Farlow et al, 2000

640

AMER. ZOOL., 40:640–663 (2000)

Theropod Locomotion1

JAMES O. FARLOW,2,* STEPHEN M. GATESY,† THOMAS R. HOLTZ, JR.,‡JOHN R. HUTCHINSON,§ AND JOHN M. ROBINSON\

*Department of Geosciences, Indiana-Purdue University at Fort Wayne, Indiana 46805†Department of Ecology and Evolutionary Biology, Brown University, Providence, Rhode Island 02912

‡Department of Geology, University of Maryland, College Park, Maryland 20742§Department of Integrative Biology, University of California, Berkeley, California 94720

\Department of Physics, Indiana-Purdue University at Fort Wayne, Indiana 46805

SYNOPSIS. Theropod (carnivorous) dinosaurs spanned a range from chicken-sizedto elephant-sized animals. The primary mode of locomotion in these dinosaurs wasfairly conservative: Theropods were erect, digitigrade, striding bipeds. Even so,during theropod evolution there were changes in the hip, tail, and hindlimb thatundoubtedly affected the way these dinosaurs walked and ran, a trend that reachedits extreme in the evolution of birds. Some derived non-avian theropods developedhindlimb proportions that suggest a greater degree of cursoriality than in moreprimitive groups. Despite this, fossilized trackways provide no evidence for changesin stride lengths of early as opposed to later non-avian theropods. However, thesedinosaurs did take relatively longer strides—at least compared with footprintlength—than bipedal ornithischian dinosaurs or ground birds. Judging from track-way evidence, non-avian theropods usually walked, and seldom used faster gaits.The largest theropods were probably not as fleet as their smaller relatives.

INTRODUCTION

Zoologists can relatively easily observethe motions of many extant animals in thefield or the laboratory. For extinct species,particularly those with a body form not ex-actly like those of any living species, it isa different story. An extinct animal’s loco-motion cannot be observed directly; rather,the paleontologist must reconstruct how thecreature may have moved from its mor-phology and other indirect sources of in-formation.

In this paper, we summarize what hasbeen inferred about the locomotion of car-nivorous dinosaurs (i.e., theropods; Fig. 1),based on the study of both body and tracefossils. Understanding how theropodsmoved is obviously necessary if we are tounderstand them as living animals. Giventhe likelihood that birds represent an extantsubgroup of theropods, however, it is equal-ly important that we understand how non-avian theropods walked and ran in order to

1 From the Symposium Evolutionary Origin ofFeathers presented at the Annual Meeting of the So-ciety for Integrative and Comparative Biology, 6–10January 1999, at Denver, Colorado.

2 E-mail: [email protected]

know which features of avian locomotionoriginated with (and within) birds, andwhich they inherited from their non-avianancestors.

Our study will therefore consider thero-pod locomotion in a phylogenetic context.We will compare and contrast features oflocomotion among different theropodgroups, including ground-living birds. Al-though it is reasonable to suppose that the-ropods could have swum well enough whenthe situation warranted (Coombs [1980] de-scribed possible trackway evidence forthis), and it is conceivable that some small-bodied theropods could have scamperedabout in trees—at least on occasion—with-out the need for obvious scansorial adap-tations, we will restrict ourselves to consid-ering what was undoubtedly the most im-portant kind of locomotion for most non-avian theropods, namely walking andrunning on the ground.

THEROPOD PHYLOGENY

Modern studies of theropod phylogenybegan with the question of the origin ofbirds, following Ostrom’s observations(e.g., 1975, 1976) of numerous featuresuniquely shared by primitive birds and

641THEROPOD LOCOMOTION

FIG. 1. Life restorations of representatives of the dinosaur groups discussed in this paper. Not to scale. A)Coelophysis (coelophysoid ceratosaur). Length ca. 3 m. B) Allosaurus (carnosaur). Length ca. 8 m. C) Gorgo-saurus (tyrannosaurid). Length ca. 8 m. D) Struthiomimus (ornithomimosaur). Length ca. 3.5 m. E) Deinonychus(dromaeosaurid). Length ca. 3 m. F) Archaeopteryx (basal bird). Length ca. 0.5 m. G) Dinornis (Dinornithifor-mes; moa). Height ca. 2.5 m. H) Iguanodon (large facultatively bipedal ornithischian). Length ca. 9 m. Drawingsby James Whitcraft.

various theropod dinosaurs. Most of thecladistic studies of the 1980s (Thulborn,1984a; Paul, 1984; Gauthier, 1986) fo-cused on the distribution of avian featuresamong various groups of carnivorous di-nosaurs. Subsequent workers conductednumerical cladistic analyses to examinethe distribution of derived character statesin the various subgroups of non-avian the-ropods (Rowe and Gauthier, 1990; Russelland Dong, 1993; Perez-Moreno et al.,1993; Holtz, 1994, 1996, 2000; Sereno etal., 1994, 1996, 1998; Novas, 1996; Char-ig and Milner, 1997; Sereno, 1997, 1999;Sues, 1997; Harris, 1998a; Makovickyand Sues, 1998; Padian et al., 1999). Al-

though these analyses have not producedidentical trees, due to differences in taxonand character choice and coding, most re-sult in very similar patterns. The phylo-genetic tree presented here (Fig. 2) rep-resents a broad consensus of the two mostcomprehensive studies (Sereno, 1997;Holtz, 2000).

There is disagreement over whether Eor-aptor and the Herrerasauridae are true the-ropods (Novas, 1993, 1996; Sereno and No-vas, 1992, 1993; Bonaparte and Pumares,1995; Sereno, 1997). However, all of thesestudies agree that Eoraptor and the herrer-asaurids more closely resemble the ances-tral dinosaurian condition, in terms of their

642 J. O. FARLOW ET AL.

FIG. 2. A) Phylogenetic tree of Theropoda; see Holtz (2000) for details. *Asterisk indicates two equally par-simonious positions for Troodontidae. B) Conflict and congruence in current coelurosaur cladograms. Heavysolid lines represent the primary phylogenetic conclusions of Gauthier (1986), which were supported in the otherstudies in question. These studies differ, however, in their placement of Therizinosauroidea (Theriz.), Troodon-tidae (Troo.), and Tyrannosauridae (Tyr.). The thin solid lines represent results from Holtz (1994, 2000); thedotted lines from Sereno (1997); the dashed lines from Makovicky and Sues (1998).

locomotor apparatus, than do the more ad-vanced theropods.

Derived theropods constitute two majorclades, Ceratosauria and Tetanurae (Gau-thier, 1986; Sereno, 1997; Holtz, 2000).The two clades together comprise Neoth-eropoda.

Ceratosauria is divided into two groups:the more gracile, long-necked coelophy-soids (e.g., the Late Triassic Coelophysisand the Early Jurassic Syntarsus) and themore robust neoceratosaurs of the Late Ju-rassic (Ceratosaurus) and Cretaceous (abel-

isaurids). Although neoceratosaurs sharewith tetanurines some features not found incoelophysoids, the current weight of evi-dence supports uniting neoceratosaurs andcoelophysoids in a monophyletic group(Holtz, 2000).

Most of the tetanurine groups belong tothe clade Avetheropoda (or Neotetanurae;Sereno et al., 1994), but several taxa pos-sess tetanurine features yet lack the derivedtraits of avetheropods. Most of these are‘‘megalosaurs’’ (e.g., Torvosaurus, Eus-treptospondylus, Piatnitzkysaurus, and Me-


galosaurus) that lack any particular distri-bution of derived features favoring onephylogenetic arrangement over others.

Avetheropods are divided into two majorgroups, Carnosauria and Coelurosauria.Carnosaurs are united mainly on cranialfeatures, and for the most part do not differin locomotor features from the condition ofmore basal tetanurines; this group includesthe well-known Allosaurus and giganticforms like Carcharodontosaurus and Gi-ganotosaurus.

Coelurosaurs are the most anatomicallydiverse group of theropods, even if birdsare excluded from consideration. A basalplexus includes small and large forms likeOrnitholestes, Proceratosaurus, Dryptosau-rus, and Deltadromeus that lack the featuresof derived coelurosaurs. The latter consti-tute the Maniraptoriformes. Although Ser-eno (1997), Makovicky and Sues (1998),and Holtz (2000) disagree over particularrelationships within this advanced coeluro-saurian clade (Fig. 2), they agree on thecomposition of this taxon, which includesmany of the most specialized groups of the-ropods: ornithomimosaurs; tyrannosaurids,troodontids, therezinosauroids, oviraptoro-saurs, dromaeosaurids, and avialians (livingbirds and their extinct relatives).

Inclusion of birds among the coeluro-saurs is a contentious matter (see, for ex-ample, Welman [1995] and Feduccia [1996]for dissenting views). Debate over this topicis as much about methodology as it is aboutevolutionary relationships per se. Shouldfunctional explanations be predicated onphylogenies, or should functional interpre-tations provide a means of evaluating phy-logenetic hypotheses? Is it reasonable to leta phylogeny based on a large number ofcharacters be ‘‘trumped’’ by one or a fewputatively paramount characters seeminglyat variance with it (see, e.g., Burke andFeduccia, 1997; Ruben et al., 1997; Wagnerand Gauthier, 1999; Feduccia, 1999)?

Such questions cannot be answered in apaper like this. However, the majority opin-ion among us is that a close phylogeneticrelationship between birds and coelurosaurs(particularly dromaeosaurids) is the mostrobust hypothesis about avian origins pres-ently available (Gauthier, 1986; Holtz,

1994, 2000; Chatterjee, 1997; Forster et al.,1998; Sereno, 1997; Padian and Chiappe,1998). If true, then some of the character-istic features of birds that are related to lo-comotion, such as striding bipedalism, dig-itigrady, elongated hindlimbs, consolidationand fusion of limb bones, thin-walled longbones, furculae, specialized carpal bones, alarge ilium, retroverted pubis, an increasedsacral vertebral count, and a reduced tail,all arose within non-avian, derived thero-pod dinosaurs. If this hypothesis turns outto be wrong, and birds in fact arose fromvery basal theropods, basal dinosaurs, oreven pre-dinosaurian archosaurs, the manyfeatures that advanced coelurosaurs sharewith primitive birds would constitute a re-markable case of parallel or convergentevolution—perhaps surpassing even that ofnimravid and neofelid sabertooths amongpredatory mammals (Martin, 1998).

To avoid cladistically correct but awk-ward, space-consuming circumlocutionssuch as ‘‘non-avian theropods,’’ in the re-mainder of this paper we will use the term‘‘theropod’’ in quotation marks to refer tomembers of the Theropoda exclusive ofbirds. Without quotation marks the term isused in a monophyletic sense.

TRENDS AND VARIATIONS IN THE

LOCOMOTORY ANATOMY OF THEROPODS

‘‘Theropods’’ shared a common loco-motor design. All were digitigrade bipedswith fully adducted hindlimbs and function-ally tridactyl feet. This commonality can beseen in the relative lengths of the femur,tibia, and metatarsus: Despite an enormousrange in body sizes, ‘‘theropods’’ all hadvery similar hind limb proportions, com-pared to the wide range of proportionsevolved by birds (Gatesy and Middleton,1997; Carrano and Sidor, 1999); there is noevidence among ‘‘theropods’’ for limbshighly specialized for diving, paddling,wading, perching, or vertical clinging. Thisconservatism supports the notion that ‘‘the-ropods’’ were primarily adapted for terres-trial bipedalism. Even so, there are distinc-tive variations on the overall ‘‘theropod’’morphological theme among the variousgroups.

The tails of basal members of most di-


nosaur lineages, like those of other reptiles,do not show marked differentiation intosections, beyond diminution of the size ofthe centrum, neural spine, and chevron sizeposteriorly. ‘‘Theropod’’ caudal vertebrae,however, show a marked trend towardtransformation of the tail into differentfunctional sections. ‘‘Theropod’’ tails havea distinct ‘‘transition point’’ (Russell, 1972;Gauthier, 1986). The neural arches andtransverse processes become strongly re-duced tailward, and are absent in at leastthe distal half of the tail. In tetanurines thistransition point occurs in the proximal halfof the caudal column, and in avetheropodsis accompanied by a dramatic change inchevron shape: Distal chevrons lose thetypical gentle curve of the primitive con-dition to take a distinct ‘‘L’’-shape. The pre-zygapophyses are elongated in the distalcaudal vertebrae of avetheropods, increas-ing the interlocked nature of the posteriortail. In advanced coelurosaurs the numberof caudals is reduced from the primitivecount of 50 or more to 44 or fewer, theprezygapophyses extend more than half thelength of the preceding vertebra, and thedistalmost chevrons are even more trans-formed, becoming very shallow dorsoven-trally but elongated anteroposteriorly. Introodontids, dromaeosaurids, and avialiansthe transition point migrates even furtherheadward, becoming located somewhere inthe first nine caudals. This trend was takenfurther still in Confuciusornis and ornith-othoracine birds, in which the distalmostcaudals are fully interlocked as a singleunit, the pygostyle.

These changes result in a tail that isstouter and more mobile proximally, butthinner and stiffer distally. This segmenta-tion presumably transformed the distal por-tion of the tail into a dynamic stabilizer,perhaps as an aid during turning (Gauthier,1986), that in dromaeosaurids may alsohave been employed for counterbalancingthe body while attacking prey (Ostrom,1969), in a manner reminiscent of a tight-rope-walker’s pole. Development of this dy-namic stabilizer incorporated progressivelymore of the tail, from the primitive condi-tion in ceratosaurs through basal tetanurines

and carnosaurs, to more profound transfor-mations in coelurosaurs.

Changes in the size and shape of the tailhad additional effects on locomotion. Thetail’s length and mass affected the way‘‘theropods’’ balanced. As obligate bipeds,‘‘theropods’’ had to stand with the body’scenter of mass over the feet in order tomaintain equilibrium. A nose-down pitchcaused by the body in front of the hip jointwas counterbalanced, at least in part, by thenose-up rotation imparted by the tail. Thetail’s ability to counter the weight of thefront part of the body depended on its mass,as well as on the distribution of that massalong its length. ‘‘Theropods’’ more closelyrelated to birds underwent a reduction intail size (Gatesy, 1995; Gatesy and Dial,1996). The above-noted progressive reduc-tion in the number of caudal vertebrae wasaccompanied by a reduction in tail diame-ter, particularly distally. These changes like-ly affected the location of the body’s centerof mass, shifting it forward relative to thehip joint and forcing the limbs to reorientin order to position the feet further forward.

In addition, changes in tail size have im-plications for the function of the caudofe-moral musculature (Gatesy, 1990). The cau-dofemoralis longus is a muscle runningfrom the tail to the femur in living saurians(i.e., lizards, Sphenodon, crocodilians, andbirds). Because of its large size and attach-ments, this muscle plays an important roleduring locomotion in both lizards (Reilly,1995, 1998) and crocodilians (Gatesy,1990). Osteological evidence of both originand insertion indicates that a substantialcaudofemoralis longus was the ancestralcondition for ‘‘theropods.’’

‘‘Theropods’’ more closely related tobirds (e.g., ornithomimosaurs, dromaeo-saurids, troodontids) show features indica-tive of a reduction in caudofemoral mus-culature. The smaller number of caudal ver-tebrae, reduced tranverse processes, anddistal specialization of the tail restricted theorigin of the caudofemoralis longus to asmaller area of the tail base. Similarly, re-duction or loss of the caudofemoral inser-tion scar, the fourth trochanter of the femur,indicates a reduction in muscle size. Thispattern of change suggests that the caudo-


femoral retraction system was primitivelyimportant in ‘‘theropods,’’ but became lessso in forms closely related to birds. Birdscontinued this trend by reducing the tail stillfurther, and losing the caudofemoralis lon-gus entirely in some species.

Theropod hips and hindlimbs showmarked morphological changes (Fig. 3) thatare consilient with functional changes dur-ing their evolution: (1) The antitrochanterrepositioned from its primitive archosaurianlocation on the ischium, facing craniodor-sally, to a more craniolateral orientation onthe ischium and ilium in dinosaurs and theirclosest relatives. The antitrochanter then en-larged and re-oriented to face cranioven-trally in birds. (2) The femoral head shiftedfrom a craniomedial orientation in basaltheropods to a more offset medial orienta-tion in avetheropods, especially birds. (3)The ectocondylar tuber of the distal femurenlarged in ‘‘theropods’’ and moved distal-ly from the proximal popliteal region ontothe distal lateral condyle in birds. (4) Themain weight-bearing axis of the crus shiftedmedially in theropods onto the tibia as thefibula and calcaneum were reduced, and el-ements of the knee and ankle joint becamemore rigidly appressed. (5) The fibular tu-bercle, the insertion of the knee flexor M.ilio-fibularis (Muller and Streicher, 1989),moved from a plesiomorphic craniolateralposition on the proximal fibula in ‘‘thero-pods’’ to a caudolateral position in birds,consistent with a change in the action ofthis muscle related to increased knee flex-ion.

Unlike human beings, which have a dis-tinctive walk/run transition, a shift betweengaits is less clear in many ground dwellingbirds (Gatesy and Biewener, 1991). Al-though the presence of an aerial phase istraditionally used to define running, birdsare able to move very quickly without anaerial phase. Guineafowl, for example, donot employ an aerial phase below speeds of2 m/sec (Gatesy, 1999). It appears thatbirds, particularly small forms, are able tomaintain contact with the ground even athigh speeds because of their highly com-pliant limbs. Such a ‘‘soft’’ limb spring al-lows birds to use a running mechanism (de-fined by the energy fluctuations of the cen-

ter of mass), while maintaining a doublesupport phase. Such a running style hasbeen called ‘‘Groucho running’’ (McMahonet al., 1987) after the famous Marx broth-er’s bent-legged locomotor style. A ques-tion that remains unanswered is when thiscompliant form of avian locomotionevolved. Did all theropods have it, or is itunique to birds? An increase in limb com-pliance may have accompanied the reori-entation of the femur into a more horizontalposition, but this hypothesis awaits testing.

Holtz (1995) showed that arctometatar-salians (tyrannosaurids, ornithomimosaurs,and possibly certain other derived coeluro-saurs) markedly increased the length of dis-tal hindlimb elements relative to the femurlength (Fig. 4A), as compared with other‘‘theropods.’’ In addition, these derived‘‘theropods’’ also developed a tightly inter-locked proximal metatarsus that more ef-fectively transmitted locomotory forcesfrom the foot to the lower leg than did theancestral theropod metatarsus. Holtz con-cluded that arctometatarsalians were prob-ably capable of a greater degree of curso-riality than were other ‘‘theropods.’’

Cursoriality is like pornography, how-ever: We’re not sure what it is, yet we thinkwe know it when we see it in animal mor-phology. How do ‘‘cursorial’’ adaptationsactually relate to locomotor performance?Would more cursorial animals be expectedto have a longer stride, and a faster topspeed and/or walking speed, than less cur-sorial animals of the same body size? Iscursoriality related mainly to maximalsprint speed, or stamina, or the cost oftransport (Garland and Janis, 1993; Janisand Wilhelm, 1993; Steudel and Beattie,1995; Carrano, 1999)? The answers to thesequestions are not clear, and this makes in-terpretation of the locomotor performanceof extinct animals like ‘‘theropods’’ espe-cially difficult.

FOOTPRINTS, TRACKWAYS, LIMB CARRIAGE,AND GAITS

Much can be learned about ‘‘theropod’’locomotion from anatomical studies of skel-etons, but such inferences are necessarilyindirect. Fossilized footprints and track-ways directly record the movements of the


FIG. 3. ‘‘Theropod’’ (left) and bird (right; kiwi [Apteryx]; after McGowan, 1979) hindlimb elements, all fromthe right side. A) Pelves in lateral view: Coelophysis (after Rowe and Gauthier, 1990) and kiwi; arrow roughlyindicates the orientation of the antitrochanter. B) Femora in caudal view; Syntarsus (after Rowe and Gauthier,1990) and kiwi; arrow roughly indicates the orientation of the femoral head. C) Allosaurus (after Madsen 1993)fibula in medial view (left) and kiwi tibiotarsus in lateral view. Abbreviations: ant 5 antitrochanter, ect 5ectocondylar tuber, ff 5 fibular fossa, fh 5 femoral head, fib 5 fibula, tfc 5 tibiofibular crest (homolog ofectocondylar tuber; Chiappe 1996), tib 5 tibia.

animals that made them, thus complement-ing the skeletal record.

Footprint shape and hindlimb carriage

Although trackways of small quadrupedshave been attributed to juvenile ‘‘thero-pods’’ (Wright, 1996), quadrupedal loco-motion was probably rare at best in carniv-orous dinosaurs. More common, but still in-frequent, are trackways attributed to ‘‘the-ropods’’ that were walking flat-footed, withthe entire metatarsus touching the ground(Thulborn and Wade, 1984; Kuban, 1989;Perez-Lorente, 1993). Most ‘‘theropod’’footprints confirm the conclusion thatwould be drawn from pedal osteology, thatthese dinosaurs usually walked in a digiti-grade fashion.

Despite the greater number of phalanges(five) in the outer digit IV than in the innerdigit II (three phalanges) of the main toesin both ‘‘theropods’’ and most large groundbirds, the aggregate lengths of these toes aremuch the same. The ostrich, which has lostdigit II, is a dramatic exception. However,this symmetry is not seen in most well-pre-served footprints attributed to ‘‘theropods.’’The impression made by digit II is com-monly much shorter than that of digit IV,resulting in a conspicuous notch along themedial edge of the print (Fig. 5A, B). Theposterior margin of the ‘‘theropod’’ foot-print seems to have been made by a padbeneath the metatarsophalangeal joint ofdigit IV (Baird, 1957; Thulborn, 1990). Themost proximal digital pad of both digits II


FIG. 4. Limb proportions in ‘‘theropods.’’ A) Plot of the length of metatarsal III against femur length. Notethe tendency for arctometatarsalian coelurosaurs (tyrannosaurids and ornithomimosaurs, and possibly troodon-tids), as compared with other groups, to lengthen the metatarsus relative to the femur. Modified from Holtz(1995). The regression line is for arctometatarsalians: log 10 metatarsal III length 5 0.664 log 10 femur length1 0.765; r2 5 0.91. However, this and other regression equations reported in this paper should be taken withcaution; data cases are individual specimens, and particular species are represented by different sample sizes.For a list of the genera that represent the various groups in this and other graphs, see Appendix 1. B) Total leglength (femur 1 tibia 1 metatarsal III) as a function of the aggregate length of the phalanges of digit III. Incontrast to the comparison of proximal and distal limb segment lengths, the present relationship shows no markeddifference between arctometatarsalians and other ‘‘theropods.’’ Regression line for arctometatarsalians: log 10total leg length 5 0.864 log 10 digit III length 1 1.088; r2 5 0.966.

and III that is commonly recorded in foot-prints was situated beneath the joint be-tween the first and second phalanges of thedigit.

In order to make prints of this shape, themetatarsophalangeal joints of digits II andIII had to have been held clear of theground, while digit IV was impressed over


FIG. 5. Footprints of ‘‘theropods’’ and large ground birds. A) Topographic map of a large Early Cretaceous‘‘theropod’’ footprint from the Paluxy River of Texas. Arrow points to the notch along the inner rear margin ofthe digit II impression created by failure of the proximal end of phalanx 1 of digit II to impress. B) Tyranno-sauripus (by convention, names assigned to footprints refer to the prints themselves, and not to the creaturesthat made them), a very large Late Cretaceous footprint from New Mexico. Due to the great depth of the track,there is a clear (and unusual) impression of digit I (left of the arrow). Arrow points to the medial embaymentalong the inner rear margin of the print. Scale bar 5 50 cm. Redrawn from Lockley and Hunt (1994). C)Footprint of an emu (Dromaius novaehollandiae). Note the well-developed, symmetrically positioned metatarsalpad impression in this and the next footprint. D) Moa footprint from the Pleistocene of New Zealand.

its entire length. This resulted in an asym-metry of the proximal (posterior) portion ofthe footprint that can be seen in ‘‘theropod’’prints small and large. (We will designatethe proximal end of ‘‘theropod’’ and birdfootprints as the ‘‘heel’’ for descriptive pur-poses, but it should be kept in mind thatthis region of the print was not made by theanatomical heel, which would usually havebeen held clear of the ground in these dig-itigrade animals.) The same asymmetry canbe seen in footprints (e.g., Anomoepus) at-tributed to primitive ornithischians (Lull,1953; Thulborn, 1990, 1994; Olsen, 1995),and so this may be a primitive feature fordinosaurs generally.

In contrast, the proximal end of Cenozoic

ground bird footprints is much more sym-metrical (Fig. 5C, D). The ‘‘heel’’ is formedby a thick metatarsal pad located beneaththe distal end of the digit III trochlea of thetarsometatarsus (Lucas and Stettenheim[1972]; Cannon [1996]; in ostriches, how-ever, this pad commonly does not touch theground). The metatarsophalangeal joints ofdigits II and IV are elevated off the ground.

Padian and Olsen (1989) observed thatthe proximal ends of the digits do not reg-ister in footprints of rheas and other ratites.They noted the relatively longer metatarsusof birds than of ‘‘theropods,’’ and suggestedthat a longer metatarsus would have to bemore vertically oriented than in typical‘‘theropods’’ to keep the animal’s center of


mass positioned above the feet. Padian andOlsen (1989) further speculated that theavian metatarsus might have elongated ‘‘tomove the center of support farther forwardto compensate for (1) the increased pectoralmuscle mass of birds . . . and (2) the loss ofthe long, fleshy tail of their dinosaurian an-cestors’’ (Padian and Olsen, 1989, p. 234).

However, the situation is probably morecomplicated than that. Lengthening themetatarsus with no other changes in limbsegment lengths or inter-segmental angleswould indeed shift the animal’s mass back-wards, causing the animal to tend to fallbackward with respect to the point whereits foot touched the ground. Shortening thetail with no other changes, in contrast,would cause the animal to fall forward.Having both changes occur simultaneouslymight cancel out their effects, obviating theneed for a more vertical metatarsus.

Another complicating factor, though, isthe size and orientation of the femur. Padianand Olsen (1989) argued that ‘‘theropods,’’like birds, held the femur in a nearly hori-zontal fashion. In contrast, Gatesy (1991)noted the relatively elongate, slender shapeof the femora in ‘‘theropods’’ as comparedwith avian femora, and argued that thethigh in ‘‘theropods’’ was not stoutlyenough constructed to withstand the stress-es to which it would have been subjectedhad it been held horizontally. Carrano(1998) extended this argument by demon-strating that a horizontal femur experiencesconsiderably greater torsional strains than avertical femur. In birds, this results in avery short, stout femur that is very unlikethe relatively slender femur of non-aviantheropods (Gatesy, 1991; Carrano, 1998).

In birds, most of the motion of the legduring protraction–retraction at slow speedsoccurs at the knee, rather than the hip (Tar-sitano, 1983; Gatesy, 1990; Carrano, 1998).The considerable shortening of the tail inbirds forces the center of mass forward, anda walking bird balances its weight over theknee, and not the hip. Perhaps this has con-tributed to the more vertical orientation ofthe metatarsus in birds than in ‘‘theropods,’’but how the more horizontal femur wouldinteract with tail shortening and metatarsal

lengthening to elevate the latter bone is un-clear.

Yet another change during theropod/birdevolution was the above-mentioned reduc-tion of the fibula and the increased impor-tance of the tibia in weight-bearing. Earlyarchosaurs had rather large fibulae, andstood/moved with digit I and V touchingthe ground (Sereno, 1991). Theropodschanged this by (1) clasping the fibula tothe tibia with a big crest, ligaments, and aproximally shifted interosseus muscle, (2)lifting digits I and V off the ground, and(3) linking the tibia and the astragalus viathe ascending process of the latter bone,while reducing the fibula and calcaneum.More derived birds than Archaeopteryxeventually reduced the fibula still further,and fusion of the metatarsals into a singletarsometatarsus may have occurred in as-sociation with a more evenly distributedcarriage of body weight across the proximalend of the foot as the metatarsus becameelevated.

In any case, a more vertical orientationof the metatarsus is very likely the reasonfor the more symmetrical shape of the‘‘heel’’ in footprints of birds as opposed tomost ‘‘theropods’’; a more vertical carriageof the metatarsus would presumably lift themetatarsophalangeal joint of digit IV off theground. Further evidence for differences inmetatarsal carriage comes from differencesin shape between footprints made by ‘‘the-ropods’’ and birds walking on very softsubstrates (Gatesy et al., 1999). It would beinteresting to know when in the course oftheropod evolution this elevation of themetatarsus happened. Would footprints ofthe first birds have looked like those ofmost small ‘‘theropods,’’ with an asymmet-rical ‘‘heel’’? Did the evolution of a pro-nounced metatarsal pad at the rear edge ofthe print occur after birds emerged as a dis-tinct clade? Or would such features havebeen seen in prints made by ‘‘theropods’’close to birds?

‘‘Theropod’’ and Cenozoic ground birdfootprints commonly differ in the angleformed by the impressions of digits II andIV (Table 1). ‘‘Theropod’’ prints tend tohave low values of this interdigital angle,while prints of birds often have higher val-


TABLE 1. Values of interdigital angle II–IV (measured as best-fit lines through digit impressions in photographsor on tracings drawn from footprint casts on plastic sheets; numbers in degrees) in footprints of selected largebirds and dinosaurs. Data from the senior author’s unpublished observations. Moa footprints are from thePleistocene of New Zealand, and dromornithid prints from the mid-Tertiary of Tasmania. Dinosaur footprintsfrom the Amherst College Collection and Dinosaur State Park are Early Jurassic, and dinosaur tracks from theF6 Ranch and Glen Rose Limestone sites are of Early Cretaceous age.

Trackmaker(s) Minimum Maximum MeanNumber offootprints

Female greater rhea (Rhea americana)Male lesser rhea (Rhea pennata)Juvenile emus (Dromaius novaehollandiae)a

Adult emusb

Cassowaries (Casuarius casuarius)c

Kiwis (Apteryx australis)d

Unidentified moaUnidentified dromornithids

4238575536

1046478

7749

111119

51119109125

57.044.085.180.846.5

111.389.9

101.4

73

4153

64

105

Phasianidse

Male kori bustard (Ardeotis kori)Amherst College 9/14 Anchisauripus

576121

1448248

110.873.729.2

563

38Dinosaur State Park (Rocky Hill, CT) ‘‘theropods’’ 31 87 54.3 113F6 Ranch site (Fort Terrett Formation, Kimble County, Texas)

‘‘theropods’’ 17 53 32.1 25Glen Rose Limestone (various localities, central Texas)

‘‘theropods’’ 29 54 40.1 28

a Data for 21 captive individuals.b Data for at least 16 captive or free-living individuals.c Data for two captive individuals.d Data for two captive individuals.e Data for at least 26 captive individuals of 19 species.

ues. However, there is overlap between thetwo groups; rheas and cassowaries havelow interdigital angles for birds, and someprints attributed to ‘‘theropods’’ have inter-digital angles as large as those of any bird(Harris, 1998b).

Several types of small Mesozoic foot-prints have been attributed to birds (Lock-ley et al., 1992), and some of these showfeatures like those of Cenozoic bird prints.Interdigital angle II–IV is usually ratherlarge in footprints attributed to Mesozoicbirds, and digits II and IV extend outwardfrom a symmetrical ‘‘heel.’’ In addition,there is often a clear impression of the hal-lux, with a large angle formed between thisimpression and that of digit II. Most Me-sozoic bird prints were made by waterbirds,and as in many modern waders the toe-marks are very slender.

However, some Mesozoic footprints,both large and small, are harder to catego-rize as having been made by ‘‘theropods’’or birds (Fig. 6). The Late Cretaceous ich-notaxon (footprint taxon) Saurexallopus(Harris et al., 1996; Harris, 1997) has a fair-

ly deep ‘‘heel’’ mark that is similar to themetatarsal pad of bird prints; the ichnotax-on is also characterized by the distinct im-pression of a long (albeit unreversed) hal-lux, large interdigital angles, and possiblyinterdigital webbing—yet more birdlikefeatures. If the Saurexallopus-maker was abird, it would have been comparable tolarge ratites in size. Harris et al. (1996) re-jected the hypothesis that the trackmakerwas a bird in part on the basis of the sizeof the prints, citing the absence of big birdsfrom the osteological record of the LateCretaceous, but the recent discovery of anostrich-sized bird from the Late Cretaceousof Europe (Buffetaut and Le Loeuff, 1998)weakens this objection. If, however, theSaurexallopus-maker was in fact a ‘‘thero-pod’’, it was a form whose feet were un-usually birdlike.

Magnoavipes from the Lower Cretaceousof Texas (Lee, 1997) is even more birdlikethan Saurexallopus, and like the latter wasmade by an animal comparable in size to alarge ratite. There seems to be a metatarsalimpression about which digits II and IV are


FIG. 6. Mesozoic tridactyl footprints of uncertain affinities. A) Topographic map of Saurexallopus, a large‘‘theropod’’ or bird print from the Late Cretaceous of Wyoming (Harris et al. 1996). The footprint is deepest inthe ‘‘heel’’ region, with a symmetrical metatarsal pad impression and a large digit I impression (extending tothe left in this view); print depth diminishes distally along the toemarks. B) Magnoavipes, a large bird (?)footprint from the Early Cretaceous of Texas; scale bar 5 10 cm. Redrawn from Lee (1997). C) Fuscinapedis,a large ‘‘theropod’’(?) footprint from the Early Cretaceous of Texas; note the apparently well developed meta-tarsal pad impression. Scale bar 5 10 cm. Redrawn from Lee (1997). D) Trisauropodiscus, a bird or birdlike‘‘theropod’’ footprint from the Early Jurassic of Africa. Scale bar 5 10 cm. Redrawn from Lockley et al. (1992).E) Plesiornis, a bird or birdlike ‘‘theropod’’ footprint from the Early Jurassic of Poland. Scale bar 5 1 cm.Redrawn from Gierlınski (1996).

symmetrically arranged, interdigital angleII–IV is rather large, and the toemarks areslender. Fuscinapedis from the same faunais even bigger, with footprints some 38 cmlong—comparable in size to large moafootprints. Lee (1997) attributed Fuscina-pedis to a ‘‘theropod,’’ but it has a markedmetatarsal impression.

Even more perplexing are the ichnoge-nera Trisauropodiscus and Plesiornis, verybirdlike footprints from the early Mesozoic(Lockley et al., 1992; Weems and Kimmel,1993; Gierlınski, 1996). Such prints showthe same features used to assign later Me-sozoic prints to avian trackmakers. Most ofthese tracks are only a few centimeterslong, but some are comparable in size toheron footprints. Intriguingly, these tracksare older than Archaeopteryx. Interpretingthem as having been made by birds wouldimply a rather earlier origin of birds than

conventionally accepted (Padian and Chiap-pe, 1998), but would be compatible withmore heterodox notions of when the firstbirds evolved (Chatterjee, 1991, 1997).

However, it is difficult to integrate infor-mation from footprints like Saurexallopus,Magnoavipes, and Trisauropodiscus intodiscussions of ‘‘theropod’’ and avian loco-motion, simply because we cannot be cer-tain of the affinities of their makers. Werethese creatures in fact birds, or rather the-ropods that had independently evolved avery birdlike pedal configuration? Suchquestions illustrate the weakness that ac-companies the strength that footprints bringto interpreting dinosaur locomotion: Whatone gains in the ability to ‘‘observe’’ thestance and motions of a trackmaker, oneloses in taxonomic resolution about whothat trackmaker was.


Trackway patterns and limb proportions

Trackways show that ‘‘theropods’’ werestriders rather than hoppers (Thulborn,1990). ‘‘Theropod’’ trackways are usually(but not always; Lockley et al., 1996) verynarrow, indicating that their makers gener-ally walked with the feet placed close to thebody midline. Footprints often angle slight-ly inward with respect to the trackmaker’sdirection of travel, but can also pointstraight ahead, or angle slightly outward(Farlow, 1987; Farlow and Galton, 2000).Tail marks are only rarely seen.

Trackway data suggest conservatism insome aspects of locomotor activity during‘‘theropod’’ evolution. Figure 7A examinesstride length as a function of footprintlength in trackways attributed to ‘‘thero-pods.’’ Unsurprisingly, bigger animals(with footprint length as a proxy for overallsize) tend to take longer strides than dosmaller animals (Thulborn, 1990; Perez-Lorente, 1996). The data plot in a broadband. Nearly all of the small later Mesozoic‘‘theropod’’ points at the upper edge of theband come from a single locality, the LarkQuarry site in Queensland, where a host ofsmall dinosaurs is thought to have beenpanicked by the approach of a large carniv-orous dinosaur (Thulborn and Wade, 1984).Other dinosaur points along the upper mar-gin of the band likewise show unusuallylong strides for a given footprint length.Unless one wishes to postulate the exis-tence of ‘‘theropods’’ with much longerlegs than any known from skeletal evi-dence, points along the top edge of the bandprobably represent trackways made by run-ning animals, and those along the loweredge of the band presumably correspond totrails of animals walking in a more leisurelyfashion.

The trackway data provide no evidencefor changes in stride lengths (at least ascompared with footprint length) duringwalking or running in comparisons of earlyas opposed to later Mesozoic ‘‘theropods.’’One might have expected progressivelymore cursorial groups (Holtz, 1995) of‘‘theropods’’ to have longer limbs for a giv-en foot length; progressively more cursorialanimals might reduce the amount of the

foot contacting the ground by increasing thedegree of digitigrady (cf. Coombs, 1978).This would then translate into longer stridelengths for a given footprint length.

The presently available trackway dataprovide no evidence for this. However,most of the later Mesozoic trackways are ofEarly Cretaceous age; there are few avail-able trail measurements from times andplaces when arctometatarsalians (predomi-nantly known from the Late Cretaceous ofAsia and western North America) wouldconstitute the likely trackmakers. If suchdata did exist, they might indicate relativelylonger strides during walking and runningthan in the data examined here.

On the other hand, although the data arescant, conservatism also seems to charac-terize the relationship between total leglength and a very rough proxy for footprintlength, the aggregate length of the phalan-ges of digit III. Ceratosaurs have a relative-ly short leg length for a given digit IIIlength, but there is otherwise no obviousphylogenetic pattern in this relationship(Fig. 4B). Derived groups seem not to haverelatively longer legs than more primitivegroups. This suggests the possibility thatrelative stride lengths did not lengthen ap-preciably—at least for walking animals—during the course of ‘‘theropod’’ evolution.A relatively longer stride may have beenused only during flat-out running (althoughthe trackway data provide no evidence forthis).

‘‘Theropods’’ may not differ muchamong themselves in the relationship be-tween stride and footprint length, but theydo differ from other kinds of bipedal archo-saurs. Figure 7B summarizes data for ‘‘the-ropods,’’ extant and extinct ground birds,and bipedal ornithischians. The manypoints for small ornithischians with longstrides are from the Lark Quarry site. Al-though large ornithischian (probably mainlyornithopod) points show considerable over-lap with those of ‘‘theropods,’’ there is aclear tendency for large ornithopods to takeshorter strides for a given footprint length.Bird points likewise show much overlapwith ‘‘theropods.’’ However, many of theavian trackmakers were zoo captives thatseemed agitated about being driven across


FIG. 7. Stride length (distance from one footprint to the next footprint made by the same foot) as a functionof footprint length. A) Early (Late Triassic-Early Jurassic) and later (Late Jurassic-Cretaceous) Mesozoic track-ways attributed to ‘‘theropods.’’ Data compiled from numerous published sources and the senior author’s un-published observations. B) Stride length as a function of footprint length in ‘‘theropods,’’ birds, and bipedalornithischians. Bird trackway data mainly from the senior author’s work on zoo and free-living animals, butalso from the literature. Although it is unlikely that any individual non-avian dinosaur is represented by morethan one point in the graph, many of the struthioniform points represent repeated trackmaking episodes by asmall number of individual emus.

substrates prepared for recording footprints;these birds were moving at faster than theirnormal walking speeds. Among these dis-contented birds were emus (Dromaius no-vaehollandiae) responsible for the longerstruthioniform strides; the same individualemus are also represented by lower-plotting

stride lengths in the graph. Note, too, thatthe dinornithiform trackways—which wereunquestionably made without human inter-ference—plot toward the bottom of thecloud of points associated with walking‘‘theropods.’’ It seems, then, that groundbirds, when moving without harassment,


FIG. 8. Leg length as a function of digit III length (a proxy for the portion of the foot likely to touch theground) in ‘‘dinosaurs’’ and birds. A) Total leg length (femur 1 tibia [or tibiotarsus] 1 metatarsal III [ortarsometatarsus]) vs. digit III length. B) ‘‘Functional’’ leg length vs. digit III length in ‘‘dinosaurs’’ and birds.In ‘‘dinosaurs’’ ‘‘functional’’ leg length is the same as total leg length, while in birds this parameter excludesthe length of the femur. Regression line for ‘‘theropods’’ in both graphs: log 10 leg length 5 1.029 log 10 digitIII length 1 0.670; r2 5 0.981. Some of the dinornithiform (moa) points represent composites of two individualbirds of comparable tarsometatarsus length.

often take relatively shorter steps for a giv-en footprint length than did non-avian the-ropods.

Conceivably this might be related to thedifferent hindlimb carriage of ‘‘theropods’’than birds, with the more vertical and rel-atively longer femur of the former resultingin a longer relative leg length. This hypoth-

esis can be tested by comparing the lengthof digit III against the combined lengths ofthe femur, tibia (or tibiotarsus), and meta-tarsus (or tarsometatarsus) in the twogroups (Fig. 8A). Contrary to the hypoth-esis, large ground birds (most living ratites,moa, elephantbirds, bustards, seriemas)tend to have a longer leg, relative to digit


III length, than do ‘‘theropods.’’ However,Mesozoic and some Cenozoic birds (button-quail, kiwi, a lithornithid, Aenigmavis, Mes-selornis, Foro, galliforms) do plot amongpoints for ‘‘theropods.’’ Large ornithopods,like big birds, have long legs relative to thelength of digit III.

If we now consider that most of the legaction in walking birds occurs at the kneerather than the hip (Gatesy, 1990, 1991; Ga-tesy and Biewener, 1991), we can comparea ‘‘functional’’ leg length against digit IIIlength in ‘‘dinosaurs’’ and birds (Fig. 8B),with ‘‘functional’’ leg length being equal tototal leg length in ‘‘dinosaurs,’’ but equiv-alent to the combined length of the tibi-otarsus and tarsometatarsus in birds. Birdsthat plotted among ‘‘theropod’’ points in thegraph of total leg length against digit IIIlength now plot below them, but the long-legged ground birds now show considerableoverlap with ‘‘theropods.’’

Neither total leg length nor ‘‘functional’’leg length as presented here is completelyrealistic; we have not considered angles be-tween limb segments or the actual motionsmade by the limbs during locomotion.Nonetheless the comparisons are probablyrealistic enough to have intriguing impli-cations for ‘‘dinosaurian’’ and avian loco-motion. First, ‘‘theropods’’ do not have rel-atively longer legs (relative to digit III andfootprint length) than birds. If we could addthe likely lengths of ‘‘theropod’’ tarsal el-ements in leg lengths, ‘‘functional’’ limblengths of ‘‘theropods’’ would likely showeven more overlap with birds in the rela-tionship depicted in Figure 8. This meansthat ‘‘theropods’’ and ground birds mayhave roughly the same functional leg lengthfor the amount of foot that touches theground.

Second, this similarity in function, ifsuch it is, is not something that was con-served during changes in limb carriage inthe evolution of birds from ‘‘theropods.’’Mesozoic and some Cenozoic birds haverelatively shorter ‘‘functional’’ limb lengthsthan ‘‘theropods’’ (Fig. 8B). Consequentlyany similarity in ‘‘functional’’ leg lengthrelative to digit III length between largeground birds and ‘‘theropods’’ reflects con-vergence rather than conservatism.

Large ornithopods have relatively longerlegs for a given digit III length than dolarge theropods (Fig. 8). This may highlighta weakness in using digit III length or foot-print length as a proxy for overall animalbody size. If we compare digit III and leglengths with overall body shape in large‘‘theropods’’ and big ornithopods, it is clearthat the latter do not have long legs so muchas extremely short toes. A better compari-son of leg lengths among ‘‘theropods,’’birds, and ornithischians might use the totallength of the dorsal vertebral series as theproxy for body size.

If birds and ‘‘theropods’’ of a given bodysize generally walk(ed) at the same speed,possibly birds have a higher stride frequen-cy than ‘‘theropods’’ did. Alternatively, ifthe longer step lengths of ‘‘theropods’’ thanbirds do in fact translate into faster speeds,perhaps this difference in locomotion be-tween ‘‘theropods’’ and ground birds has abehavioral rather than a structural explana-tion. Most Cenozoic ground birds havebeen herbivores or omnivores rather thancarnivores, while most ‘‘theropods’’ weremanifestly predators. Conceivably track-ways of ‘‘theropods’’ were generally madeduring foraging bouts, when the trackmak-ers were actively searching for prey. In con-trast, many trackways made by groundbirds may record less purposeful motion.The more abundant, better dispersed natureof foodstuffs available to non-predatory an-imals may permit short, unhurried stepsduring many foraging bouts. This idea is,however, merely speculation.

‘‘Theropod’’ trackways and running

In a study based on 267 trackways of bi-pedal dinosaurs, Thulborn (1984b) exam-ined the frequency distribution of the ratioof stride length to estimated hip height. Un-surprisingly, he found that dinosaur track-ways usually had a low stride/hip height ra-tio, indicating that walking predominatedover faster styles of locomotion. More in-terestingly, Thulborn’s data suggested thatthe distribution of stride/hip height ratiowas bimodal. He speculated that what hecalled a ‘‘trot’’ (which he defined as slowrunning in which the ratio of stride lengthto estimated hip height ranged from 2.0 to


FIG. 9. Relative stride length (stride/footprint length) in trackways of bipedal dinosaurs. Unshaded histogramis for all trackways (N 5 664), including many from the Lark Quarry site, where a large number of smalldinosaurs apparently were frightened into mass flight by the approach of a large ‘‘theropod.’’ Shaded histogram(N 5 577) excludes trackways from the Lark Quarry site.

2.9) was a transitional gait of high energeticcost that bipedal dinosaurs generally avoid-ed using.

However, 92 of Thulborn’s dinosaurtrackways came from his Lark Quarry site,which—as already has been noted—repre-sent an unusual circumstance, where a largegroup of small dinosaurs was stampeded bythe approach of a large carnivore (Thulbornand Wade, 1984). Figure 9 plots the stride/footprint length ratio for 664 bipedal dino-saurs, more than twice Thulborn’s (1984b)sample size. There is no indication of bi-modality in this larger data set. This con-clusion is even stronger if data for the LarkQuarry site are excluded; if bipedal dino-saurs really had avoided a ‘‘trot,’’ thisshould be apparent in trackway data gen-erally. The bimodality in Thulborn’s datawas an artifact of the large number of track-ways from a single, unusual site (a possi-bility that Thulborn conceded). Bipedal di-nosaurs usually walked, and used increas-ingly faster gaits decreasingly often.

The occurrence of ‘‘Groucho’’ running inbirds raises other questions about the gaitsused by non-avian theropods. If this styleof compliant locomotion actually evolvedin pre-avian theropods, then some ‘‘thero-

pod’’ trackways with relatively long stridesmay not record true running. However, be-cause some of the characteristic features as-sociated with avian bipedalism (knee-basedkinematics, horizontal femur, presence of apygostyle) may not have been present inbasal birds (Gatesy, 1990, 1991; Carrano,1998), then—if ‘‘Groucho’’ running is in-timately tied to a horizontal femoral orien-tation—it is likely that most ‘‘theropods’’did not employ this style of fast locomo-tion. If any ‘‘theropods’’ were ‘‘Grouchos,’’we suspect that they were small forms veryclosely related to birds.

Alexander (1976) published an equationthat related an animal’s speed to its hipheight and stride length. Because the lattertwo parameters can be either measured orestimated from trackways, Alexander’sequation became the basis for estimatingthe speeds of dinosaurs. Most such esti-mates for bipedal dinosaurs are 10 km/hr orless (Thulborn, 1990; Perez-Lorente, 1996),but estimated speeds of as much as 40 km/hr have been reported (Farlow, 1981; Vieraand Torres, 1995; Irby, 1996). Althoughthese estimates seem reasonable, there is noway of testing their accuracy.


FIG. 10. One of many possible interpretations ofhindlimb joint angulation in Tyrannosaurus that areconsistent with a three-dimensional analysis of jointconfigurations.

Stance, Running, and Gigantism in‘‘Theropods’’

Gigantism was a recurring theme in di-nosaurian evolution (Farlow et al., 1995;Sereno, 1997). The earliest theropods weresmall to medium-sized (1–3 m body length)animals, but tetanurines soon evolved verylarge forms.

Biomechanical, scaling, and kinematicstudies of living tetrapods suggest that ter-restrial vertebrates adopt more extendedhindlimb posture at larger body sizes ascompared with smaller tetrapods (Biewener,1989; Bertram and Biewener, 1990; Gatesyand Biewener, 1991). If the pattern exhib-ited by mammals and birds has general ap-plication, then larger theropods should havemore straightened hindlimbs than theirsmaller relatives.

In contrast, Paul (1988) contended that a‘‘crouched’’ hindlimb stance was retainedeven in enormous ‘‘theropods’’ like Tyran-nosaurus, basing his argument on the con-figuration of the hip, knee, and ankle joints.From this he concluded that Tyrannosaurusretained the same capacity for rapid loco-motion seen in smaller, related forms likeornithomimosaurs.

However, a 3D computer analysis usingdigitized elements of the hindlimb of Ty-rannosaurus (J. R. Hutchinson, unpub-lished observations) reveals more joint mo-bility at the knee and hip than is evidentfrom 2D drawings of the kind made byPaul. The hip and knee joints can be posi-tioned in a variety of intuitively pleasingpositions (Fig. 10) that span the entire‘‘crouched vs. columnar’’ dichotomy with-out disarticulating these joints.

Christiansen (1999) estimated limb an-gulations of ‘‘theropods’’ of various sizesby measuring the angle between the femo-ral distal epiphysis and the long axis of thebone. In contrast to Paul (1988), his datasuggested that large ‘‘theropods’’ did in-deed have less flexed limbs than theirsmaller relatives, but nonetheless had lesscolumnar limbs than elephants, suggestingsomewhat greater potential for fast loco-motion than in the latter.

Alexander (1985) defined an index of thebending strength of limb bones that consid-

ered animal mass, bone length, and the sec-tion modulus of the shaft of the bone. Mod-ern large animals capable of rapid loco-motion have high values of this ‘‘strengthindicator,’’ while slowly moving formshave low values. Applied to dinosaurs, Al-exander’s approach indicated that the femurof Tyrannosaurus was comparable in bend-ing strength to that of an elephant, sug-gesting that the huge theropod was littlebetter than an elephant at moving quickly.Farlow et al. (1995) recalculated Alexan-der’s strength indicator for Tyrannosaurus,using better data than were available to Al-exander, but nonetheless found values ofthe strength indicator comparable to thoseobtained by Alexander. Christiansen (1999)noted that a component of the strength in-dicator that lessens its value, limb bonelength, is disproportionately longer in large‘‘theropods’’ than in large mammals capa-ble of rapid locomotion. He concluded thatthe limbs of enormous ‘‘theropods’’ wereonly marginally stronger than those of ele-phants and hippos.

Farlow et al. (1995) considered anotherissue that could have affected the locomotor


FIG. 11. Injury threshold impact acceleration a2 as a function of animal body mass M2.

performance of very large ‘‘theropods’’: thelikelihood of injury in the event of a fallwhile running. Simple mathematical modelsbased on animal mass, the height of the di-nosaur’s head and torso above the ground,the depth of the hole gouged in the groundby the animal’s fall, and the animal’sground speed suggested that the risk of se-rious injury or even death would have beensignificant had a Tyrannosaurus-sized ‘‘the-ropod’’ stumbled while moving at speed.Farlow et al. (1995) concluded that this wasan additional reason for thinking that verylarge ‘‘theropods’’ did not run at sprintspeeds comparable to those of smaller ‘‘the-ropods.’’

Alexander (1996) evaluated this argu-ment using a novel source of data: injuriessustained by humans in car crashes. Fromscaling arguments he calculated that, if a70-kg human can survive an impact decel-eration of 15g, a 6000-kg tyrannosaurshould be able to withstand a decelerationof 3.4g. More than that and the animalwould likely be seriously injured. The mod-els presented by Farlow et al. (1995) esti-mated impact decelerations considerablygreater than 3.4g, and Alexander (1996)concluded that a falling Tyrannosaurus

would indeed very likely risk damage to it-self.

We have extended Alexander’s (1996) ar-gument (appendix 2) to generalize the re-lationship between injury threshold accel-eration and body mass (Fig. 11).The max-imum tolerable impact deceleration a2 de-creases drastically for animals with bodymass greater than 70 kg (the mass of anaverage-sized human being). This alsomeans, though, that ‘‘theropods’’ smallerthan Tyrannosaurus would have risked lessinjury in the event of a fall: a2 for a 2000-kg animal is 4.9g; for a 1500-kg animal5.4g; for a 500-kg animal 7.8g; for a 200-kg animal 10.6g. To the extent that injuryrisk was a factor in limiting locomotor per-formance, theropods like Gorgosauruscould be expected to have been somewhatmore likely than Tyrannosaurus to engagein ‘‘Jurassic Park’’-style performances,while ornithomimosaurs could safely haverun at the higher speeds that their mor-phology suggests (Russell, 1972). Further-more, juveniles of very large-bodied the-ropods could more safely have been rapidrunners than their elders were.

However, Alexander (1996, p. 121) ques-tioned ‘‘whether animals can be expected to


be cautious, or to live dangerously.’’ He cit-ed the galloping of giraffes, the sprinting ofostriches, and the treetop maneuvers of gib-bons as activities in which a fall might beexpected to result in injury. Whether andhow natural selection factors such risks be-fore rolling the locomotory dice is a ques-tion we cannot answer. However, it doesseem likely that smaller animals like gib-bons and ostriches would be better able tomake sudden, quick corrections of motionto correct for a potential mishap than couldan elephant-sized animal moving at a rapidspeed. Furthermore, Farlow et al. (1996) ar-gued that a huge biped might find it harderto compensate for a misstep than could alarge quadruped like a giraffe.

The trackway of a Tyrannosaurus-sizedtheropod with an enormous stride woulddecisively show that such animals were in-deed capable of rapid sprinting. Thus far,however, the only known ‘‘theropod’’ track-ways that suggest flat-out running weremade by small to medium-sized forms(Thulborn, 1990; Irby, 1996). As noted byPaul (1988), this is merely negative evi-dence. However, as fossilized trails of largetheropods continue to be discovered, ifnone of them suggests rapid locomotion,this negative evidence will become increas-ingly hard to ignore.

CONCLUDING REMARKS

Birds may be theropods, but ‘‘thero-pods’’ were not birds, and ‘‘the evolutionof theropods was more than a ‘bird facto-ry’’’ (Holtz, 1998, p. 1276). For more than150 million years a diversity of large andsmall ‘‘theropods’’ constituted the dominantlarge-vertebrate predators of continental bi-otas. Throughout that time they showedconsiderable variability in adaptations forprocuring and processing animal prey, withremarkable specializations of the jaws andhands. At the same time, however, theirhindlimb features remained fairly conser-vative. Early in their history, theropods hitupon a successful style of striding, digiti-grade bipedalism that they employed for theremainder of their successful reign.

Whether and how the origin of feathersrelates to theropod locomotion depends onwhen in avian evolution these peculiar

structures first appeared, a matter that re-mains contentious (Chen et al., 1998; Ji etal., 1998; Unwin, 1998; Ruben and Jones,2000; L. Witmer, personal communication).If feathers or ‘‘protofeathers’’ arose in non-volant, non-avian theropods, their initalfunction could not have been related toflight (Padian, 1998), and neither wouldthere be any direct connection betweenfeather origins and the typical striding bi-pedalism of ‘‘theropods.’’ On the otherhand, bipedal locomotion freed theropodforelimbs from an obligatory connectionwith ground locomotion (Gatesy and Dial,1996). The forelimb was therefore availableto be pressed into service for an entirelynew style of locomotion, whenever it wasthat feathers appeared or were elaborated ina fashion that made flight possible.

Thus one lineage of extremely special-ized theropods, although they retainedmany of the features of theropod bipedallocomotion, nevertheless considerablymodified the forelimb for flight. For thesetheropods, which survived whatever it wasthat eliminated their more conservative,ground-dwelling kin, the sky was literallythe limit of their evolutionary potential.

ACKNOWLEDGMENTS

We thank Paul Maderson and DominiqueHomberger for inviting us to participate inthis symposium, and the National ScienceFoundation for travel grants to bring partic-ipants to the meeting. Matt Carrano, KevinMiddleton and two anonymous reviewersmade helpful comments on the manuscript.Peggy Maceo, Kirk Johnson, Kieran Shep-herd, Tony Fiorillo, and Raymond Cooryprovided us with useful footprint casts; JanaMcClain and personnel of the Fort Wayneand other zoos helped in collecting printsof modern birds. Jim Hopson shared mea-surements of bird hindlimbs, and Jim Whit-craft produced some of our artwork. Thisresearch was supported by grants from theDinosaur Society and the National ScienceFoundation to Farlow. This is University ofCalifornia Museum of Paleontology contri-bution number 1723.

REFERENCES

Alexander, R. McN. 1976. Estimates of speeds of di-nosaurs. Nature 261:129–130.


Alexander, R. McN. 1985. Mechanics of posture andgait of some large dinosaurs. Zool. J. Linn. Soc.83:1–25.

Alexander, R. McN. 1996. Tyrannosaurus on the run.Nature 379:121.

Baird, D. 1957. Triassic reptile footprint faunules fromMilford, New Jersey. Bull. Mus. Comp. Zool., Ha-vard Univ. 117:449–520.

Bertram J. E. A. and A. A. Biewener. 1990. Differ-ential scaling of the long bones in the terrestrialCarnivora and other mammals. J. Morph. 204:157–169.

Biewener, A. A. 1989. Scaling body support in mam-mals: limb posture and muscle mechanics. Science245:45–48.

Bonaparte, J. F. and J. A. Pumares. 1995. Notas sobreel primer craneo de Riojasaurus incertus (Dino-sauria, Prosauropoda, Melanorosauridae) delTriasico Superior de La Rioja, Argentina. Amegh-iniana 32:341–349.

Buffetaut, E. and J. Le Loeuff. 1998. A new giantground bird from the Upper Cretaceous of France.J. Geol. Soc. Lond. 155:1–4.

Burke, A. C. and A. Feduccia. 1997. Developmentalpatterns and the identification of homologies inthe avian hand. Science 278:666–668.

Cannon, R. L. 1996. Anatomical studies of the ratitepelvic limb. Master’s thesis, Johns Hopkins Uni-versity, Baltimore.

Carrano, M. T. 1998. Locomotion in non-avian dino-saurs: Integrating data from hindlimb kinematics,in vivo strains, and bone morphology. Paleobiol-ogy 24:450–469.

Carrano, M. T. 1999. What, if anything is a cursor?Categories vs. continua for determining locomotorhabit in mammals and dinosaurs. J. Zool., Lond.247:29–42.

Carrano, M. T. and C. A. Sidor. 1999. Theropod hindlimb disparity revisited: Comments on Gatesy andMiddleton (1997). J. Vert. Paleontol. 19:602–605.

Charig, A. J. and A. C. Milner. 1997. Baryonyx walk-eri, a fish-eating dinosaur from the Wealden ofSurrey. Bull. Nat. Hist. Mus. Lond. (Geol.) 53:11–70.

Chatterjee, S. 1991. Cranial anatomy and relationshipsof a new Triassic bird from Texas. Phil. Trans.Roy. Soc. Ser. B 332:277–346.

Chatterjee, S. 1997. The rise of birds: 225 millionyears of evolution. Johns Hopkins UniversityPress, Baltimore.

Chen, P.-J., Z.-M. Dong, and S.-N. Zhen. 1998. Anexceptionally well-preserved theropod dinosaurfrom the Yixian Formation of China. Nature 391:147–152.

Chiappe, L. M. 1996. Late Cretaceous birds of south-ern South America: Anatomy and systematics ofPatagopteryx and Enantiornithes. In G. Arratia(ed.), Contributions of southern South America tovertebrate paleontology, pp. 203–244. MunchnerGeowiss. Abh. a(3), F. Pfeil, Munich.

Christiansen, P. 1999. Long bone scaling and limb pos-ture in non-avian theropods: Evidence for differ-ential allometry. J. Vert. Paleo. 19:666–680.

Coombs, W. P., Jr. 1978. Theoretical aspects of cur-

sorial adaptations in dinosaurs. Quart. Rev. Biol.53:393–418.

Coombs, W. P., Jr. 1980. Swimming ability of carniv-orous dinosaurs. Science 207:1198–1200.

Farlow, J. O. 1981. Estimates of dinosaur speeds froma new trackway site in Texas. Nature 294:747–748.

Farlow, J. O. 1987. Lower Cretaceous dinosaur tracks,Paluxy River Valley, Texas. South Central Sec-tion, Geol. Soc. Amer., Waco, TX.

Farlow, J. O., P. Dodson, and A. Chinsamy. 1995. Di-nosaur biology. Ann. Rev. Ecol. Syst. 26:445–471.

Farlow, J. O. and P. M. Galton. 2000. Dinosaur track-ways of Dinosaur State Park, Rocky Hill, Con-necticut. In P. M. LeTourneau and P. E. Olsen(eds.), Aspects of Triassic-Jurassic rift basin geo-science, Columbia University Press, New York.(In press).

Farlow, J. O., M. B. Smith, and J. M. Robinson. 1995.Body mass, bone ‘‘strength indicator,’’ and cur-sorial potential of Tyrannosaurus rex. J. Vert. Pa-leont. 15:713–725.

Feduccia, A. 1996. The origin and evolution of birds.Yale University Press, New Haven, CT.

Feduccia, A. 1999. 1,2,3 5 2,3,4: Accomodating thecladogram. Proc. Natl. Acad. Sci. U.S.A. 96:4740–4742.

Forster, C. A., S. D. Sampson, L. M. Chiappe, and D.W. Krause. 1998. The theropod ancestry of birds:New evidence from the Late Cretaceous of Mad-agascar. Science 279:1915–1919.

Garland, T., Jr. and C. M. Janis. 1993. Does metatarsal/femur ratio predict maximal running speed in cur-sorial mammals? J. Zool., Lond. 229:133–151.

Gatesy, S. M. 1990. Caudofemoral musculature andthe evolution of theropod locomotion. Paleobiol-ogy. 16:170–186.

Gatesy, S. M. 1991. Hind limb scaling in birds andother theropods: Implications for terrestrial loco-motion. J. Morph. 209:83–96.

Gatesy, S. M. 1995. Functional evolution of the hindlimb and tail from basal theropods to birds. In J.J. Thomason (ed.), Functional morphology in ver-tebrate paleontology, pp. 219–234. CambridgeUniv. Press, Cambridge, U.K.

Gatesy, S. M. 1999. Guineafowl hindlimb function I:Cineradiographic analysis and speed effects. J.Morph. 240:127–142.

Gatesy, S. M. and A. A. Biewener. 1991. Bipedal lo-comotion: Effects of speed, size and limb posturein birds and humans. J. Zool., Lond. 224:127–147.

Gatesy, S. M. and K. P. Dial. 1996. Locomotor mod-ules and the evolution of avian flight. Evolution50:331–340.

Gatesy, S. M. and K. M. Middleton. 1997. Bipedalism,flight, and the evolution of theropod locomotordiversity. J. Vert. Paleont. 17:308–329.

Gatesy, S. M., K. M. Middleton, F. A. Jenkins, Jr.,andN. H. Shubin. 1999. Three-dimensional preser-vation of foot movements in Triassic theropod di-nosaurs. Nature 399:141–144.

Gauthier, J. 1986. Saurischian monophyly and the or-igin of birds. In K. Padian (ed.), The origin of


birds and the evolution of flight, pp. 1–55. Mem.Calif. Acad. Sci. 8.

Gierlınski, G. 1996. Avialian theropod tracks from theEarly Jurassic strata of Poland. Zubıa 14:79–87.

Harris, J. D. 1997. Four-toed theropod footprints anda paleomagnetic age from the Whetstone FallsMember of the Harebell Formation (Upper Cre-taceous: Maastrichtian), northwestern Wyoming:A correction. Cret. Res. 18:139.

Harris, J. D. 1998a. A reanalysis of Acrocanthosaurusatokensis, its phylogenetic status, and paleobi-ogeographic implications, based on a new speci-men from Texas. New Mexico Mus. Nat. Hist. Sci.Bull. 13, Albuquerque.

Harris, J. D. 1998b. Dinosaur footprints from GardenPark, Colorado. Modern Geol. 23:291–307.

Harris, J. D., K. R. Johnson, J. Hicks, and L. Taux.1996. Four-toed theropod footprints and a paleo-magnetic age from the Whetstone Falls Memberof the Harebell Formation (Upper Cretaceous:Maastrichtian), northwestern Wyoming. Cret. Res.17:381–401.

Holtz, T. R., Jr. 1994. The phylogenetic position of theTyrannosauridae: Implications for theropod sys-tematics. J. Paleont. 68:1100–1117.

Holtz, T. R., Jr. 1995. The arctometatarsalian pes, anunusual structure of the metatarsus of CretaceousTheropoda (Dinosauria: Saurischia). J. Vert. Pa-leont. 14:480–519.

Holtz, T. R., Jr. 1996. Phylogenetic taxonomy of theCoelurosauria (Dinosauria: Theropoda). J. Pa-leont. 70:536–538.

Holtz, T. R., Jr. 1998. Spinosaurs as crocodile mimics.Science 282:1276–1277.

Holtz, T. R., Jr. 2000. A new phylogeny of the carniv-orous dinosaurs. Gaia. (In press).

Irby, G. V. 1996. Paleoichnological evidence for run-ning dinosaurs worldwide. In M. Morales (ed.),The continental Jurassic, pp. 109–112. Mus. N.Ariz. Bull. 60, Flagstaff, AZ.

Janis, C. M. and P. B. Wilhelm. 1993. Were theremammalian pursuit predators in the Tertiary?Dances with wolf avatars. J. Mamm. Evol. 1:103–125.

Ji, Q., P. J. Currie, M. A. Norell, and S.-A. Ji. 1998.Two feathered dinosaurs from northeastern China.Nature 393:753–761.

Kuban, G. J. 1989. Elongate dinosaur tracks. In D. D.Gillette and M. G. Lockley (eds.), Dinosaur tracksand traces, pp. 57–72. Cambridge Univ. Press,Cambridge, U.K.

Lee, Y.-N. 1997. Bird and dinosaur footprints in theWoodbine Formation (Cenomanian), Texas. Cret.Res. 18:849–864.

Lockley, M. G. and A. P. Hunt. 1994. A track of thegiant theropod dinosaur Tyrannosaurus from closeto the Cretaceous/Tertiary boundary, northernNew Mexico. Ichnos 3:213–218.

Lockley, M. G., C. A. Meyer, and V. F. dos Santos.1996. Megalosauripus, Megalosauropus, and theconcept of megalosaur footprints. In M. Morales(ed.), The continental Jurassic, pp. 113–118. Mus.N. Ariz. Bull. 60, Flagstaff, AZ.

Lockley, M. G., S. Y. Yang, M. Matsukawa, F. Flem-

ing, and S. K. Lim. 1992. The track record ofMesozoic birds: Evidence and implications. Phil.Trans. Roy. Soc. Lond. B 336:113–134.

Lucas, A. M. and P. R. Stettenheim. 1972. Avian anat-omy: Integument. Part I. Agriculture Handbook362, U.S. Dept. Agriculture, Washington, D.C.,340 pp.

Lull, R. S. 1953. Triassic life of the Connecticut Val-ley. Connecticut State Geol. Nat. Hist. Surv. Bull.81, Hartford, CT.

Madsen, J. H. 1993. Allosaurus fragilis: A revised os-teology. Utah Geol. Surv. 109:1–163.

Makovicky, P. J. and H.-D. Sues. 1998. Anatomy andphylogenetic relationships of the theropod dino-saur Microvenator celer from the Lower Creta-ceous of Montana. Amer. Mus. Novitates 3240:1–27.

Martin, L. D. 1998. Felidae. In C. M. Janis, K. M.Scott, and L. L. Jacobs (eds.), Evolution of Ter-tiary mammals of North America, pp. 236–242.Cambridge Univ. Press, Cambridge, U.K.

McGowan, C. 1979. The hind limb musculature of thebrown kiwi, Apteryx australis mantelli. J. Morph.160:33–74.

McMahon, T. A., G. Valiant, and E. C. Frederick.1987. Groucho running. J. Appl. Physiol. 62:2326–2337.

Muller, G. B. and J. Streicher. 1989. Ontogeny of thesyndesmosis tibiofibularis and the evolution of thebird hindlimb: A caenogenetic feature triggersphenotypic novelty. Anat. Embryol. 179:327–339.

Novas, F. E. 1993. New information on the systematicsand postcranial skeleton of Herrerasaurus ischig-ualastensis (Theropoda: Herrerasauridae) from theIschigualasto Formation (Upper Triassic) of Ar-gentina. J. Vert. Paleont. 13:400–423.

Novas, F. E. 1996. Dinosaur monophyly. J. Vert. Pa-leont. 16:723–741.

Olsen, P. E. 1995. Paleontology and paleoenviron-ments of Early Jurassic age strata in the WalterKidde Dinosaur Park (New Jersey, USA). In J. E.B. Baker (ed.), Contributions to the paleontologyof New Jersey, pp. 156–190. Geol. Assoc. N. Jer-sey, Wayne, N.J.

Ostrom, J. H. 1969. Osteology of Deinonychus antir-rhopus, an unusual theropod from the Lower Cre-taceous of Montana. Peabody Mus. Nat. Hist.Bull. 30:1–165.

Ostrom, J. H. 1975. The origin of birds. Ann. Rev.Earth Planet. Sci. 3:55–77.

Ostrom, J. H. 1976. Archaeopteryx and the origin ofbirds. Biol. J. Linn. Soc. 8:81–182.

Padian, K. 1998. When is a bird not a bird? Nature393:729–730.

Padian, K. and L. M. Chiappe. 1998. The origin andearly evolution of birds. Biol. Rev. 73:1–42.

Padian, K., J. R. Hutchinson, and T. R. Holtz, Jr. 1999.Phylogenetic definitions and nomenclature of themajor taxonomic categories of the carnivorous Di-nosauria (Theropoda). J. Vert. Paleo. 19:69–80.

Padian, K. and P. E. Olsen. 1989. Ratite footprints andthe stance and gait of Mesozoic theropods. In D.D. Gillette and M. G. Lockley (eds.), Dinosaur


tracks and traces, pp. 231–241. Cambridge Univ.Press, Cambridge, U.K.

Paul, G. S. 1984. The archosaurs: A phylogeneticstudy. In W.-E. Reif and F. Westphal (eds.), Thirdsymposium on Mesozoic terrestrial ecosystems,Short Papers, pp. 175–180. Atempto Verlag, Tu-bingen.

Paul, G. S. 1988. Predatory dinosaurs of the world: Acomplete illustrated guide. Simon and Schuster,New York.

Perez-Lorente, F. 1993. Dinosaurios plantıgrados en LaRioja. Zubıa 5:189–228.

Perez-Lorente, F. 1996. Pistas teropodas en cifras. Zu-bıa 14:37–55.

Perez-Moreno, B. P., J. L. Sanz, J. Sudre, and B. Sige.1993. A theropod dinosaur from the Lower Cre-taceous of southern France. Rev. Paleobiol. Vol.Spec. 7:173–188.

Reilly, S. M. 1995. Quantitative electromyography andmuscle function of the hind limb during quadru-pedal running in the lizard Sceloporus clarkii. Zo-ology 98:263–277.

Reilly, S. M. 1998. Sprawling locomotion in the lizardSceloporus clarkii: Speed modulation of motorpatterns in a walking trot. Brain Behav. Evol. 52:126–138.

Rowe, T. and J. Gauthier. 1990. Ceratosauria. In D. B.Weishampel, P. Dodson, and H. Osmolska (eds.),The Dinosauria, pp. 151–168. Univ. CaliforniaPress, Berkeley, C.A.

Ruben, J. A., T. D. Jones, N. R. Geist, and W. J. Hil-lenius. 1997. Lung structure and ventilation in the-ropod dinosaurs and early birds. Science 278:1267–1270.

Ruben, J. A. and T. D. Jones. 2000. Selective factorsin the origin of fur and feathers. Amer. Zool. 40:585–596.

Russell, D. A. 1972. Ostrich dinosaurs from the LateCretaceous of western Canada. Can. J. Earth Sci.9:375–402.

Russell, D. A. and Z.-M. Dong. 1993. The affinities ofa new theropod from the Alxa Desert, Inner Mon-golia, People’s Republic of China. Can. J. EarthSci. 30:2107–2127.

Sereno, P. C. 1991. Basal archosaurs: Phylogenetic re-lationships and functional implications. J. Vert.Paleon. 11(Supplement to No. 4):1–53.

Sereno, P. C. 1997. The origin and evolution of dino-saurs. Ann. Rev. Earth Planet. Sci. 25:435–489.

Sereno, P. C. 1999. A rationale for dinosaurian tax-onomy. J. Vert. Paleo. 19:788–790.

Sereno, P. C., A. L. Beck, D. B. Duthiel, B. Gado, H.C. E. Larsson, G. H. Lyon, J. D. Marcot, O. W.M. Rauhut, R. W. Sadleir, C. A Sidor, D. D. (sic)Varrichio, G. P. Wilson, and J. A. Wilson. 1998.A long-snouted predatory dinosaur from Africaand the evolution of spinosaurids. Science 282:1298–1302.

Sereno, P. C., D. B. Dutheil, M. Iarochene, H. C. E.Larsson, G. H. Lyon, P. M. Magwene, C. A. Sidor,D. J. Varrichio, and J. A. Wilson. 1996. Predatorydinosaurs from the Sahara and Late Cretaceousfaunal differentiation. Science 272:986–991.

Sereno, P. C. and F. E. Novas. 1992. The complete

skull and skeleton of an early dinosaur. Science258:1137–1140.

Sereno, P. C. and F. E. Novas. 1993. The skull andneck of the basal theropod Herrerasaurus ischig-ualastensis. J. Vert. Paleont. 13:451–476.

Sereno, P. C., J. A. Wilson, H. C. E. Larsson, D. B.Dutheil, and H.-D. Sues. 1994. Early Cretaceousdinosaurs from the Sahara. Science 266:267–271.

Steudel, K. and J. Beattie. 1995. Does limb length pre-dict the relative energetic cost of locomotion inmammals? J. Zool., Lond. 235:501–514.

Sues, H.-D. 1997. On Chirostenotes, a Late Cretaceousoviraptorosaur (Dinosauria: Theropoda) fromwestern North America. J. Vert. Paleont. 17:698–716.

Tarsitano, S. 1983. Stance and gait in theropod dino-saurs. Acta Palaeont. Polonica 28:251–264.

Thulborn, R. A. 1984a. The avian relationships of Ar-chaeopteryx, and the origin of birds. Zool. J. Linn.Soc. 82:119–158.

Thulborn, R. A. 1984b. Preferred gaits of bipedal di-nosaurs. Alcheringa 8:243–252.

Thulborn, R. A. 1994. Ornithopod dinosaur tracksfrom the Lower Jurassic of Queensland. Alcher-inga 18:247–258.

Thulborn, T. 1990. Dinosaur tracks. Chapman andHall, London.

Thulborn, T. and M. Wade. 1984. Dinosaur trackwaysin the Winton Formation (Mid-Cretaceous) ofQueensland. Mem. Queensland Mus. 21:413–517.

Unwin, D. M. 1998. Feathers, filaments and theropoddinosaurs. Nature 391:119–120.

Viera, L. I. and J. A. Torres. 1995. Analisis compara-tivo sobre dos rastros de Dinosaurios Theropodos:Forma de marcha y velocidad. Munibe 47:53–56.

Wagner, G. P. and J. A. Gauthier. 1999. 1,2,3 5 2,3,4:A solution to the problem of the homology of thedigits in the avian hand. Proc. Natl. Acad. Sci.U.S.A. 96:5111–5116.

Weems, R. E. and P. G. Kimmel. 1993. Upper Triassicreptile footprints and coelacanth fish scales fromthe Culpeper Basin, Virginia. Proc. Biol. Soc.Wash. 106:390–401.

Welman, J. 1995. Euparkeria and the origin of birds.S. Afr. J. Sci. 91:533–537.

Wright, J. L. 1996. Fossil terrestrial trackways: Func-tion, taphonomy, and palaeoecological signifi-cance. Ph.D. thesis, University of Bristol.

APPENDIX 1

A list of the genera that represent cladesand informal groups in Figures 4, 7, and8 of this study. A given clade or informalgroup is not necessarily represented bythe same genera in all graphs.

Basal theropods 5 Eoraptor and Herrerasaurus.Ceratosaurs 5 Segisaurus, Procompsognathus, Dilo-

phosaurus, Coelophysis, Syntarsus, Elaphrosaurus,Podokesaurus, and Ceratosaurus.

‘Megalosaurs’ 5 Afrovenator and Eustreptospondylus.Carnosaurs 5 Sinraptor, Acrocanthosaurus, Allosau-


rus, Chilantaisaurus, Saurophaganx, and Szechu-anosaurus.

Basal coelurosaurs 5 Compsognathus, Chuandongo-coelurus, Deltadromeus, Dryptosaurus, Kiajiango-saurus, Ornitholestes, and Tugulusaurus.

Dromaeosaurids 5 Deinonychus and Saurornitholes-tes.

Tyrannosaurids 5 Maleevosaurus, Aublysodon, Gor-gosaurus, Albertosaurus, Daspletosaurus, and Ty-rannosaurus.

Ornithomimosaurs 5 Ornithomimus, Struthiomimus,Dromiceiomimus, Gallimimus, Garudimimus, andArchaeornithomimus.

Alvarezsaurids 5 Avimimus (questionably) and Parv-icursor.

Oviraptorids 5 Chirostenotes, Caudipteryx, and Ingen-ia.

Troodontids 5 Sinornithoides.Struthioniformes 5 Dromaius and Rhea.Large ratites 5 Struthio, Rhea, Casuarius, Dromaius,

Dinornis, Emeus, Euryapteryx, Megalapteryx, An-omalopteryx, Pachyornis, and Aepyornis.

Galliformes (Fig. 7) 5 Ammoperdix, Alectoris, Rollu-lus, Lophophorus, Gallus, Lophura, Phasianus,Chrysolophus, Afropavo, Pavo, Meleagris, and Cal-lipepla.

Coraciiformes 5 Bucorvus.Gruiformes 5 Bustards: Ardeotis, Otis, Eupodotis;

Seriemas: Cariama and Chunga; other gruiforms 5Grus, Gallirallus, and Psophia.

Charadriiformes 5 Charadrius.Falconiformes 5 Sagittarius.Galliformes (Fig. 8) 5 Phasidus, Numida, Guttera,

Chrysolophus, Crossoptilon, Syrmaticus, Tragopan,Alectoris, and Francolinus.

Ornithischians 5 Lesothosaurus, Heterodontosaurus,Agilisaurus, Yandusaurus, Parksosaurus, Tenonto-saurus, Iguanodon, Prosaurolophus, Saurolophus,Edmontosaurus, Anatotitan, and Corythosaurus.

APPENDIX 2Derivation of an equation relating injurythreshold acceleration to body mass.

Alexander (1996) used animal mass and scaling the-ory to estimate how injury theshold forces depend onthe mass of an animal. We expand upon this approach,which assumes that the critical factor in determininginjury is the ratio of the impact force F to surface areaA of the animal. This ratio is assumed to be equal atinjury threshold for all animals. Thus:

F /A 5 F /A ,2 2 1 1

where the subscripts 1 and 2 label two different animals.Because mass is proportional to the cube, and area tothe square of the linear dimension,

M } A * L,

where A is area and L is the length of the animal. There-fore

3/3 1/3 2/3A} M/L } M /M } M

Thus2/3F 5 (A /A )F 5 (M /M ) F2 2 1 1 2 1 1

Dividing both sides of the last equation by M2 in orderto obtain the injury threshold acceleration a2, we find

2/3 2/3 3/3F /M 5 a 5 (M )(F )/(M )(M )2 2 2 2 1 1 2

21/3 1/3 3/35 (M )(F )(M )/M2 1 1 1

1/3 1/35 (M )(F )/(M )(M )1 1 2 1

1/35 (M /M ) (F /M )1 2 1 1

From accident data, Alexander (1996) estimated thatthe injury threshold acceleration F1/M1 for a 70-kg hu-man is about 15g, where g is the acceleration of grav-ity, 9.8 m/sec2. Substituting these numbers into the fi-nal expression above, we obtain the result

1/3a 5 (70/M ) 15g.2 2

Farlow et al, 2000

Documents

Transcript of Farlow et al, 2000