Introduction

The prosauropod Plateosauru engelhardti Meyer, 1837 from the Late Triassic of Central Europe is well known from many specimens from a number of locations, several of them in nearly perfect articulation. Among early dinosaurs, it is one of the best-known genera. The finds have caused heated debate about the locomotory adaptations of the animal. Huene (1907-08, 1926, 1928) argued for digitigrade bipedality and attributed a grasping function to the manus. He was convinced that Plateosaurus was an obligate biped, much as the similar Anchisaurus that had been described as exclusively bipedal and digitigrade by Marsh (1893a, 1893b). A good indicator for this is the highly divergent length of fore- and hindlimbs. Huene had the mounts of GPIT1 and GPIT2 (Figure 1.1) set up in Tübingen in bipedal postures and adamantly stuck to this interpretation of the material, despite criticism from many sources.

Other researchers have suggested practically any possible stance: obligate quadrupedality and plantigrady 'like lizards' was proposed by Jaekel (1910), who later changed his mind and concluded a clumsy, kangaroo-like hopping as the only possible mode of locomotion (Jaekel 1911, 1913-14). Fraas (1912, 1913) referred to the (dorsoventrally compressed by sediment compaction) position of the skeletal finds in the field, arguing for a sprawling obligatorily quadrupedal gait. He had the skeleton SMNS 13200 mounted in this position in the Stuttgart museum.

Later, researchers began to agree with Huene on the issue of digitigrady, although plantigrady made a comeback in Sullivan et al. (2003). Weishampel and Westphal (1986) depicted Plateosaurus running digitigrade and bipedally, but they argued for facultative quadrupedality. Interestingly, the metacarpals were shown widely spread, in marked contrast to the interpretation by Huene (1926, but contra the reconstruction drawing in that publication), a position that does not seem to fit an active role of the manus in locomotion. Paul (1997) also argued for bipedality, but his outlined skeletal drawing seemed to imply permanent quadrupedality, not bipedality. Facultative bipedality was also proposed by Van Heerden (1997).

Among others, Galton (1971a, 1976, 1990, 2000) advocated facultative bipedality in prosauropods. He based his opinion on the hindlimb to trunk ratio, which was also invoked by Bonaparte (1971). Wellnhofer (1994) also depicted Plateosaurus in a quadrupedal stance, based on characteristics of the tail of material from Ellingen now in the BSP, which he figured with a strong downward curve making a bipedal stance impossible.

In the first functional morphology approach on Plateosaurus locomotion, Christian et al. (1996) studied the vertebral column's resistance to bending in various vertebrates in order to determine their locomotory modes. Since Plateosaurus shows an intermediate pattern between obligate bipeds and obligate quadrupeds, exhibiting a medium peak of resistance to bending over the shoulders instead of either the small peak of bipeds or the large peak of quadrupeds, Christian et al. (1996) argued that the animal was probably facultatively bipedal at high speeds only. Christian and Preuschoft (1996) investigated the shape of the acetabulum and agreed with Huene 1926 on a near-vertical position of the femur in anterior view instead of a more sprawled configuration.

The latest extensive publication on the osteology of Plateosaurus, Moser (2003), claimed that Plateosaurus would only have been capable of tiny shuffling steps when walking bipedally. Moser also described the remounting of the skeletal mount previously exhibited in the BSP for the Naturhistorische Gesellschaft in Nuremberg. The animal was forced into what Moser called the track of a quadrupedal prosauropod (Moser did not specify which track exactly; probably he referred to the track depicted in Moser (2003: fig. 28, Tetrasauropus unguiferus, from Ellenberger [1972]), despite the fact that the track exhibits medially curving and rotated short toes and fingers, while almost all articulated finds of Plateosaurus and the morphology of the phalangal articular surfaces indicate no longitudinal rotation of the toes or fingers. Also, the toes of Plateosaurus are long and slender. Moser (2003) suggested that this discrepancy may indicate an early sauropod instead of a prosauropod as the trackmaker. Galton (1971a, 1971b) already suggested medial curving of the fingers in Plateosaurus, on the basis of GPIT1, and proposed a walking configuration of the manus that kept the laterally diverging first digit's claw off the ground. Baird (1980) concluded that the track of Navahopus falcipollex from the Navaho Sandstone of Arizona was made by a quadrupedally walking plateosaurid dinosaur. The ichnofossil Otozoum, originally described by Hitchcock (1847) and redescribed by Rainforth (2003), is another candidate for a prosauropod track: a bipedal track with two manus imprints on which the fingers point outward at a right angle to the direction of movement. The imprints fit the shape of the Plateosaurus manus and pes in a semi-plantigrade position. The pes of Plateosaurus has been claimed also to fit the pes print of, suggested as another possible quadrupedal prosauropod track by Lockley and Meyer (2000). Porchetti and Nicosia (2007) concluded that a Plateosaurus-like prosauropod is a possible creator of Pseudotetrasauropus, again requiring a more or less plantigrade position.

The latest development in the controversy about the locomotory abilities of Plateosaurus is an assessment of the range of motion of the forelimb of Plateosaurus and the close relative Massospondylus by Bonnan and Senter (2007), which indicated that manus pronation was impossible. Therefore, Bonnan and Senter (2007) concluded that Plateosaurus was an obligate biped.

Recent research (Fechner 2006; Remes 2006, 2008) suggests that the paradigm of obligatorily bipedal ancestral dinosaurs may be wrong, and that the first dinosaurs were at most facultatively bipedal, holding their hind limbs in strongly flexed poses during quadrupedal locomotion. Such a posture allows subequal functional limb and thus stride lengths despite significantly different total limb lengths, and is similar to the posture of many small mammals. If this model of early dinosaurian locomotion is correct, then the obligate quadrupedal posture of sauropods is a primitive character, most small ornithischians with highly different limb lengths may have been facultatively quadrupedal as well, and bipedal posture would have evolved de novo and potentially separately in several dinosaurian lineages, such as theropods, some 'prosauropods', and some ornithopods.

The best approach to assess the locomotion capabilities of an extinct animal is to create an exact 3D digital mount of it, based on a well-preserved complete individual, and test all potentially possible postures for osteological and for kinematic probability. For the latter, it is necessary to create an accurate 3D model to determine the position of the center of mass (COM). Impossible postures can thus be eliminated. Here, I test the hypothesis that Plateosaurus was a facultative bipedal, using four-legged gaits for slow speed, while running bipedally. I used two possible basic postures of Plateosaurus engelhardti in various variations and assessed the position of the center of mass (COM) for a spread of mass distributions for each posture. If only the first posture, a quadrupedal stance, results in a stable and probable posture, it is safe to assume that Plateosaurus was an obligate quadruped. The second posture, bipedal, would indicate an obligate biped if it is the sole feasible posture. If both postures are possible, at similar or different walking speeds, it is probable that Plateosaurus may indeed have been capable of both locomotory modes.

NOTE: Different researchers have articulated the bones of Plateosaurus differently, both in museum mounts and in reconstruction drawings. Those reconstructions that have a bearing on the investigations described here will be discussed below with the details of the articulation of the digital skeleton of Plateosaurus.