killing and/or nutritious marrow, and if not, the major ben- feeding as evidenced by various punctures efit of bone cracking would have been non- and scores on the preserved bones of their existent. Thus, the limited carnivorous nature prey and a large coprolite containing many of theropod dental morphology is appropriate bone fragments, but there is no evidence of for the predominant food (dinosaurs) that was significant gnawing (Currie and Jacobsen available to nonherbivorous species. Alterna- 1995; Erickson and Olson 1996; Chin et al. tively, the lack of differentiation in theropod 1998). Moreover, a survey of tooth-damaged dentitions might reflect constraints on tooth bone in six dinosaur localities found very little form imposed by the nature of their develop- evidence of bone crushing, supporting the ment. Theropod teeth are replaced multiple idea that habitual consumption of large bones times over their life span and thus the tooth was rare (Fiorillo 1991). These findings sug- row always contains teeth of varying age and gest that all theropods with sectorial teeth height. Consequently precise tooth-to-tooth were highly carnivorous (hypercarnivorous) occlusion between any pair of opposing upper and that ecological separation among coexist- and lower teeth is difficult to sustain, and thus ing taxa would have depended on differences the evolution of more complex teeth may not in habitat choice, prey size, or prey type. The have been favored. diet of the edentulous forms is unclear, but a Comparisons of the postcranial adaptations recent discovery of gastroliths in an unde- for hunting and killing in dinosaurs and car- scribed Chinese ornithomimosaur may indi- nivorans reveal few similarities. Among ex- cate a herbivorous/granivorous diet for at tant large carnivorans, there are short-dis- least this species (Kobayashi et al. 1999). Pre- tance ambush species (e.g., felids, some ur- vious work with stable nitrogen isotope anal- sids) that use their muscular forelimbs and ysis did not find a significant difference in val- clawed feet to grapple with prey while they ues between North American ornithomimo- administer a deadly bite. Alternatively, there saurs and clearly carnivorous theropods such are those that are incapable of grappling, and as tyrannosaurids and dromaeosaurids (Os- instead kill with jaws alone after a long-dis- trom et al. 1993). However, only two values tance pursuit (e.g., canids, hyaenids). Two were obtained, one similar to that for ceratop- comparable types can be recognized among sians and the other greater than that of the theropods, ‘‘head-hunters’’ and ‘‘grappler/ dromaeosaurid examined. The ornithomimo- slashers.’’ Presumably, the head-hunters were saurs may have had a catholic diet, like living the allosaurs, later ceratosaurs, and tyranno- foxes or bears, or different species may have saurs, all relatively large species with reduced 534 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR forelimbs and no evidence of enlarged pedal claws. These species killed through actions of their jaws and teeth with little or no assistance from their forelimbs. The best developed grappler/slashers were the dromaeosaurs, ag- ile theropods of small to medium size with well-developed, clawed forelimbs and slash- ing pedal claws. No doubt, these species used both hands and feet to produce mortal wounds, as felids will in some circumstances. Coelurosaurs and troodontids lacked the large, terrible claws of dromaeosaurs and can be considered less well equipped grapplers FIGURE 1. Log10-Log10 plot of maximum skull length (mm) against body mass (kg) for canid (solid squares, n that almost always took prey smaller than ϭ 17), felid (open circles, n ϭ 16), and theropod (trian- themselves. gles, dashed line, n ϭ 14) species. Linear regression The two different killing styles in theropods equations: for canids, log10 y ϭ 1.85 ϩ 0.332 (log10 body 2 mass), r ϭ 0.89, p Ͻ 0.001; for felids log10 y ϭ 1.8 ϩ 0.273 are not so clearly associated with different 2 (log10 body mass), r ϭ 0.9, p Ͻ 0.001; for theropods, log10 2 hunting modes (ambush vs. long-distance y ϭ 1.92 ϩ 0.314 (log10 body mass), r ϭ 0.93, p Ͻ 0.001. pursuit) as they are in carnivorans. On the one Data for canids are from Van Valkenburgh and Koepfli (1993); data for felids are from Van Valkenburgh and hand, the light build of most of the grappler/ Ruff (1987). Data for theropods are from Osborn (1916), slashers suggests speed, and therefore the po- Gilmore (1920, 1946), Colbert (1962), Russell (1970), tential for long-distance pursuit. Their mod- Barsbold (1974), Maleev (1974), Paul (1988), and Britt erate size, however, would also have made (1991). them excellent ambush predators, as they could be concealed fairly easily in dense veg- ropods, our sample is confined to 28 species etation. On the other hand, many or all of the from four formations, the Morrison (Late Ju- head-hunters were so large as to make am- rassic), Judith River (Late Cretaceous), Hell bush hunting seem ludicrous. What could a T. Creek (Late Cretaceous), and Nemegt (Late rex hide behind? Although dense vegetation Cretaceous) (Table 1). The Hell Creek assem- might conceal a T. rex, could such a large an- blage was not included in our paleoguild com- imal move through it quietly enough to stalk parisons because it has only three species as its prey? If ambush was not an option, then the opposed to the 10–14 for the included assem- tyrannosaurs must have overtaken their prey, blages, and this suggests that it is likely to be although this was probably done with a fast missing species that were present. In fact, one walk rather than a run (Farlow et al. 1995; Car- of the three species, Nanotyrannus lancensis, rano 1999; Hutchinson and Garcia 2002). may be a juvenile T. rex (Carr 1999), leaving Although qualitative comparisons of the- only two theropods in the Hell Creek. We in- ropod and carnivoran adaptations can be clude the Hell Creek species in the allometric stimulating, it is more satisfying to establish analysis to expand the sample of theropods similarities in structure based on quantitative for comparison with mammals. An even larg- data. Fortunately, recent studies of the allom- er sample of theropod species might strength- etry of teeth, jaws, and skulls of mammalian en our conclusions and we would like to ex- carnivores allow for some direct comparison pand the analysis in the future. Nevertheless, of scaling relationships between mammalian we feel that the 28 sampled species are broad- and dinosaurian predators. The data present- ly representative of the diversity of theropod ed here are limited, in that only a subset of craniodental morphology. theropod diversity is represented and com- Relative Head Size. Relative head size varies parisons were constrained by the availability among extant large carnivorans (Fig. 1). Felids of similar data (e.g., measures of jaw depth or have small heads relative to their body mass, length) for both groups. Because we were in- both because their bodies tend to be fairly terested in examining sympatric suites of the- heavily muscled and because their skulls are DINOSAURIAN AND MAMMALIAN PREDATORS 535

theropods. Maximum muzzle width is positively allometric in all three groups, with larger species exhibiting relatively broader muz- zles. The smallest theropods for which we have data overlap canids on the plot of maximum snout width against skull length. The relationship of snout width to skull length in theropods is close to isometry (1.19), but our sample size is quite small (n ϭ 10), and the 95% confidence interval for the slope is large (0.62–1.76). In canids, the broader muzzle of larger species was correlated with enlarged FIGURE 2. Log10-Log10 plot of maximum snout width (mm) against skull length (mm) for canids (solid canine and incisor teeth and with diets that in- squares, n ϭ 33), felids (open circles, n ϭ 18), and the- clude relatively large prey (Van Valkenburgh ropods (triangles, dashed line, n ϭ 10). Linear regres- and Koepfli 1993). For example, wolves, sion equations: for canids, log10 y ϭϪ1.14 ϩ 1.26 (log10 2 skull length), r ϭ 0.92, p Ͻ 0.001; for felids log10 y ϭ dholes, and African hunting dogs often take 2 Ϫ0.89 ϩ 1.16, r ϭ 0.98; for theropods, log10 y ϭϪ1.2 ϩ 2 prey larger than themselves, whereas similar- 1.19 (log10 skull length), r ϭ 0.74, p Ͻ 0.01. Data for canids are from Van Valkenburgh and Koepfli (1993); fe- sized canids such as the maned wolf or Ethi- lids were measured for this paper. Data for theropods as opian wolf are more narrow-snouted and tend in Figure 1. to hunt small prey (e.g., lagomorphs and ro- dents). Theropods may have been similar, somewhat small. Canids on the other hand with the largest forms such as the tyranno- have relatively larger skulls and more slender saurids tackling the most difficult prey. No- bodies. The difference is most pronounced tably, the large theropod Torvosaurus tanneri among species of large size and reflects killing has a very narrow snout for its skull length behavior to some extent. Felids tend to kill (TTA in Fig. 2), suggesting it may have had a with a single, strong bite to the neck or muz- weaker bite and might have favored smaller zle, whereas canids tend to use repeated, shal- prey than similar-sized tyrannosaurs. lower bites to subdue their prey. The short Relative Tooth Size. Because of the absence snout of felids enhances the leverage of their of specific serial homology between the teeth jaw muscles for a killing bite with the canines of theropods and mammals, it is difficult to by reducing the moment arm of resistance of compare them in terms of relative tooth size. an anteriorly placed load (Biknevicius and It is not clear whether one should compare Van Valkenburgh 1996). Theropods were pro- theropod cheek teeth to the premolars or car- portioned more like canids (Fig. 1). With the nassials of carnivorans, for example. We chose exception of the small coelurosaur Ornitholes- the carnassials because they are usually the tes hermanni, all sampled theropods have rel- largest teeth in the jaw and function some- atively larger skulls than felids of similar what similarly in felids and canids. A com- body mass. Thus, we do not see a clear divi- parison of the largest lower teeth of theropods sion of theropods into the two alternative kill- with the lower carnassials of canids and felids ing types that are apparent today. Rather, they reveals that the teeth of theropods are sub- all appear to have been built fairly similarly stantially shorter mesiodistally than those of and there is little or no evidence of catlike the- canids and felids of similar jaw length (Fig. 3). ropods in our sample. It would be interesting The large carnosaurs and tyrannosaurs also to be able to include more taxa, especially appear to have shorter teeth than would be ex- smaller species. pected for a mammalian carnivore of their Skull Shape. A comparison of skull shape body size. This is not surprising given that among canids, felids, and theropods reveals theropod jaws contain many more teeth, all of additional similarities between canids and more similar size, than do jaws of mammals. theropods (Fig. 2). Felids have broader snouts Although felids and canids appear to be sim- relative to skull length than either canids or ilar in tooth length/jaw length proportions, 536 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

FIGURE 3. Log10-Log10 plot of the mesiodistal length of the longest lower tooth (mm) against maximum jaw length (mm) for canid (solid squares, n ϭ 32), felid (open FIGURE 4. Log10-Log10 plot of jaw depth (mm) against circles, n ϭ 16), and theropod (triangles, n ϭ 10) species. maximum jaw length (mm) for felids (open circles, n ϭ For canids and felids, the lower first molar was used. 10), canids (solid squares, n ϭ 10), hyaenids (diamonds, Linear regression equations: for canids and felids to- n ϭ 2), and theropod (triangles, n ϭ 10) species. Jaw gether, log y ϭϪ0.93 ϩ 1.05 (log jaw length), r2 ϭ depth is measured at the midpoint of the tooth row in 10 10 theropods, and as the maximum depth along the tooth 0.85, p Ͻ 0.001; for theropods, log10 y ϭ 1.94 ϩ 1.25 (log10 jaw length), r2 ϭ 0.77, p Ͻ 0.01. For data sources, see Fig- row in carnivorans. Linear regression equations: for fe- 2 ure 1. lids, log10 y ϭϪ0.8 ϩ 1.04 (log10 jaw length), r ϭ 0.99, p Ͻ 0.001; for canids, log10 y ϭϪ1.57 ϩ 1.37 (log10 jaw 2 length), r ϭ 0.95, p Ͻ 0.001; for theropods, log10 y ϭ 2 Ϫ1.33 ϩ 1.22 (log10 jaw length), r ϭ 0.97, p Ͻ 0.001. Data the slopes of the log-log regressions differ sig- for carnivorans are from Biknevicius and Ruff (1992). nificantly, with canids showing positive al- Data for theropods are from Osborn (1916), Gilmore (1920, 1924), Russell (1969, 1970), Colbert and Russell lometry as opposed to near isometry in felids. (1969), Barsbold (1974), Maleev (1974), Britt (1991), Car- Interestingly, theropods also show positive al- penter (1992), Osmolska (1996), and Glut (1997). lometry with larger species displaying even larger teeth (slope ϭ 1.25). However, our sample size for theropods is small and the addi- maximum depth in theropods occurs behind tion of more species could easily alter the ap- the tooth row in the region where jaw adduc- parent positive relationship. tor muscles insert. Because we are interested Relative Jaw Depth. The ratio of jaw depth in the strength of the jaw when biting, we to length should be indicative of jaw strength chose the tooth-row midpoint as a reasonable during biting. Although it would be prefera- position for the estimation of jaw rigidity. ble also to include information on jaw width Despite the potential problems of compar- and cortical bone distribution internally as ing non-analogous measures, a log-log regres- has been done for carnivorans (Biknevicius sion of jaw depth on length for carnivorans and Ruff 1992), few data are available for the- and theropods reveals similar scaling relation- ropods. When tetrapods bite an object, their ships, with both groups showing positive al- jaws are loaded somewhat as beams with the lometry (slope ϭ 1.12, theropods; 1.04 felids; jaw joint acting as a fulcrum. Bending 1.37, canids; Fig. 4). The slopes of the regres- strength is enhanced by deepening the jaw rel- sion lines are not significantly different be- ative to its length, and consequently, carnivor- tween theropods and felids (ANCOVA: p ϭ ans such as the bone-cracking hyenas have rel- 0.357), theropods and canids(ANCOVA: p ϭ atively deep jaws. Here we compare jaw depth 0.137), or theropods against all 22 carnivoran taken midway along the tooth row of thero- species (ANCOVA: p ϭ 0.894). Apparently, pods with previously published data on max- however, the jaws of theropods were relatively imum depth in carnivorans (canids, felids, weaker for their length than those of carni- hyaenids). Maximum jaw depth in carnivor- vorans. The positive allometry in both groups ans tends to occur posterior to the midpoint suggests that larger species loaded their jaws of the tooth row, near the carnassials. Al- more heavily, because deeper jaws are much though it might have been preferable to com- stronger in resisting loads applied dorsoven- pare maximum jaw depth in both groups, trally, as during jaw closure. This component DINOSAURIAN AND MAMMALIAN PREDATORS 537 of jaw strength is largely dependent on jaw depth and increases approximately with the square of depth (see Biknevicius and Van Val- kenburgh 1996 for details). As was the case for muzzle width, the positive allometry of jaw depth suggests that larger theropods killed prey that were very large relative to their own body size. Thus, the data presented here suggest that the jaws of Tyrannosaurus rex were at least as strong as and probably relatively stronger than those of other carnosaurs, and do not support the idea that T. rex was a less capable killer and relied more on scavenging FIGURE 5. Log -Log plot of typical prey size (kg) than other carnosaurs (see also Erickson et al. 10 10 against maximum jaw length (mm) for ten species each 1996; Molnar 2000). of canids and felids (solid circles). The linear regression The fact that theropods appear to have had equation derived from these data was used to estimate weaker jaws than similarly sized carnivorans typical prey sizes of theropods (triangles) based on their estimated body masses. Linear regression equation for might be explained by our choice of tooth row carnivorans (n ϭ 10): log10 y ϭϪ7.25 ϩ 6.22 (log10 jaw midpoint for the depth comparison. As noted length), r2 ϭ 0.84, p Ͻ 0.001. Data for typical prey size above, the measure for carnivorans was taken of carnivorans are from Van Valkenburgh and Hertel (1998). posterior to the midpoint, and it is clear that theropod jaws do deepen behind the midpoint but beyond the end of the tooth row. To re- (one thousand tons), a size that exceeds the solve this issue, it would be useful to compare maximum weight estimates made for any di- jaw depth in carnivoran and theropod jaws at nosaur (Peczkis 1994). On the other hand, the multiple positions in a more comprehensive prey size estimates for the three smaller the- analysis. In addition, it should be noted that ropods, Ornitholestes, Saurornithoides,andDro- theropod dentary bones are solid (Molnar maeosaurus, might be reasonable, as they are 2000), whereas those of carnivorans have a much more similar in size to the mammalian medullary cavity. A solid dentary is some- carnivores used for the regression. It is diffi- what stronger than a hollow one, but because cult to assign much confidence to any of these strength in bending is enhanced more by in- estimates given the differences in body form creasing overall diameter, the carnivorans between dinosaurs and carnivorans. However, probably retain the stronger jaws. In the two it is likely that there was a positive relation- examples where theropods and carnivorans ship between prey size and jaw depth in the- overlap in size (Ornitholestes hermanni and ropods, and consequently, jaw depth might be Dromaeosaurus albertensis), carnivorans have useful in examining ecological separation jaws that are approximately twice as deep. No- among sympatric theropods. Unfortunately, tably, a recent finite element analysis of cra- we could not do so because of the absence of nial and bite strength in Allosaurus fragilis also jaw depth data for many of our species. found that this species had a relatively weaker Species Richness. The mammalian predator bite than do carnivorans (Rayfield et al. 2001). guilds include species whose diets range from Jaw depth is significantly correlated with omnivorous to hypercarnivorous and thus are typical prey size in a limited sample of ten not strictly comparable to the theropod species of canids and felids (Fig. 5). If this re- guilds, which include putative hypercarni- gression is used to predict the typical prey of vores only. If the edentulous and presumably theropods, it produces overestimates, at least less carnivorous theropods are considered as for those species that were much larger than part of the predator guild, then total species any of the carnivorans used in the regression. diversity is similar in both mammalian and For example, it predicts that Tyrannosaurus rex theropod guilds (Fig. 6). The total number of was usually killing prey that weighed 106 kg species within the mammalian guilds ranges 538 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

enced by energetics; if the energetic needs of dinosaurs were reduced relative to mammals, a given environment might sustain a greater abundance and diversity of species. Species richness values might be inflated for the theropod guilds relative to the mammalian guilds if time-averaging is more severe for the former. Each of the theropod guilds is derived from sediments that span at least 1 Myr (Judith River) to as much as 4 Myr (Morrison). FIGURE 6. Species richness in theropod (Nemegt, Judith River, Morrison), extinct mammalian predator (Orellan, It is impossible to be certain that all the spe- Irvingtonian, Rancholabrean), and extant mammalian cies listed as guild members were sympatric predator (Yellowstone, Chitawan, Malaysia, Serengeti) guilds. Number of hypercarnivorous species is shown in for some period of time, but given that verte- black. For species composition of all mammalian guilds brate species typically last for about 2–4 Myr except the Rancholabrean, see Van Valkenburgh 1985, (Stanley 1979), it seems likely that they would 1991. Rancholabrean guild includes Canis lupus, C. dirus, C. latrans, Smilodon fatalis, Homotherium serus, Panthera have overlapped in time. Consequently, we do atrox, P. o n c a , P.concolor, Lynx rufus, Miracinonyx trumani; not feel that time-averaging explains the rich- Arctodus simus, Ursus arctos,andU. americanus. All but ness of the theropod faunas. Moreover, the ex- the last three and C. latrans were considered hypercarnivorous. For species composition of theropod guilds, tinct mammalian guilds (Orellan, Irvingtoni- see Table 1. an, Rancholabrean) also span millions of years and broad geographic areas. Spatial mixing of theropod species that were never contempo- from 8 (Malaysia, Chitawan) to 15 (Irvington- raneous but rather lived in the area at different ian) with a mean of 11. The mean diversity of times also seems unlikely. As mentioned be- the theropod guilds is also 11, with 14 species fore, large carnivores tend to have large home in the Judith River, 11 in the Nemegt, and 10 ranges and broad geographic distributions as in the Morrison. If species richness of just the a result of their metabolic requirements (see hypercarnivores is compared, theropod and Gittleman and Harvey 1982; Farlow 1993) and mammalian guilds are also not significantly this makes it difficult for species to avoid each different, with an average of 6 hypercarni- other over extended time spans. Alternatively, vores in mammalian guilds and 8 in the the- theropod richness might be inflated because ropod guilds (Fig. 6). This is somewhat sur- of taxonomic oversplitting relative to mam- prising for two reasons. First, as mentioned mals. This is difficult to assess but would earlier, sympatric theropods appear more probably only decrease species richness by similar to one another than do sympatric hy- one or two taxa per guild. Most of the thero- percarnivorous mammals, and thus one pod species are readily distinguished from would expect fewer species could coexist. one another. Such a limit on co-occurrence is known to oc- The theropod guild with the greatest num- cur among some extant lizards, in that Va r a nu s ber of hypercarnivorous species (ten) is the komodoensis, a hypercarnivore, appears to late Jurassic Morrison. If this is not due to eliminate other, smaller varanids from its taphonomic bias or oversplitting, the in- range (Auffenberg 1981). However, this does creased predator richness might reflect a not appear to be true among the varanids of greater availability of prey, as is the case for Australia (Pianka 1994). Second, given that mammalian guilds. For example, of the four there is typically a negative relationship be- extant predator guilds, the Serengeti includes tween species diversity and body size (see a greater number of species in total, as well as Hutchinson and MacArthur 1959; Van Valen a greater number of hypercarnivores, and has 1973), the larger mean body size of carnivo- by far the greatest biomass of prey (Van Val- rous theropods would be expected to be as- kenburgh 1989). Unlike the two Late Creta- sociated with reduced taxonomic diversity. Of ceous faunas, the Morrison included several course, this relationship is likely to be influ- large sauropod species (Weishampel 1990) DINOSAURIAN AND MAMMALIAN PREDATORS 539

TABLE 2. Average size difference between adjacent- pair or trio of very similarly sized species. It sized species within guilds computed as the difference would be interesting to look for other evi- between log10 body mass estimates for each species. G1- Barton-David statistic, the ratio of the smallest differ- dence of character divergence within these ence between species over the largest difference be- clusters, such as differences in locomotor or tween species. p ϭ probability that the distribution of feeding adaptations, as is apparent among body size differences can be explained by chance. mammals. For example, the Serengeti includes Mean three species of similar mass—leopard, chee- Guild size ratio G1 p tah, and spotted hyena. Each of these is quite Malaysia 0.238 0.097 0.285 distinct morphologically and behaviorally. Serengeti 0.177 0.038 0.533 The hyena exploits carcasses more fully than Yellowstone 0.257 0.068 0.743 Chitawan 0.344 0.10 0.648 either cat because of its bone-cracking abili- Rancholabrean 0.202 0.151 0.063 ties. The two cats separate by habitat and Irvingtonian 0.149 0.179 0.154 hunting style; the cheetah is a sprinter of more Orellan 0.156 0.147 0.24 Nemegt 0.426 0.056 0.678 open country whereas the leopard is an am- Judith River 0.54 0.113 0.707 busher that prefers gallery forest. By contrast, Morrison 0.290 0.047 0.397 the trios and pairs of similar-sized theropods are not composed of combinations of head hunters and grappler/slashers. Instead, there and this could have favored a greater diversity is no overlap in size between these two; grap- of large theropods. pler/slashers always filled the low end of the Body Size Diversity. Predator body size is size range. positively associated with prey body size (Van The average difference between successive- Valkenburgh and Koepfli 1993; Van Valken- ly sized species among the seven mammalian burgh and Hertel 1998), and thus ecological predator guilds was 0.201 (n ϭ 38, SD ϭ 0.173) separation among sympatric hypercarnivores (Table 2). The same value was larger (0.399; n can be effected by differences in body size. As ϭ 19, SD ϭ 0.348) for the three theropod noted above, studies of modern carnivore guilds, and this difference was significant us- guilds have documented remarkably even ing a t-test (F ϭ 8.275; p Ͻ 0.01), but not using spacing between species in the size of certain a Mann-Whitney nonparametric test (p ϭ features, such as jaw length or tooth size (e.g., 0.11). The nonparametric test is more appro- Dayan et al. 1992; Kiltie 1988). We would have priate given that ratio data are being com- liked to do the same sorts of quantitative anal- pared, but it is much more conservative. Given yses of the theropod guilds but the data are that the sample size for theropods is much less not sufficiently complete. Unfortunately, size than that for the mammals, the lack of statis- ratio analyses are not useful when the same tical significance should not be considered as measurement is not available for all species definitive. It does appear from examination of within the guild. The only parameter for the distribution of body sizes within guilds which we had nearly complete data was esti- that sympatric theropods tend to separate by mated body mass, and it has not been ob- size more than mammals (Fig. 7). This might served to show constant size ratios in modern reflect an increased pressure to segregate by carnivore guilds. Because the body mass data size among theropods because of their overall for theropods used here are estimates taken similarity in body form. In addition, suites of from the literature that were derived by dif- sympatric theropods span a much greater ferent methods, our results must be viewed range of body sizes than do those of mam- with caution. mals; consequently, there is more room to For both carnivorans and theropods, size spread out. ratios were not constant within any guild, and In sum, theropod guilds do not seem to B-D statistics indicate that the distribution of demonstrate ecological separation as clearly size ratios could be explained by chance (Table as do the mammalian guilds. Size separation 2, Fig. 7). Unlike the mammal guilds, each of is apparent, but when there is overlap, it is not the three theropod guilds includes at least one clear how the species differ along another di- 540 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

FIGURE 7. Body size (log 10) distribution of hypercarnivorous species in theropod and mammalian predator guilds. Species abbreviations: FVI ϭ Felis vivverina; FTE ϭ F.temmincki;CALϭ Cuon alpinus; NNE ϭ Neofelis nebulosa;PPA ϭ Panthera pardus; PTI ϭ P.tigris; FSE ϭ F.serval;CCAϭ Caracal caracal;LPIϭ Lycaon pictus; CCR ϭ Crocuta crocuta; AJU ϭ Acinonyx jubatus;PLEϭ P. l e o ; LRU ϭ Lynx rufus;CLUϭ Canis lupus;CDIϭ C. dirus; PCO ϭ Puma concolor; MTR ϭ Miracinonyx trumani;PONϭ Panthera onca; HSE ϭ Homotherium serus;SFAϭ Smilodon fatalis;PATϭ Panthera atrox; FRE ϭ Felis(?) rexroadensis; BDI ϭ Borophagus diversidens;CARϭ Canis armbrusteri; COS ϭ Chasmoporthetes ossifragus; SGR ϭ Smilodon gracilis; HCR ϭ Hyaenodon crucians;HPRϭ Hoplophoneuus primaevus; DSP ϭ Dinictis sp.; HHO ϭ Hyaenodon horridus; ESI ϭ Eusmilus sicarius;HOCϭ Hoplophoneus occidentalis; BOS ϭ Bagaraatan ostromi; SJU ϭ Saurornithoides junior; BGR ϭ Borogovia gracilicrus; TNE ϭ Tochisaurus nemegtensis;MNOϭ Maleevosaurus novojilovi; TBA ϭ Tyrannosaurus bataar;DMIϭ Deinocheirus mirificus; SLA ϭ Saurornitholestes langstoni;DALϭ Dromaeosaurus albertensis; TFO ϭ Troodon formosus; ASP ϭ Aublysodon sp.; DTO ϭ Daspletosaurus torosus; ALI ϭ Albertosaurus libratus;OHEϭ Ornitholestes hermanni; CFR ϭ Coelurus fragilis; SCL ϭ Stokesosaurus clevelandi; ESP ϭ Elaphosaurus sp.; MBI ϭ Marshosaurus bicentesimus;CNAϭ Ceratosaurus nasicornis;TTAϭ Torvosaurus tanneri; AFR ϭ Allosaurus fragilis; SMA ϭ Saurophaganax maximus. DINOSAURIAN AND MAMMALIAN PREDATORS 541 mension. In large part this is due to a lack of than do coexisting carnivorans. These differ- comparable information; it is rare to find rea- ences could have favored coexistence because sonably complete cranial, dental, and limb larger species tend to take larger prey. The as- data for dinosaurs because the material is too sociation between predator and prey body fragmentary. Some species are known only size is well established for a wide array of ex- from teeth and jaws, others from limb ele- tant vertebrates (frogs, snakes, lizards, birds, ments. As a result, we could not examine the- mammals [Emerson et al. 1994]) and probably ropod guilds using a multidimensional mor- applied to dinosaurs as well. Notably, it is not phospace as has been done with carnivorans unusual to find two or three similar-sized spe- (cf. Van Valkenburgh 1985, 1988, 1991). Nev- cies of theropod in the same paleofauna that ertheless, the interspecific differences would do not differ markedly in their locomotor or seem to be subtler than those between sym- feeding anatomy. This contrasts with carni- patric carnivorans. As has been pointed out voran guilds where similar-sized sympatric above, dental diversity is limited among the- species usually do differ in either locomotor ropods, and major postcranial differences lie or feeding behavior. in the relative size of the forelimbs (i.e., grap- It seems very likely that theropod guilds pler/slasher vs. head hunter). were characterized by interspecific battles over carcasses, intraguild predation, and the Summary and Conclusions tendency of smaller species to avoid larger, Our admittedly preliminary look at the more dominant species, all processes that are structure of dinosaurian and mammalian observed in mammalian predator guilds. To- predator guilds revealed some intriguing re- day’s bullies of the Serengeti plains, lions and sults that deserve further investigation with spotted hyenas, were paralleled by tyranno- an expanded sample of theropods and more saurs in Jurassic and Cretaceous ecosystems. morphological data. The allometric compari- Smaller theropods probably adapted to life sons demonstrated a surprising similarity be- with tyrannosaurs by hunting in groups, feed- tween carnivorans and theropods in the scaling rapidly, and avoiding encounters with ing of jaw depth with length that suggests their larger cousins whenever possible. This similar biomechanical constraints. In addi- would have favored improved hunting skills tion, theropods and carnivorans shared the and associated adaptations such as agility, tendency to show positive allometry in the speed, intelligence, and increased sensory scaling of most craniodental measures, sug- awareness. With additional discoveries of gesting that in both groups, larger species kill more complete specimens of theropods, it will relatively larger prey. However, theropods be fascinating to compare more fully the sen- could not be shoehorned into doglike, hyena- sory, feeding, and locomotor capabilities of like, or catlike roles. Instead, they all appear sympatric theropods. Much remains to be dis- more similar in their skulls and teeth than covered. their mammalian counterparts. Other than the edentulous species, most if not all thero- Acknowledgments pods appear to have been hypercarnivorous, We gratefully acknowledge two thoughtful and this may reflect a Mesozoic dearth of al- anonymous reviewers as well as S. Wing, W. ternative resources (fruits, nutrient-rich bone DiMichele, J. Downs, and G. Erickson for their marrow) used by some mammalian predators. helpful and insightful comments. We also Theropod predator guilds included at least thank D. Simberloff for assistance with the as many species as comparable mammalian Barton-David statistics. predator guilds. This was not expected given the aforementioned morphological and in- Literature Cited ferred ecological similarity among sympatric Anderson, J. F., A. Hall-Martin, and D. A. Russell. 1985. Long- theropods, and the great size of most of the bone circumference and weight in mammals, birds and dinosaurs. Journal of Zoology 207:53–61. species. Coexisting theropods usually dif- Auffenberg, W. 1981. The behavioral ecology of the Komodo fered in body mass to a slightly greater degree monitor. University Presses of Florida, Gainesville. 542 BLAIRE VAN VALKENBURGH AND RALPH MOLNAR

Barsbold, R. 1974. Saurornithoididae, a new family of small the- wolves, jackals and foxes of Israel. Biological Journal of the ropod dinosaurs from central Asia and North America. Acta Linnean Society 45:315–331. Palaeontologia Polonica 30:5–22. Dodson, P. 1971. Sedimentology and taphonomy of the Oldman Barton, D. E., and F. N. David. 1956. Some notes on ordered ran- Formation (Campanian), Dinosaur Provincial Park, Alberta dom intervals. Journal of the Royal Statistical Society B 18:79– (Canada). Palaeogeography, Palaeoclimatology, Palaeoecolo- 94. gy 10:21–74. Biknevicius, A. R., and C. B. Ruff. 1992. The structure of the ———. 1983. A faunal review of the Judith River (Oldman) for- mandibular corpus and its relationship to feeding behaviours mation, Dinosaur Provincial Park, Alberta. Mosasaur 1:89– in extant carnivorans. Journal of Zoology 228:479–507. 118. Biknevicius, A., and B. Van Valkenburgh. 1996. Design for kill- Eaton, R. L. 1979. Interference competition among carnivores: a ing: craniodental adaptations of predators. Pp. 393–428 in J. model for the evolution of social behavior. Carnivore 2:9–16. L. Gittleman, ed. Carnivore behavior, ecology, and evolution, Eberth, D. A., R. G. Thomas, and A. Deino. 1992. Preliminary K- Vol. II. Cornell University Press, Ithaca, N.Y. Ar dates from bentonites in the Judith River and Bearpaw For- Britt, B. 1991. Theropods of Dry Mesa Quarry (Morrison For- mations (Upper Cretaceous) of Dinosaur Provincial park, mation, Late Jurassic, Colorado) with emphasis on the oste- southern Alberta, Canada. Pp. 296–304 in N. J. Mateer and ology of Torvosaurus tanneri. Brigham Young University Geo- P.-J. Chen, eds. Aspects of nonmarine Cretaceous geology. logical Studies 37:1–72. China Ocean Press, Beijing. Carbone, C., G. M. Mace, S. C. Roberts, and D. W. Macdonald. Emerson, S. B., H. W. Greene, and E. L. Charnov. 1994. Allo- 1999. Energetic constraints on the diet of terrestrial carni- metric aspects of predator-prey interactions. Pp. 123–139 in P. vores. Nature 402:286–288. C. Wainwright and S. M. Reilly, eds. Ecological morphology: Carpenter, K. 1992. Tyrannosaurids (Dinosauria) of Asia and integrative organismal biology. University of Chicago Press, North America. Pp. 250–268 in N. J. Mateer and P.-J. Chen, eds. Chicago. Aspects of nonmarine Cretaceous geology. China Ocean Erickson, G. M., and K. H. Olson. 1996. Bite marks attributable Press, Beijing. to Tyrannosaurus rex: preliminary description and implica- Carr, T. D. 1999. Craniofacial ontogeny in tyrannosauridae (Di- tions. Journal of Vertebrate Paleontology 16:175–178. nosauria, Coelurosauria). Journal of Vertebrate Paleontology Erickson, G. M., S. D. Van Kirk, J. Su, M. E. Levenston, W. E. Cal- 19:497–520. er, and D. R. Carter. 1996. Bite-force estimation for Tyranno- Carrano, M. T. 1999. What, if anything, is a cursor? Journal of saurus rex from tooth-marked bones. Nature 382:706–708. Zoology 247:29–42. Ewer, R. F. 1973. The carnivores. Weidenfeld and Nicolson, Lon- Chin, K., T. T. Totaryl, G. M. Erickson, and L. C. Calk. 1998. A don. king-sized theropod coprolite. Nature 393:680–682. Fanshawe, J. H., and C. D. Fitzgibbon. 1993. Factors influencing Chure, D. J. 1994. Koparion douglassi, a new dinosaur from the the hunting success of an African Wild Dog pack. Animal Be- Morrison Formation (Upper Jurassic) of Dinosaur National haviour 45:479–490. Monument; the oldest troodontid (Theropoda: Maniraptora). Farlow, J. O. 1993. On the rareness of big, fierce animals: spec- Brigham Young University Geology Studies 40:11–15. ulations about the body sizes, population densities, and geo- ———. 1995. A reassessment of the gigantic theropod Sauro- graphic ranges of predatory mammals and large carnivorous phagus maximus from the Morrison Formation (Upper Jurassic) dinosaurs. American Journal of Science 293A:167–199. of Oklahoma, USA. Pp. 103–106 in A. Sun and Y. Wang, eds. Farlow, J. O., D. L. Brinkman, W. L. Abler, and P. J. Currie. 1991. Short papers, Sixth symposium on Mesozoic terrestrial eco- Size, shape and serration density of theropod dinosaur lateral systems and biota, 1995. China Ocean Press, Beijing. teeth. Modern Geology 16:161–198. Colbert, E. H. 1962. The weights of dinosaurs. American Mu- Farlow, J. O., M. B. Smith, and J. R. Robinson. 1995. Body mass, seum Novitates 2076:1–16. bone ‘‘strength indicator’’ and cursorial potential of Tyran- Colbert, E. H., and D. A. Russell. 1969. The small Cretaceous nosaurus rex. Journal of Vertebrate Paleontology 15:713–725. dinosaur Dromaeosaurus. American Museum Novitates 1900: Fiorillo, A. R. 1991. Prey bone utilization by predatory dino- 1–20. saurs. Palaeogeography, Palaeoclimatology, Palaeoecology Currie, P. J. 1990. Elmisauridae. Pp. 245–248 in D. B. Weisham- 88:157–166. pel, P. Dodson, and H. Osmolska, eds. The Dinosauria. Uni- Foster, J. R., and D. J. Chure. 1998. Patterns of theropod diversity versity of California Press, Berkeley. and distribution in the Late Jurassic Morrison Formation, Currie, P. J., and A. R. Jacobsen. 1995. An azhdarchid pterosaur Western USA. Pp. 30–31 in Abstracts and Programs for the eaten by a velociraptorine theropod. Canadian Journal of Fifth International Symposium on the Jurassic System, Inter- Earth Sciences 32:922–925. national Union of Geological Sciences Subcommission on Ju- Currie, P. J., and D. A. Russell. 1988. Osteology and relation- rassic Stratigraphy. Vancouver, Canada. ships of Chirostenotes pergracilis (Saurischia, Theropoda) from Gilmore, C. W. 1920. Osteology of the carnivorous Dinosauria the Judith River (Oldman) Formation of Alberta, Canada. Ca- in the United States National Museum, with special reference nadian Journal of Earth Sciences 25:972–986. to the genera Antrodemus (Allosaurus)andCeratosaurus. Bul- Dayan, T., and D. Simberloff. 1994. Character displacement, sex- letin of the United States National Museum 110:1–154. ual dimorphism, and morphological variation among British ———. 1924. A new coelurid dinosaur from the Belly River Cre- and Irish mustelids. Ecology 75:1063–1073. taceous of Alberta. Bulletin of the Canadian Department of Dayan, T., E. Tchernov, Y. Yom-Tov, and D. Simberloff. 1989a. Mines, Geological Survey 38:1–12. Ecological character displacement in Saharo-Arabian Vulpes: ———. 1946. A new carnivorous dinosaur from the Lance For- outfoxing Bergmann’s rule. Oikos 55:263–272. mation of Montana. Smithsonian Miscellaneous Collections Dayan, T., D. Simberloff, E. Tchernov, and Y. Yom-Tov. 1989b. In- 106:1–19. ter- and intraspecific character displacement in mustelids. Gittleman, J. L., and P. H. Harvey. 1982. Carnivore home-range Ecology 70:1526–1539. size, metabolic needs, and ecology. Behavioral Ecology and ———. 1990. Feline canines: community-wide character dis- Sociobiology 10:57–63. placement among the small cats of Israel. American Natural- Glut, D. F. 1997. Dinosaurs, the encyclopedia. McFarland, Jef- ist 136:39–60. ferson, N.C. ———. 1992. Canine carnassials: character displacement in the Gradzinski, R., and T. Jerzykiewicz. 1972. Additional geograph- DINOSAURIAN AND MAMMALIAN PREDATORS 543

ical and geological data from the Polish-Mongolian palaeon- the Oldman Formation (Cretaceous) of Alberta. Canadian tological expeditions. Acta Palaeontologia Polonica 27:17–30. Journal of Earth Sciences 6:595–612. Gradzinski, R., J. Kazmierczak, and J. Lefeld. 1968. Geograph- ———. 1970. Tyrannosaurs from the Late Cretaceous of western ical and geological data from the Polish-Mongolian palaeon- Canada. Publications in Palaeontology, National Museum of tological expeditions. Acta Palaeontologia Polonica 19:33–82. Natural Sciences (Ottawa) 1:1–34. Haynes, G. 1982. Utilization and skeletal disturbances of North Seidensticker, J. 1976. On the ecological separation between ti- American prey carcasses. Arctic 35:266–281. gers and leopards. Biotropica 8:225–234. Hutchinson, G. E., and R. H. MacArthur. 1959. A theoretical eco- Seidensticker, J., M. E. Sunquist, and C. McDougal. 1990. Leop- logical model of size distributions among species of animals. ards living at the edge of Royal Chitawan National Park, Ne- pal. Pp. 415–423 in J. C. Daniel and J. S. Serrao, eds. Conser- American Naturalist 93:117–125. vation in developing countries: problems and prospects. Bom- Hutchinson, J. R., and M. Garcia. 2002. Tyrannosaurus was not a bay Natural History Society, Bombay. fast runner. Nature 415:1018–1021. Simberloff, D., and W. Boecklen. 1981. Santa Rosalia reconsid- Jaksic, F., H. W. Greene, and J. L. Yanez. 1981. The guild struc- ered: size ratios and competition. Evolution 35:1206–1228. ture of a community of predatory vertebrates in central Chile. Simberloff, D., and T. Dayan. 1991. The guild concept and the Oecologia 49:21–28. structure of ecological communities. Annual Review of Ecol- Kiltie, R. F. 1988. Interspecific size regularities in tropical felid ogy and Systematics 22:115–143. assemblages. Oecologia 76:97–105. Stanley, S. M. 1979. Macroevolution: pattern and process. W. H. Kobayashi, Y., J.-C. Lu, Z.-M. Dong, R. Barsbold, Y. Azuma, and Freeman, San Francisco. Y. Tomida. 1999. Herbivorous diet in an ornithomimid dino- Tiffney, B. H. 1984. Seed size, dispersal syndromes, and the rise saurs. Nature 402:480–481. of the angiosperms: evidence and hypothesis. Annals of the Kruuk, H. 1972. The spotted hyena. University of Chicago Press, Missouri Botanical Garden 71:551–576. Chicago. Turner, C. E., and F. Peterson. 1999. Biostratigraphy of dino- Losos, J. B., and H. W. Greene. 1988. Ecological and evolutionary saurs in the upper Jurassic Morrison Formation of the western implications of diet in monitor lizards. Biological Journal of interior, U.S.A. Pp. 77–114 in D. D. Gillette, ed. Vertebrate pa- the Linnean Society 35:379–407. leontology in Utah. Utah Geological Survey Miscellaneous Madsen, J. H., Jr., and S. P. Welles. 2000. Ceratosaurus (Dinosau- Publication 99-1. ria, Theropoda) a revised osteology. Miscellaneous Publica- Van Valen, L. 1973. Body size and numbers of plants and ani- tion, Utah Geological Survey, 00–2:1–80. mals. Evolution 27:27–35. Maleev, E. A. 1974. Giganskei karnozavrii semeistva Tyranno- Van Valkenburgh, B. 1985. Locomotor diversity within past and sauridae. Sovmestnaia Sovetsko-Mongoliskaia Paleontologi- present guilds of large predatory mammals. Paleobiology 11: cheskaia Yekspeditsiia, Trudi 1:132–191. 406–428. ———. 1988. Trophic diversity within past and present guilds Molnar, R. E. 2000. Mechanical factors in the design of the skull of large predatory mammals. Paleobiology 14:156–173. of Tyrannosaurus rex (Osborn, 1905). Gaia 15:193–218. ———. 1989. Carnivore dental adaptations and diet: a study of Muckenhirn, N. A., and J. F. Eisenberg. 1973. Home ranges and trophic diversity within guilds. Pp. 410–436 in J. L. Gittleman, predation of the Ceylon leopard. Pp. 142–175 in R. L. Eaton, ed. Carnivore behavior, ecology, and evolution, Vol. I. Cornell ed. The world’s cats, Vol. 1. Winston Wildlife Safari, Winston, University Press, Ithaca, N.Y. Ore. ———. 1991. Iterative evolution of hypercarnivory in canids Osborn, H. F. 1916. Skeletal adaptations of Ornitholestes, Stru- (Mammalia: Canidae): evolutionary interactions among sym- thiomimus, Tyrannosaurus. Bulletin of the American Museum patric predators. Paleobiology 17:340–362. of Natural History 35:733–771. ———. 1994. Extinction and replacement among predatory Osmolska, H. 1996. An unusual theropod dinosaur from the mammals in the North American Late Eocene–Oligocene: Late Cretaceous Nemegt Formation of Mongolia. Acta Pa- tracking a guild over twelve million years. Historical Biology laeontologica Polonica 41:1–38. 8:1–22. Ostrom, P. H., S. A. Macko, M. H. Engel, and D. A. Russell. 1993. ———. 2001. The dog-eat-dog world of carnivores: a review of Assessment of trophic structure of Cretaceous communities past and present carnivore community dynamics. Pp. 101–121 based on stable nitrogen isotope analysis. Geology 21:491– in C. Stanford and H. T. Bunn, eds. Meat-eating and human 494. evolution. Oxford University Press, Oxford. Palomares, F., and T. M. Caro. 1999. Interspecific killing among Van Valkenburgh, B., and F. Hertel. 1998. The decline of North mammalian carnivores. American Naturalist 153:492–508. American predators during the Late Pleistocene. In J. J. Saun- Paul, G. S. 1988. Predatory dinosaurs of the world. Simon and ders, B. W. Styles, and G. F. Baryshnikov, eds. Quaternary pa- Schuster, New York. leozoology in the Northern Hemisphere. Illinois State Muse- Peczkis, J. 1994. Implications of body-mass estimates for dino- um Scientific Papers 27:357–374. saurs. Journal of Vertebrate Paleontology 14:520–533. Van Valkenburgh, B., and K. Koepfli. 1993. Cranial and dental adaptations for predation in canids. In N. Dunstone and M. Pianka, E. R. 1994. Comparative ecology of Va ra nus in the Great L. Gorman, eds. Mammals as predators. Symposia of the Zoo- Australian Desert. Australian Journal of Ecology 19:395–408. logical Society of London 65:15–37. Oxford University Press, Polis, G. A., and R. D. Holt. 1992. Intraguild predation: the dy- Oxford. namics of complex trophic interactions. Trends in Ecology Van Valkenburgh, B., and C. B. Ruff. 1987. Canine tooth strength and Evolution 7:151–154. and killing behaviour in large carnivores. Journal of Zoology Rayfield, E. J., D. B. Norman, C. C. Horner, J. R. Horner, P. M. 212:379–398. Smith, J. J. Thomason, and P. Upchurch. 2001. Cranial design Weishampel, D. B. 1990. Dinosaurian distribution. Pp. 63–139 in and function in a large theropod dinosaur. Nature 409:1033– D. B. Weishampel, P. Dodson and H. Osmolska, eds. The Di- 1037. nosauria. University of California Press, Berkeley. Root, R. B. 1967. The niche exploitation pattern of the blue-gray Wing, S. L., and B. H. Tiffney. 1987. The reciprocal interaction gnatcatcher. Ecological Monographs 37:317–350. of angiosperm evolution and tetrapod herbivory. Review of Russell, D. A. 1969. A new specimen of Stenonychosaurus from Palaeobotany and Palynology 50:179–210.