Why Archaeopteryx did not run over water

Ming San Ma, Whee Ky Ma, Ilja Nieuwland and Ronald R. Easley Why Archaeopteryx did not run over water Zusammenfassung In der 2000-Nummer von Archaeopteryx schlägt der Marinebiologe John Videler ein neues Modell für die Erwerbung des Vogelfluges vor, wobei die Vorfahren von Archaeopteryx lithographica über das Wasser liefen in einer den modernen Basilisken oder Jesus-Christus-Eidechsen (Basiliscus basiliscus) ähnlichen Art. In diesem Szenario konnte größere evolutionäre Konkurrenzfähigkeit durch die Entwicklung von Flügeln erreicht werden. Eine biomechanische Analyse zeigt aber, daß der von Basiliscus genutzte Mechanismus nicht auf Archaeopteryx angewandt werden kann. Dazu fehlen bei Archaeopteryx einige essentielle anatomische Bedingungen für das Laufen über Wasser. Weiterhin ist es unwahrscheinlich, daß die Umstände, die in der Solnhofener Gegend vorherrschten, die Archaeopteryx zwangen, auf das Wasser zu fliehen. Letztendlich basiert Videlers Bezeichnung einer Trennungsfunktion der Krallen sich auf einer Fehlinterpretation. Abstract In the 2000 issue of Archaeopteryx, the marine biologist John Videler proposes a new model for the acquisition of flight in birds, in which the ancestors of Archaeopteryx lithographica were able to run over water in a way similar to the modern basilisk or Jesus Christ lizard (Basiliscus basiliscus). In this scenario, greater fitness might have been achieved through the development of wings. However, a biomechanical analysis shows that the mechanism exploited by Basiliscus cannot have been applied by Archaeopteryx. Moreover, Archaeopteryx appears to lack several crucial locomotional and anatomical conditions for water running. The scenario implicitly suggests an unlikely reverse development of modern avian characteristics out of a water-running final stage. Furthermore, the Solnhofen environment did probably not provide enough selective pressure for Archaeopteryx to flee onto the water. Finally, Videler s allocation of wing fence function to Archaeopteryx finger nails is based on misinterpretation. Biomechanics The biomechanics of Basiliscus were described by Glasheen and McMahon (GLASHEEN & MCMAHON 1996a, 1996b). They model the movement of its foot through the water. The movement consists of three phases: slap, stroke, and protraction. First, the foot swings through the air and slaps the water surface; this causes an upward force. Then it strokes downwards, creating an air cavity; this yields two drag forces, one due to the inertia of the water and the other due to the hydrostatic pressure difference between the air cavity and the water. During the stroke, the foot is kept perpendicular to its direction of motion. Finally, the foot is 1

retracted from the air cavity (protraction), minimising the friction with the water. It is necessary that the time between two subsequent steps is shorter than the time in which the air cavity closes. There is rarely a moment when either two feet or no feet are in the water. Together, the slap and stroke can produce the support needed to prevent an adult lizard from sinking. Moreover, it was found that the larger the animal is, the larger the part of the impulse that it has to obtain from the stroke. A rough estimate of the quantities involved readily shows that Archaeopteryx can never have employed this mechanism; for its ancestors, the calculation will be comparable. If T step is the time between two subsequent steps, M is the body mass, and g is the gravitational acceleration, the slap and stroke impulses should satisfy p + p MgT (1) slap stroke step to allow for water-running. It turns out that the stroke impulse can be expressed as (GLASHEEN & MCMAHON 1996a) 1 Ê gh ˆ p stroke = 2 CSrhÁu +, (2) Ë u where C is a constant coefficient (the drag coefficient), S is the area of the foot, u is the root-mean-square value of the downward velocity of the middle of the foot, h is the largest depth of the middle of the foot, and r is the density of water. In Table 1 we list the values of the parameters for a typical lizard and for the (subadult) Berlin specimen of Archaeopteryx. Basiliscus Archaeopteryx S (cm 2 ) 6.1 10 h (cm) 7.6 10 M (g) 90 260 T step (s) 0.06 0.10 C 0.70 0.70 u (m/s) 2.5? Table 1: Relevant parameters for a typical lizard and for Archaeopteryx (estimated). In lizards, C was found to be almost constant, regardless of size, speed or cavity depth. The value of u for the lizard is an upper limit. Values for Basiliscus are adopted from GLASHEEN & MCMAHON 1996b. For Archaeopteryx, the values for S and M are based on PAUL 1988, h is half of the hind leg, and the other values are estimates. We made rather favourable estimates for Archaeopteryx: it is unlikely that it could make 10 steps a second, and a drag coefficient as high as 0.70 requires special adaptations. To get an indication of the minimum u needed for Archaeopteryx to walk on water, we use the allometric formula from (GLASHEEN & MCMAHON 1996a) for the maximum slap impulse relative to the minimum needed impulse with T step = 0.10 s, pslap,max -0.244 = 0.424M = 0.11 MgT step (M in grams); there is no reason why Archaeopteryx would do significantly better on this. 1 It follows from Eq. (1) that p stroke 0.89 MgT step = 0.23 kg m/s, and with Eq. (2), 1 If we assume instead that the ratio between the slap velocity and u is similar to this ratio for the 2

u + u gh 6.5 m/s. This gives u = 6.3 m/s, which is much higher than the lizard s upper limit; its absurdity is even more evident when we calculate the average power needed for the stroke: = 1 CSru( u 2 + gh) = 91 W. P stroke 2 There is no evidence at all that Archaeopteryx was able to generate such an immense power. For comparison, the stroke power is 3.7 W for the lizard in our example. Anatomy In the former we assumed that Archaeopteryx was able to create an air cavity. However, it remains to be seen whether the animal possessed the necessary anatomical conditions to do so. Firstly, the position of the art. coxae and the morphology of the femur in lizards of the genus Basiliscus causes the animals to have their legs oriented laterally from their bodies, a condition which originates in the spread gait of the animal s ancestors. In Basiliscus, this condition has allowed it to develop its unique way of locomotion, namely swinging the leg forward through the air between protraction and slap. Such a movement cannot be applied in the erect gait of Archaeopteryx or any other maniraptoran dinosaur, making it impossible for the animal s foot to slap the water with its foot parallel to the water surface. It is therefore highly unlikely that an air cavity could thus be created. Secondly, apart from the mechanism of locomotion, the foot of Basiliscus possesses fringes, which flare out when the foot is moved downward, increasing the effective surface of the foot considerably. This adaptation appears to be essential in creating an air cavity (GLASHEEN & MCMAHON 1996a). No indications for such anatomical faculties are present in Archaeopteryx. Instead, the reversed hallux of Archaeopteryx would not only have been useless but even a disturbing factor in creating and maintaining an air cavity. Thirdly, even if Archaeopteryx would have been able to form an air cavity, the animal could probably not have retracted its foot efficiently. To minimise the waterinduced friction during protraction, the foot needs to be retracted out of the air cavity while being fully stretched with respect to the lower leg, making the angle between them practically 180, as can be observed in the basilisk. However, when comparing the morphology of the hind legs of Archaeopteryx and Basiliscus, a crucial difference can be noticed. In Archaeopteryx, the orientation of the joints of tibia, fibula and metatarsals with the tarsals reduce the maximum angle between lower leg and foot to about 140, after which it blocks. This condition can be witnessed both in the small theropod ancestors of Archaeopteryx and in most modern birds (between tibiotarsus and tarsometatarsus). Additionally, in birds the foot is dorsally flexed during this movement, making the angle even tighter. Consequently, the air cavity would be disrupted by the withdrawal of the foot (fig. 1), leading to an enormous amount of friction. In fact, the stretching of the basilisk s foot is a special adaptation for running on water. Finally, in addition to the problems Archaeopteryx would have faced due to the anatomy of the hind limbs when running over water, its locomotion would have been further obstructed by water absorption and friction caused by its feathers. In basilisk, and take the mass of the foot to be of the order of 10 g, then the results are even more unfavourable. 3

contrast with the smoothly skinned basilisk lizard, Archaeopteryx was almost entirely feathered. We assumed that the bird would sink about 10 cm into the water, which means at least up to its upper legs. Its wings and tail would be very close to the surface, like the rest of its plumage, increasing its mass by absorbing splashing water. The hind limbs are feathered up to the metatarsus which would cause an increase of body mass by absorption as well as an unacceptably high amount of friction. Fig. 1 A) Hind limb of Basiliscus. The foot is retracted from an air cavity in such a way that the long axis of the foot is parallel to the direction of movement (indicated with 1). B) Hind limb of Archaeopteryx. Supposing that Archaeopteryx was able to create an air cavity, it would have been disrupted by fact that the ankle could not have been fully stretched (long axis not parallel to the direction of movement). Additional disruption is caused by dorsal flexion of the foot during movement (indicated with 2) which is absent in Basiliscus. Nails Videler s paper also treats the function of Archaeopteryx s claws. He asserts that the last phalangeal joints show features that escaped the attention of earlier investigators, and goes on to suggest that the claws could be positioned in an alternative socket (VIDELER 2000, 28-30). This phenomenon has been observed before by Dames in his classic description of the Berlin specimen of Archaeopteryx (DAMES 1884). Also, it is important to realise that a similar condition can be found in theropod dinosaurs, such as Deinonychus antirrhopus (OSTROM 1969, GISHLICK 2001). Even though it may be true that neither [the arboreal nor the cursorial] scenario [for the origin of flight] offers a convincing function [ ] for the unworn horny sheaths of the finger, it remains unclear what the relevance of this observation to the Jesus Archaeopteryx hypothesis might be. In fact, why would one expect the claws to show marked wear when it is far more likely that, as in modern birds, the horny sheaths of these claws would have prevented damage to the bone itself? Aerodynamics However, Videler does present a second hypothesis, in which Archaeopteryx claws become important, although there is no obvious connection to the first, waterrunning hypothesis. This second part attempts to deal with Archaeopteryx 4

adaptations for flapping flight. Great significance is attached to the lack of a m. supracoracoideus and triosseal canal, leading to the conclusion that Archaeopteryx could not supinate its wing quickly enough to permit the vortex-ring gait needed for slow flight. First of all, the meaning of slow flight in this context remains undefined. When experimental subjects like pigeons and magpies fly slowly, they use a vortexring gait, the same gait as is used in near-vertical takeoff. Most likely, this is a derived set of adaptations and behaviours. Also, when a pigeon switches from a continuous-vortex gait to a vortex-ring gait in preparation for landing, it is likely doing so to decelerate quickly while using direct thrust to maintain proper orientation for a precise, controlled landing. The supposed inability to supinate the wing or raise it quickly does not necessarily mean that Archaeopteryx was incapable of slow flight. It was just incapable of slow flight using the vortex-ring gait. Probably, it could not producing sufficiently high angles of attack during the wing stroke that would promote flow separation, and would therefore have no need for anything like a wing fence. The strict definition of a wing fence is an aerodynamic structure that extends along nearly the entire chord of the airfoil on that point and physically separates the airflow between two regions of the wing. Such a structure is unlikely to be present on Archaeopteryx. One interesting aspect of Videler's hypothesis is his suggestion of the hand wing acting as a delta wing. One advantage of the delta platform that may be useful to slow-flying birds is that a delta wing produces a strong, stabilizing leading-edge vortex (HUENECKE 1987). At the cost of a slight penalty in drag, these leading-edge vortex flows confer a more gradual lift loss with increasing angles of attack and give more predictable stall characteristics (WHITFORD 1987). However it is difficult, if not impossible, to translate these characteristics of a fixed wing into that of a flapping wing and expect the same results to occur. This can only be examined by preceding theoretical and experimental work exploring this problem. No such work has been done. Also, the "hand wing" of most birds is composed of separated primary feathers and may not act like a fixed delta wing under any set of circumstances. While this portion of Videler's argument may be worthy of further investigation, at this time it remains speculative at best. Ecology Furthermore, a number of ecological problems present themselves. Chief among these is the motivation that any Archaeopteryx might have felt to change its secure terrestrial surroundings for the water surface. First of all, it is unlikely that it was to find additional sources of nutrition there (BARTHEL et al. 1990, 73). Rather, it would have moved from an environment in which food was in ample supply to one in which it would have been very hard to find. But more importantly, the animal appears not to have had any reason to seek security on the water. Archaeopteryx was at the top of the Solnhofen food chain, and might have put himself at unnecessary risk rather than in security by leaving the shores, especially since it was the water that contained most of Solnhofen s predators. 2 Also, one may safely question whether the Solnhofen waters were suitable to run over. The presence of life in the upper layers of the hypersaline Solnhofen lagoon indicates that the necessary oxygen was brought in by surface turbulence (VIOHL, pers. comm.). Even though there was no direct connection to the Tethys ocean, and the coral reefs may have given protection from the worst, there is still a 2 We appreciate the presence of Compsognathus longipes, but we do not consider Archaeopteryx as a prey for that animal, considering their comparable adult sizes. 5

considerable difference between this surface and the stagnant ponds that Basilisks run over (BARTHEL et al. 1990, 56-8). This, in turn, will almost certainly have had detrimental effects for the water-running mechanism. Evolution Following the ecological arguments pointed out earlier, it can be stated that the environmental conditions probably did not provide enough selective pressures to force Archaeopteryx onto the water. The water-running of Basiliscus is an aspect of a highly specialized ecological niche in which the animal operates, and hence can be seen as an evolutionary final stage. It is not likely that a species, which is optimally adapted to a given environment, were to choose quite a different evolutionary path (development of flapping flight) unless being forced to do so by radically changing environmental conditions, for which there are no indications in the case of Archaeopteryx (BARTHEL ET AL. 1990). Although we demonstrated the supposed inability of Archaeopteryx foot to create an air cavity, an evolutionary scenario could be proposed in which a direct ancestor of Archaeopteryx demonstrated basilisk-like features. However, the anatomy and morphology of the feet in the theropod ancestors of Archaeopteryx show several shared characteristics with those in modern birds. This means that the scenario should be one in which the bird-like hind limb anatomy of Archaeopteryx ancestors evolved into Basilicus-like structures, and, having acquired the ability to fly, subsequently changed back to modern bird-like characteristics (Archaeopteryx). This suggests an highly implausible reverse-evolutionary path. Conclusion On the basis of all the above arguments one can safely state that Archaeopteryx kept on land (or in the air). Although the anatomy of the ancient bird does not allow a similar way of locomotion as the one exploited by the highly specialised Basiliscus, a certain theoretical possibility exists that Archaeopteryx made use of another water-running mechanism, as do some present-day water fowl. However, the ability of the latter to walk short distances on water seems to be largely dependent on wing-generated lift and the increased surface of the foot. Still, every anatomical adaptation needed for any mechanism has to face the unlikeliness of subsequent reverse evolution, and still the question remains how Archaeopteryx benefited by switching from a relatively profitable niche to a relatively unprofitable one. Since the fingernails of Archaeopteryx cannot have been exploited as wing fences as claimed, they do not seem relevant in the present discussion. The debate on bird flight remains a thorny and complex issue, both from a scientific and from a sociological point of view, and we believe Videler s article does not do it justice. His suggestion that opinion has always been divided into two opposing paradigms misrepresents the diversity and historicity of the debate about the origin of bird flight (SHIPMAN, 1998). There is certainly a large number of unresolved issues, but that does not justify poorly construed attempts at finding alternatives. In the words of Carl Sagan, extraordinary claims require extraordinary evidence. Unfortunately, that is not presented in Videler s article. Acknowledgements We wish to thank Bert Boekschoten, Jim Glasheen, John Hutchinson, Kevin Padian, Anne Schulp, Marco Signore, and Günter Viohl for pleasant and useful discussions on Basiliscus, Archaeopteryx and the Solnhofen environment. 6

References BARTHEL, K.W., CONWAY MORRIS, S., SWINBURNE, N.M. (1990): Solnhofen. Cambridge. DAMES, W. (1884): Über Archaeopteryx. Palaeontologische Abhandlungen 2/3. Berlin. GISHLICK, A.D. (2001): The Function of the Manus and Forelimb of Deinonychus Antirrhopus and its importance for the origin of avian flight. In: GAUTHIER, J. & GALL, L.F., eds., New Perspectives on the Origin and Early Evolution of Birds: Proceedings of the Interantional Symposium in Honour of John H. Ostrom (New Haven), 301-318. GLASHEEN, J.W. & MCMAHON, T.A. (1996a): A Hydrodynamic Model of Locomotion in Basilisk Lizard. Nature 380, 340-342. GLASHEEN, J.W. & MCMAHON, T.A. (1996b): Size-dependence of Water-Running Ability in Basilisk Lizards (Basiliscus basiliscus). The Journal of Experimental Biology 199, 2611-2618. HUENECKE, K. (1987): Modern Combat Aircraft Design. Annapolis. OSTROM, J.H. (1969): Osteology of Deinonychus antirrhopus, an unusual Theropod from the Lower Cretaceous of Montana. Peabody Museum Bulletin 30. PAUL, G. (1988): Predatory Dinosaurs of the World. New York. PADIAN, K. & CHIAPPE, L.M. (1998): The origin and early evolution of birds. Biological Review of the Cambridge Society 73, 1-42. SHIPMAN, P. (1998): Taking Wing. Archaeopteryx and the evolution of bird flight. New York. WHITFORD, R. (1987): Design for Air Combat. London. Correspondence addresses: MING SAN MA, c/o G.J. Boekschoten, Department of Earth and Life Sciences, Free University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands; m.s.ma@student.rug.nl Dr. WHEE KY MA, Groningen University, Center for Theoretical Physics, Nijenborgh 4, 9747 AG Groningen, The Netherlands; w.k.ma@phys.rug.nl ILJA NIEUWLAND, Groningen University, History Department, P.O. Box 716, 9700 AS Groningen, The Netherlands; nieuwland.ilja@berlin.de RONALD R. EASLEY, 6536 Markley Way, Carmichael, CA 95608-6306, U.S.A; rre@lanset.com 7