Research Article

Dynamic Locomotor Capabilities Revealed by Early Dinosaur Trackmakers from Southern Africa

  • Jeffrey A. Wilson mail,

    Affiliation: Museum of Paleontology & Department of Geological Sciences, University of Michigan, Ann Arbor, Michigan, United States of America

  • Claudia A. Marsicano,

    Affiliation: Departamento de Ciencias Geológicas, Universidad de Buenos Aires, Buenos Aires, Argentina

  • Roger M. H. Smith

    Affiliation: Department of Karoo Paleontology, Iziko South African Museum, Cape Town, South Africa

  • Published: October 06, 2009
  • DOI: 10.1371/journal.pone.0007331
  • Featured in PLOS Collections



A new investigation of the sedimentology and ichnology of the Early Jurassic Moyeni tracksite in Lesotho, southern Africa has yielded new insights into the behavior and locomotor dynamics of early dinosaurs.

Methodology/Principal Findings

The tracksite is an ancient point bar preserving a heterogeneous substrate of varied consistency and inclination that includes a ripple-marked riverbed, a bar slope, and a stable algal-matted bar top surface. Several basal ornithischian dinosaurs and a single theropod dinosaur crossed its surface within days or perhaps weeks of one another, but responded to substrate heterogeneity differently. Whereas the theropod trackmaker accommodated sloping and slippery surfaces by gripping the substrate with its pedal claws, the basal ornithischian trackmakers adjusted to the terrain by changing between quadrupedal and bipedal stance, wide and narrow gauge limb support (abduction range = 31°), and plantigrade and digitigrade foot posture.


The locomotor adjustments coincide with changes in substrate consistency along the trackway and appear to reflect ‘real time’ responses to a complex terrain. It is proposed that these responses foreshadow important locomotor transformations characterizing the later evolution of the two main dinosaur lineages. Ornithischians, which shifted from bipedal to quadrupedal posture at least three times in their evolutionary history, are shown to have been capable of adopting both postures early in their evolutionary history. The substrate-gripping behavior demonstrated by the early theropod, in turn, is consistent with the hypothesized function of pedal claws in bird ancestors.


The earliest dinosaurs were small, bipedal forms that are first recorded ca. 228 million years ago in the Late Triassic, when the Earth's continental landmasses were interlocked as Pangea [1]. Already present in these earliest dinosaur-bearing sediments are two basal lineages, Ornithischia and Theropoda, both of whose locomotor evolution would diverge dramatically from the ancestral dinosaurian condition during the Jurassic [2]. The ornithischian line gave rise to three different lineages, each capable of quadrupedal locomotion, during the Early and Middle Jurassic (ca. 200–161 Ma), and the theropod line gave rise to birds capable of powered flight by the close the Late Jurassic (145 Ma). We report new evidence from an Early Jurassic tracksite in southern Africa [3] that provides a rare glimpse at dynamic locomotor capabilities of these two dinosaur lineages at an early phase in their evolution (Fig. 1). These dynamic capabilities, which are not directly apparent from the body fossil record, imply a greater locomotor plasticity in early ornithischians than in early theropods and bring to light a previously unknown functionality in theropods.


Figure 1. The Moyeni trackway locality and its stratigraphic and phylogenetic context.

The thumbnail map of Africa shows areal extent of main Karoo Basin (grey) and the country of Lesotho (yellow). The phylogeny depicts the basic interrelationships of major dinosaur clades [41] on a timescale. Subsequent analyses have resolved Lesothosaurus at the base of Thyreophora [45], but most agree that it is positioned near the base of clade uniting Thyreophora, Ornithopoda, and Marginocephalia. The Moyeni tracksite (marked by yellow band) preserves tracks made by early dinosaur trackmakers, well after the initial divergence of saurischians and ornithischians (early Late Triassic) but well before the origin of flight (Upper Jurassic) and prior to three independent acquisitions of quadrupedal posture (Early–Middle Jurassic) [41], [42]. The graded stratigraphic range for Marginocephalia reflects uncertainty in its first appearance date; icons atop diagram are representative theropod and ornithischian dinosaurs [41].


Moyeni Tracksite, Lesotho

Extensive, well-preserved fossil tetrapod trackways of the Stormberg Group in the Karoo Basin of southern Africa provide significant information about faunal diversity and turnover during the early Mesozoic in southern Pangea [4][6]. Many of these trackways occur in erosion-resistant sandstones of the Elliot Formation that are positioned near the Triassic-Jurassic boundary (199.6 Ma; [7]). Perhaps the most spectacular is the Moyeni tracksite in southern Lesotho, which was discovered and first described in detail by Paul Ellenberger [8][10]. The Moyeni tracksite is generally considered to be Early Jurassic in age [6], but uncertainty about the exact position of the Triassic–Jurassic boundary within the Elliot Formation precludes more specific temporal resolution.

Originally, Ellenberger [3] described 13 different track-types at Moyeni, all of which he considered endemic to the Karoo Basin, and provided detailed reconstructions of the behavior implied by many of the longer trackways. A critical influence on Ellenberger's reconstructions of trackmaker behavior was his interpretation of the depositional environment as an emergent, elongated sandbank formed at the mouth of a river flowing into a large lake. Consequently, he interpreted many of the tracks we discuss herein as having been made by swimming dinosaurs that had webbed hind feet (see Text S1).

The Moyeni tracksite records more than 250 tetrapod footprints and associated invertebrate traces on a 100 m2 sandstone surface (Fig. 2). The remarkable richness and level of detail preserved at the Moyeni tracksite offer a unique opportunity to investigate the locomotor habits of early dinosaurs.


Figure 2. Map of major geological features and tracks at the Moyeni tracksite.

Trackways of Anomoepus and Grallator are in solid green and blue, respectively; all other trackways are in 50% grey. Trackway numbering (Arabic numerals) begins with the first recognized step of each trackmaker. Grallator tracks 1 and 2 are not shown because they are now underneath a retaining wall [3]. Red wavy lines indicate ripple marks; solid light grey fields indicate algal-matted surface. The inset highlights postural and gait changes in the Anomoepus trackway, which was made by a basal ornithischian. A shift from wide-gauge to narrow-gauge posture (between tracks 13 and 14) is marked by shift in pace angulation (blue dotted line) and was accompanied by a brief pause, during which the tail registered on the substrate when tracks 12 and 13 were impressed. The shift from a quadrupedal to a bipedal gait (between tracks 15 and 16) occurs atop the algal matted surface. Note that Grallator track 11 (blue) overprints Anomoepus track 15, indicating it was made later. Grid pattern forms 1 meter squares for both maps. Abbreviations: lm, left manus; lp, left pes; mt, metatarsus; rm, right manus; rp, right pes; t, tail; Roman numerals indicate pedal digits.


Paleoenvironmental Interpretation

We interpreted the sedimentary sequence below and above the track-bearing surface as having accumulated on the inside bank of a meander loop rather than in a lacustrine setting, with the main trackway surface preserved on one of several scroll bars making up a low-angle point bar [11]. There, fossil trackways probably were imprinted over the course of days or weeks on the sandy inner bank of a channel meander, recording a brief moment in a southern African terrestrial ecosystem. The trackway surface has three distinctive sections that differ in their slope and surface texture: a lower, ripple-marked portion; a smooth bar slope inclined at approximately 20°; and a relatively flat bar top with a distinctive algal-matted texture (Fig. 2). Tracks cover nearly the entire area, but they are most abundant and best preserved on the algal-matted top surface. Here we focus on the two longest trackways, which cross all three surfaces, and the trackmakers' dynamic responses to differences in slope and slipperiness.

Ichnotaxonomy and Trackmaker Identification

Although Ellenberger [3] recognized similarities with tracks described from the Newark Supergroup [12], [13], he created 7 new ichnogenera and 13 new ichnospecies to encompass the track diversity at Moyeni. The two dinosaur track-types that we will focus on in this contribution were named Neotrisauropus deambulator and Moyenisauropus natator [3](Text S1). Five other ichnospecies were named for the latter ichnogenus and distinguished by somewhat subtle features—many of which may reflect behavioral and/or preservational, rather than anatomical, variation. Revisions of Ellenberger's ichnotaxonomy synonymized all but one of the Moyeni ichnotaxa, subsuming Neotrisauropus and Moyenisauropus within Grallator and Anomoepus, respectively [4], [14][16]. Based on our first-hand observations of the Moyeni footprints, we concur that there is little morphological distinction between the Moyeni and Newark ichnogenera or between ichnospecies of Moyenisauropus and apply the revised ichnogenus names here. Future taxonomic work must determine whether the various ichnospecies named by Ellenberger must also be synonymized.

The fossilized tracks and trackways at Moyeni provide direct evidence of how early dinosaurs behaved in life. However, because trackways record the interaction of soft tissues with the substrate—typically only the undersides of the manus and pes—trackmaker identifications are imprecise and rely on a combination of stratigraphic, geographic, and morphological coincidence with body fossils [17]. Despite this constraint, fossil tracks can provide powerful insight into trackmaker paleobiology. In this contribution, we use soft tissue features preserved in the tracks to identify skeletal synapomorphies diagnosing particular trackmaker subgroups [18]. This method results in coarse yet falsifiable identifications that can be used to interpret trackmaker paleobiology. Tetrapod footprints at Moyeni include tracks produced ornithischian and theropod dinosaurs (ichnogenera Anomoepus, Grallator), basal tetrapods (ichnogenus Episcopopus), and crurotarsal archosaurs (“chirotheroid”-type ichnotaxon) [11].


The map in Figure 2 was made by gridding the entire trackway surface, photographing each 1 m square, and drawing all footprints at 1:10 scale. Trackway measurements reported in Text S2 follow standard protocol, summarized in [19]. Stride length is measured as the straight-line distance between homologous points on successive footfalls of the same foot. Pace length is measured as the straight line distance between homologous points on left and right manus or pes prints; successive pace lengths (i.e., R–L–R or L–R–L) form an angle that is measured as pace angulation. Linear measurements were made to the nearest 0.5 centimeter and were measured in situ unless noted; angular measurements were made to the nearest 0.5 degree using a high-resolution map of the trackway surface. Some tracks are no longer accessible because they are no longer visible (e.g., beneath retaining wall) or no longer preserved. These track measurements were estimated to the nearest 1.0 cm or degree based on maps in [3]. Gauge width was measured as the straight-line distance between a foot and the midpoint between two footfalls of its opposite (i.e., the midpoint of the stride). Three-dimensional data were collected using a HandyScan HZ at 0.63–0.88 mm resolution.


The two dinosaur trackmakers recorded numerous tracks and trackways at Moyeni. We focus on two lengthy trackways of Anomoepus and Grallator [3], [4] that traversed the heterogeneous surface of the point bar in opposite directions in close temporal succession, as indicated by a Grallator footprint that overprinted a fresh Anomoepus footprint as it crossed its path (Figs. 2, 3). We identify these trackmakers as dinosaurs because both trackways indicate animals capable of bipedal posture walking with a parasagittal gait on functionally tridactyl (three-toed) pes. Grallator can be identified more specifically as a theropod dinosaur because its tracks lack metatarsal traces, and the functionally tridactyl foot bears an elongate digit III and clearly marked, sharply pointed pedal ungual traces. Anomoepus, in turn, can be identified as a basal ornithischian by its symmetrical, functionally tridactyl foot bearing blunt unguals and a pentadactyl manus with subequal digits with gently rounded ends [16], [19]. The lack of sharp ungual traces in Anomoepus excludes basal sauropodomorphs, theropods, and heterodontosaurids as potential trackmakers, because all can be expected to create manus prints with one or more sharply defined ungual traces.


Figure 3. Moyeni dinosaur tracks.

Photographs of plaster casts (positives) of Grallator track 6 (A) and Anomoepus track 8 (B) made at the Moyeni tracksite by the authors. Tracks are shown at the same scale (10 cm), and hatching pattern indicates broken surfaces. The Grallator hind foot print was made by pedal digits II–IV; the trackmaker's phalangeal formula was 3–4–5. Digits I and V did not contact the substrate. The rugose texture surrounding the print is the algal mat. The Anomoepus manus–pes couple registers all five manual digits (i–v), four pedal digits (I–IV), the metatarsus (mt), and toe drag marks (dm). Additional structures to the left of the pes are incidental marks made by a different trackmaker.


Theropod Trackmaker (Grallator)

Few noticeable changes are apparent along the one definitive theropod trackway preserved at Moyeni, despite substantial differences in substrate. The trackway consists of 25 tridactyl pes prints that traverse the point bar top and curve gently down the slope onto the rippled surface at or very near the water's edge (Fig. 2). Digital pads are well-defined, and a typical 3–4–5 phalangeal formula can be inferred for the trackmaker's three weight-bearing digits (Fig. 3A). There is usually no trace of digit V or the metatarsus on any theropod track, unless the animal is sitting [20]. Grallator pes prints at Moyeni average 28 cm in length, which is slightly larger than typical Grallator prints but intermediate between the maximal pes lengths recorded in Late Triassic and Early Jurassic theropod trackmakers of the Newark Supergroup [21]. Footprint length can be used to estimate a hip height of 1.4 m and a body length of approximately 6–7 m [19], [22], which is close to the size of the coeval southern African theropod Dracovenator regenti [23]. Despite differences in consistency and inclination of the substrate, the Grallator trackmaker never deviated from a digitigrade, bipedal posture with a parasagittal gait. Pace angles, measured across three successive footfalls (left–right–left or right–left–right) along the length of the trackway, are consistently high (ca. 170°; Text S2). In contrast, stride length was affected by changing topography, decreasing to 75% maximum stride length over the last 7 steps as the trackmaker left the flat surface of the bar and descended the slope onto the river margin. These shortened strides, which suggest reduced speed, were accompanied by a slight forward shift in body weight, registration of the digit I on the substrate (noted by [3]), and the appearance of subrectangular-to-oval shaped claw impressions on the second and third digits (Fig. 4A). Registration of a digit I trace is usually associated with deep tracks [24], but it occasionally occurs in shallow tracks of the Grallator-like ichnotaxon Gigandipus from North America, often in association with tail drags [13], [25]. At the Moyeni tracksite, impression of digit I occurs without a tail drag or any appreciable deepening of the track, and it is associated with subrectangular-to-oval shaped claw impressions that we interpret as the result of strong flexion of the second and third pedal claws. In these flexed traces, the ungual phalanx and its keratinous sheath deeply penetrated the substrate and left a trace of its cross-section at the surface. This suggests that the inner three digits flexed together and actively gripped the substrate as a real time behavioral response to a sloping and slippery surface.


Figure 4. Dynamic locomotor adjustments to paleosurface heterogeneity. A, “flexed-ungual” locomotion in the Grallator trackway, which is attributed to a theropod.

Images at right show the impression of terminal end of the third pedal digit in tracks (tr) 5, 16, 17, and 24. Differences in the shape of terminal impression reflect deeper penetration of the ungual into the substrate, as shown in corresponding schematic interpretations of digit III at right. In the first and last panel, the tip of the ungual makes a narrow, pointed impression. In the middle two panels, the ungual has penetrated deeper into the substrate so that its base makes a rounded impression at the surface. Silhouette morphology and proportions based on Allosaurus. B, postural changes in Anomoepus. Panels show our interpretation of limb posture during plantigrade, wide-gauge locomotion (top), and digitigrade, narrow-gauge locomotion (bottom) in lateral (left) and anterior (right) views. Silhouettes are based on limb proportions and skeletal morphology common to basal ornithischians [28][34].


Two lines of evidence support this interpretation over an alternative explanation for these tracks as a natural consequence of a forward shift in body weight. First, the pre-ungual portion of the pes is shallowly impressed and pre-ungual phalanges are identical in size and shape to previous steps in the trackway. Second, flexion of the second and third ungual phalanges measurably lengthened the second and third digits (absolutely and relative to the fourth digit) compared to those impressed on the bar top (Fig. 4A). Although somewhat counterintuitive, the flexed digits are longer because more of their arc of rotation during the step cycle is recorded in the substrate (Fig. 5).


Figure 5. Arc of rotation of ungual during the step cycle.

Flexed ungual tracks are longer than typical tracks because the ungual leaves a longer impression when it is submerged within the sediment (compare upper and lower lines).


Basal Ornithischian Trackmaker (Anomoepus)

In contrast to the theropod trackway, the basal ornithischian trackways at Moyeni shift between three distinct locomotor styles, each of which is associated with a different substrate consistency and slope. North American Anomoepus has been interpreted as made by a facultative quadruped, based on abundant bipedal and rare quadrupedal tracks assigned to the same ichnotaxon [16], but the Moyeni tracksite offers a unique opportunity to record bipedal and quadrupedal locomotion in a single trackway and to describe the transition between them in detail. This is best demonstrated in the longest Anomoepus trackway, which consists of 17 steps that begin on the rippled channel margin and turn sharply up the slope of the point bar and onto the algal-matted upper surface of the point bar (Figs. 2, 3B). Anomoepus manus and pes prints along this trackway average 13 cm and 21 cm in length, respectively, indicating a hip height of 1.2 m and body length similar to that of the Grallator trackmaker (5–6 m; [19], [22]). Digital pads are not well defined, and manual and pedal phalangeal counts are not known. Although digit I left a trace in some Anomoepus tracks, the pes was functionally tridactyl, as indicated by deeper impressions of the internal three digits (II, III, IV), which are large, relatively broad, splayed, and terminate in blunt claws (Fig. 3). When present, manus tracks are more lightly impressed, but they are clearly rotated outwards, pentadactyl, and tipped by rounded unguals.

The first 13 steps in the Anomoepus trackway register “wide-gauge” manus and pes impressions that are placed at a distance from the trackway midline (Fig. 2). In these wide-gauge tracks, both digit I and the metatarsus are recorded. The shallow depth of the tracks indicates that the uniform registration of both digit I and the metatarsus is a consequence of a plantigrade postural feature, rather than of deep penetration of the pes into a poorly consolidated substrate [24]. In addition, the shape and size of the metatarsus impressions match those of Anomoepus resting traces found elsewhere on the tracksite [11]. Lengthy toe drag marks extending between successive footfalls may have been the result of reduced clearance of the foot during the swing phase of locomotion, indicating that the trackmaker did not “spring up” during the step cycle. Together, the plantigrade hind foot posture and wide-gauge stance suggest that the Anomoepus trackmaker adopted a more stable, crouching pose and moved more slowly along the ripple-marked riverbed surface (Fig. 4B). We note here that Ellenberger [3] interpreted these wide-gauge tracks and drag marks as made by a swimming animal, an interpretation we disagree with based on the strong registration of the pes and metatarsus on the substrate (which indicates the animal was supporting its body weight), the regularity of the footfall pattern, and the inferred shallow water in the riverbed [11]. Moreover, the Anomoepus trackway lacks features typically preserved in swimming traces, such as shortened prints (which indicates that the animal was buoyed by the water mass) and sigmoid scratch marks of variable length with sediment piled up at their base [26], [27].

The last 5 steps are “narrow gauge” with the limbs positioned along the midline, underneath the body. The transition between the straddling, wide-gauge, plantigrade posture and a parasagittal, narrow-gauge, digitigrade posture was punctuated by a tail impression made during a brief pause (Fig. 2). Pace angulations for the 5 narrow-gauge tracks imprinted on the bank and upper point bar surface are significantly higher than the 13 wide-gauge tracks, all but one of which were made on the rippled surface (186° vs. 107°; Text S2). Additionally, there is no impression of the metatarsus, digit I, or dragged toes on any of the narrow-gauge tracks (steps 14–17). This indicates an elevated, digitigrade pedal posture in which each foot was lifted clear of the substrate during the stride. The final two narrow-gauge steps (steps 16–17) bear no manus prints and demonstrate a third postural change from quadrupedal to bipedal locomotion on the upper algal-matted point bar surface.

The wide-gauge walking recorded in the rippled portion of the trackway indicates a greater range of abduction than implied by Anomoepus resting traces preserved elsewhere at the Moyeni tracksite [3]. In resting traces, the implied gauge width is approximately 62% the combined length of the metatarsus and pes; gauge width is twice this in the rippled portion of the trackway (120%). In addition, the differences in gauge width measured along a single trackway allow estimation of abduction angles for the limb. Assuming hindlimb proportions common to basal ornithischians [28][34], a hip height of 1.2 m, and taking into account differences in foot posture, we estimate that limb abduction angle ranged between 22° at maximum gauge width (46 cm) and −9° at minimum gauge width (−4 cm), in which the feet stepped across the midline (Fig. 4B). This implies the Anomoepus trackmaker was capable of a range of at least 31° of limb abduction. Although it seems likely that most of this mobility was exercised at the proximal, ball-and-socket joint (i.e., hip), some of it may have been taken up at the more distal hinge joints (i.e., knee, ankle).

In summary, in the riverbed, Anomoepus maintained a wide-gauge, quadrupedal, gait with a crouching, plantigrade posture. In this crouched position, the animal was apparently not able to fully lift its toes clear of the substrate and left long drag marks that are truncated by the succeeding footprint. On the slope of the bank, the trackmaker transitioned to a more elevated and parasagittal, but still quadrupedal gait. On the stable bar top, the trackmaker adopted a bipedal, parasagittal gait with an upright posture (Figs. 2, 4B). No other Anomoepus trackway crossed all three parts of the point bar, but several crossed the inclined slope and algal matted bar top portions. In each case and irrespective of direction of travel, impressions of the manus, metatarsus, and digit I are associated with the inclined slope and absent on the bar top. All Anomoepus trackways on the inclined slope and bar top are narrow-gauge.


High-fidelity preservation of long, continuous trackways of early dinosaurs crossing a heterogeneous paleosurface at Moyeni documents real time responses to substrate quality and inclination. This particular set of taphonomic and sedimentological circumstances is rare and allows detailed reconstruction of the locomotor behavior of the trackmakers. Nonetheless, we infer that the locomotor behaviors inferred from the Moyeni tracks are likely to be common to other basal ornithischians and theropods, due to their morphological similarity to contemporaneous North American track-types [4], [14][16]. Accordingly, we suggest that the dynamic adjustments left by early dinosaur trackmakers at Moyeni have the following implications for dinosaur locomotor evolution.

Despite the unevenness of the ground, the theropod (Grallator trackmaker) did not modify foot posture or adopt a more stable, wide-gauge gait. Although there are examples of theropods adopting a wider stance at low speeds [35] and at rest [20], [36], as well as rare examples of metatarsal traces in deep tracks [24], the vast majority of theropod trackways known to us indicate a narrow-gauge, bipedal gait with digitigrade foot posture [37]. This suggests that theropods were either anatomically incapable or behaviorally reluctant to widen their gait while moving at typical speeds, perhaps because they were able to accommodate surface heterogeneity in other ways, such as the “flexed-ungual” posture observed at the Moyeni tracksite. We suspect that the ability to grip the substrate with the pes was important because it relieved the forelimb from a role in body support. Although seldom recorded in footprints, this feature may have been present in theropods and their immediate ancestors, but absent in ornithischians and sauropodomorphs. This functionality was inherited by descendant theropods close to the bird line, whose ability to climb inclined surfaces using their unguals [38] has recently been implicated in the “wing assisted inclined running” hypothesis for origin of flight [39], [40].

In marked contrast, the Anomoepus trackways demonstrate that basal ornithischians possessed a broad range of functional responses to substrate changes. These include changes to trackway gauge (wide vs. narrow) facilitated by limb abduction, foot posture (plantigrade vs. digitigrade) accommodated by ankle flexibility, and the number of supporting limbs (quadrupedal vs. bipedal). The ability of the Anomoepus trackmaker to facultatively enlist the front limbs in locomotion in real time foreshadows the three independent evolutionary acquisitions of facultative or fully quadrupedal posture in ornithischian history [41], [42]. This ability has been long inferred for basal ornithischians [43], based on osteological evidence and on their phylogenetic intermediacy between bipedal ancestors and quadrupedal descendants, but it is demonstrated in the Moyeni trackway. Trackway evidence from Moyeni suggests that basal ornithischians were capable of facultative quadrupedalism less than 30 million years after their origin from an obligatorily bipedal ancestor and contemporaneous with the appearance of the first quadrupedal ornithischian body fossils [44].

Supporting Information

Text S1.

Ellenberger translation. This is an English-language translation of two portions of Ellenberger (1974) that name and describe the ichnotaxa discussed the text. The first part is the author's description of Neotrisauropus. The second is a lengthier section on Moyenisauropus. The original text is in French. Translators: Emile Moacdieh and John Whitlock; editor: Jeff Wilson.


(0.14 MB PDF)

Text S2.

Measurements of two dinosaur trackways from Moyeni, Lesotho. Includes measurements of lengthy trackways of the dinosaur ichnotaxa Anomoepus and Grallator.


(0.07 MB PDF)


We thank T. Baumiller, S. Gatesy, and P. Meyers for discussions, and M. Carrano, M. D'Emic, T. Ikejiri, D. Pol, and E. Rainforth for comments on previous drafts of this paper. F. Knoll and A. R. C. Milner provided thoughtful and thorough reviews that improved this paper. We thank B. Miljour for assistance with the figures. We thank B. Britt and J. Madsen for lending us a cast of an Allosaurus pes. E. Moacdieh and J. Whitlock translated excerpts of Ellenberger (1974) from the French (Text S1). C. Sidor provided assistance mapping the tracksite. We are grateful to S. O'Grady and the UM 3D lab for assistance with the laser scanner and J. Whitlock and D. Fisher for assistance with 2D rendering. Prof. David Ambrose MBE, KCMMOM, National University of Lesotho, provided useful information about Moyeni and other tracksites in Lesotho.

Author Contributions

Conceived and designed the experiments: JAW CAM RMHS. Performed the experiments: JAW CAM RMHS. Analyzed the data: JAW CAM RMHS. Contributed reagents/materials/analysis tools: JAW CAM RMHS. Wrote the paper: JAW CAM RMHS.


  1. 1. Rogers RR, Swisher CC III, Sereno PC, Monetta AM, Forster CA, et al. (1993) The Ischigualasto tetrapod assemblage (Late Triassic, Argentina) and 40Ar/39Ar dating of dinosaur origins. Science 260: 1–4.
  2. 2. Sereno PC, Forster CA, Rogers RR, Monetta AM (1993) Primitive dinosaur skeleton from Argentina and the early evolution of Dinosauria. Nature 361: 64–66.
  3. 3. Ellenberger P (1974) Contribution à la classification des pistes de vertèbres du Trias: les types du Stormberg d'Afrique du Sud (II partie): le Stormberg Superieur-I. Le biome de la zone B/1 ou niveau de Moyeni: ses biocenoses. Paleovertebrata Mémoire Extraordinaire 1: 1–202.
  4. 4. Olsen PE, Galton PM (1984) A review of the reptile and amphibian assemblages from the Stormberg of southern Africa, with special emphasis on the footprints and the age of the Stormberg. Paleontol Africana 25: 87–110.
  5. 5. Knoll F (2004) Review of the tetrapod fauna of the “Lower Stromberg Group” of the main Karoo Basin (southern Africa): implication for the age of the Lower Elliot Formation. Bull Soc Géol France 175: 73–83.
  6. 6. Knoll F (2005) The tetrapod fauna of the Upper Elliot and Clarens formations in the main Karoo Basin (South Africa and Lesotho). Bull Soc Géol France 176: 81–91.
  7. 7. Gradstein FM, Ogg JG, Smith AG (2004) A Geologic Time Scale 2004. Cambridge: Cambridge University Press.
  8. 8. Ellenberger F, Ellenberger P, Fabre J, Mendrez C (1963) Deux nouvelles dalles à pistes de vertébrés fossiles découvertes au Basutoland (Afrique du Sud). C R Somm Séance Soc Géol France 1963: 315–317.
  9. 9. Ellenberger F, Ellenberger P, Ginsburg L (1967) The appearance and evolution of dinosaurs in the Trias and Lias: a comparison between South African Upper Karoo and western Europe based on vertebrate footprints. In: International Symposium on Gondwana Geology. Mar del Plata: UNESCO. pp. 333–354.
  10. 10. Ellenberger P (1970) Les niveaux paléontologiques de premiere apparition des mammiferes primordiaux en Afrique du Sud et leur ichnologie. Establissement de zones stratigraphiques detaillees dans le Stormberg de Lesotho (Afrique du Sud) (Trias Superieur a Jurassique). Proc II Gondwana Symp Pretoria 2: 340–370.
  11. 11. Smith RMH, Marsicano C, Wilson JA (in press) Sedimentology and paleoecology of a diverse Early Jurassic tetrapod tracksite in Lesotho, southern Africa. Palaios 24:
  12. 12. Lull RS (1915) Triassic life of the Connecticut Valley. Bull Conn Geol Nat Hist Surv 24: 1–285.
  13. 13. Lull RS (1953) Triassic life of the Connecticut Valley. Bull Conn Geol Nat Hist Surv 81: 1–336.
  14. 14. Haubold H (1986) Archosaur footprints at the terrestrial Triassic–Jurassic transition. In: Padian K, editor. Beginning of the Age of Dinosaurs. New York: Cambridge University Press. pp. 189–201.
  15. 15. Gierlinski G (1991) New dinosaur ichnotaxa from the Early Jurassic of the Holy Cross Mountains, Poland. Palaeogeogr Palaeoclimatol Palaeoecol 85: 137–148.
  16. 16. Olsen PE, Rainforth EC (2003) The Early Jurassic ornithischian dinosaurian ichnogenus Anomoepus. In: Letourneau PM, Olsen PE, editors. The Great Rift Valley of Pangea in eastern North America, volume two. New York: Columbia University Press. pp. 314–353.
  17. 17. Olsen PE, Baird D (1986) The ichnogenus Atreipus and its significance for Triassic biostratigraphy. In: Padian K, editor. Beginning of the Age of Dinosaurs. New York: Cambridge University Press. pp. 61–87.
  18. 18. Carrano MT, Wilson JA (2001) Taxon distributions and the tetrapod track record. Paleobiol 27: 564–582.
  19. 19. Thulborn RA (1990) Dinosaur Tracks. London: Chapman and Hall.
  20. 20. Lockley M, Matsukawa M, Li J (2003) Crouching theropods in taxonomic jungles: ichnological and ichnotaxonomic investigations of footprints with metatarsal and ischial impressions. Ichnos 10: 169–177.
  21. 21. Olsen PE, Kent DV, Sues HD, Koeberl C, Huber H, et al. (2002) Ascent of dinosaurs linked to an Iridium anomaly at the Triassic-Jurassic boundary. Science 296: 1305–1307.
  22. 22. Henderson DM (2003) Footprints, trackways, and hip heights of bipedal dinosaurs—testing hip height predictions with computer models. Ichnos 10: 99–104.
  23. 23. Yates A (2005) A new theropod dinosaur from the Early Jurassic of South Africa and its implications for the early evolution of theropods. Palaeontol Africana 41: 105–122.
  24. 24. Gatesy SM, Middleton KM, Jenkins FA Jr, Shubin NH (1999) Three-dimensional preservation of foot movements in Triassic theropod dinosaurs. Nature 399: 141–144.
  25. 25. Rainforth EC (2002) Tails of saurischian dinosaurs in the Early Jurassic of the Newark Supergroup (eastern North America). GSA Abs Progs 34: 61.
  26. 26. Ezquerra R, Doublet S, Costeur L, Galton PM, Pérez-Lorente F (2007) Were non-avian theropod dinosaurs able to swim? Supportive evidence from an Early Cretaceous trackway, Cameros Basin (La Rioja, Spain). Geology 35: 507–510.
  27. 27. Milner ARC, Lockley MG, Kirkland JI (2006) A large collection of well-preserved theropod dinosaur swim tracks from the Lower Jurassic Moenave Formation, St. George. In: Harris JD, Lucas SG, Spielmann JA, Lockley MG, Milner ARC, et al., editors. The Triassic-Jurassic Terrestrial Transition: New Mexico Mus Nat Hist Sci Bull. 37. : 315–328.
  28. 28. Casamiquela RM (1967) Un nuevo dinosaurio ornitisquio Triasico (Pisanosaurus mertii; Ornithopoda) de la Formación Ischigualasto, Argentina. Ameghiniana 4: 47–64.
  29. 29. Thulborn RA (1972) The post-cranial skeleton of the Triassic ornithischian dinosaur Fabrosaurus australis. Paleontol 15: 29–60.
  30. 30. Santa Luca AP (1980) The postcranial skeleton of Heterodontosaurus tucki (Reptilia, Ornithischia) from the Stormberg of South Africa. Ann S Afr Mus 79: 159–211.
  31. 31. Colbert EH (1981) A primitive ornithischian dinosaur from the Kayenta Formation of Arizona. Mus N Arizona Press Bull Ser 53: 1–61.
  32. 32. Sereno PC (1991) Lesothosaurus, “fabrosaurids,” and the early evolution of Ornithischia. J Vertebr Paleontol 11: 168–197.
  33. 33. Butler RJ (2005) The ‘fabrosaurid’ ornithischian dinosaurs of the Upper Elliot Formation (Lower Jurassic) of South Africa and Lesotho. Zool J Linn Soc 145: 175–218.
  34. 34. Butler RJ, Smith RMH, Norman DB (2007) A primitive ornithischian dinosaur from the Late Triassic of South Africa, and the early evolution and diversification of Ornithischia. Proc R Soc London B 274: 2041–2046.
  35. 35. Day JJ, Norman DB, Upchurch P, Powell HP (2002) Dinosaur locomotion from a new trackway. Nature 415: 494–495.
  36. 36. Milner ARC, Harris JD, Lockley MG, Kirkland JI, Matthews NA (2009) Bird-like anatomy, posture, and behavior revealed by an Early Jurassic theropod dinosaur resting trace. PLoS ONE 4: e4591. doi:4510.1371/journal.pone.0004591.
  37. 37. Farlow JO, Gatesy SM, Holtz TR Jr, Hutchinson JR, Robinson JM (2000) Theropod locomotion. Am Zool 40: 640–663.
  38. 38. Bock WJ, Miller WD (1959) The scansorial foot of the woodpeckers with comments on the evolution of perching and climbing feet in birds. Am Mus Novitates 1931: 1–45.
  39. 39. Dial KP (2003) Wing-assisted incline running and the evolution of flight. Science 499: 402–404.
  40. 40. Dial KP, Randall RJ, Dial TR (2006) What use is half a wing in the ecology and evolution of birds? BioScience 56: 437–445.
  41. 41. Sereno PC (1999) The evolution of dinosaurs. Science 284: 2137–2147.
  42. 42. Carrano MT (2000) Homoplasy and the evolution of dinosaur locomotion. Paleobiol 26: 489–512.
  43. 43. Norman DB, Witmer LM, Weishampel DB (2004) Basal Ornithischia. In: Weishampel DB, Dodson P, Osmólska H, editors. The Dinosauria, 2nd edn. Berkeley: University of California Press. pp. 325–334.
  44. 44. Owen R (1861) A monograph on fossil Reptilia from the Liassic formations. I. Scelidosaurus harrisonii. Monog Palaeontograph Soc 13: 1–14.
  45. 45. Butler RJ, Upchurch P, Norman DB (2007) The phylogeny of the ornithischian dinosaurs. J Syst Palaeontol 6: 1–40.