ON THE RECONSTRUCTION OF PTEROSAURS AND THEIR MANNER OF FLIGHT, WITH NOTES ON VORTEX WAKES

Classical pterosaur reconstructions are variants on a ‘bat-analogy’, whereby the wing is conceived as a simple membrane with no inherent bending strength, stretched between the arm and leg skeletons. The legs are considered to be splayed out to the sides, as in bats, so that the animal would have to adopt a quadrupedal stance on the ground, supported on its feet and the metacarpo-phalangeal joints. In recent years an alternative ‘bird-analogy’ has come to be generally accepted. This hypothesis, most elements of which are due to Padian (1983 a, b) calls for the animal to stand upright on its legs like a bird. The wings are independent of the legs, as in birds, are stiffened by skeletal fibres in the membrane, and have a very narrow, sharply pointed shape. There are difficulties in reconciling the bird-analogy with the evidence. The long-tailed rhamphorhynchs might conceivably have balanced their weight about their hip joints but this would not have been possible for the short-tailed pterodactyls. The bird pelvis shows modifications which permit bipedal standing in spite of the reduction of the tail, but no equivalent adaptations are seen in pterodactyls. Besides, all known pterosaur pelvises, except that of the giant pterodactyl Pteranodon were open ventrally, which would have precluded the legs from being brought to a parasagittal position, as required for bipedal walking. The notion that the wing was not attached to the legs is based on negative evidence, in that no clear impressions of the inner end of the wing membrane are preserved in the fossils. However one pterodactyl fossil shows a membrane edge approaching the ankle joint. In fossils that are preserved with the wings forward, the legs have been pulled forwards by the ankles. A tendon connecting the ankle to the wing tip is consistent with the evidence. The ‘fibres’ in the wing membranes are actually impressions of surface ridges, with no internal structure, and are better interpreted as surface wrinkles in the skin, caused by contraction of elastic fibres within the membrane. The bird analogy also results in a very unsatisfactory wing from an aerodynamic point of view. The structure of an animal wing is best understood in terms of the type of vortex wake it is adapted to generate. Hummingbirds, and insects capable of economical hovering, have wings that can be inverted on the upstroke, and when hovering, generate a wake consisting of two vortex rings per wingbeat cycle. The span of such wings is fixed, which implies that they create a ‘ladder wake’ in cruising flight, consisting of a pair of undulating wing-tip vortices, joined by a transverse vortex at each transition from downstroke to upstroke and back. Normal birds cannot invert their wings, and so are less efficient in hovering, but they can shorten the wing during the upstroke in cruising flight. This creates a ‘concertina wake’, with no transverse vortices. Hummingbirds show very limited migration performance, compared with normal birds, with the implication that a wing capable of creating a concertina wake is more economical in cruising flight than one creating a ladder wake, and is an essential adaptation for long-distance migration. A revised reconstruction of the pterosaur wing starts from the observations that, contrary to the currently popular bird-analogy, pterosaurs were not bipedal, their wings did not contain stiffening fibres but did contain elastic fibres, and the trailing edge of the membrane was supported by a tendon joining the tip of the wing finger to the ankle. A hypothetical arrangement of elastic fibres, that accounts well for the observed pattern of wrinkles in contracted wings, also allows the planform shape of the wing to be adjusted in much the same way as seen in birds, although using a completely different mechanism. It opens the possibility that pterosaurs could fly with a concertina wake, and thus could have been long-distance migrators like modern birds. Although this hypothetical wing is mechanically somewhat bat-like, it is not a return to the classical bat-analogy. It would not have the high degree of control over profile shape, which gives bats their outstanding manoeuvrability. On the other hand bats do not have the degree of control over their wingspan that is suggested here for pterosaurs, and consequently are not notable for migration performance.