Walking with Whales.Pdf

Home , Ambulocetus, Archaeoceti, Diacodexis, Dorudon, Pakicetus, Phenacodus

news and views Walking with whales Christian de Muizon

Whales must have evolved from land-based mammals, but fossil evidence of some of the steps in between has been patchy. Newly discovered skeletons with legs fill in the gaps.

n page 277 of this issue, Thewissen and colleagues1 describe their discovery of u.p. Opartial fossil skeletons from the earliest cetaceans — a group of mammals that today consists of whales, porpoises and dolphins. The fossils, which are some 50 million years Diacodexis old and were found in Pakistan, take us a huge step forwards in understanding the origins and evolutionary relationships of Phenacodus Pachyaena whales. Until now, the limbs of all known u.p. early cetaceans reflected an amphibious or 2–6 wholly aquatic lifestyle . But the newly Pakicetus discovered fossils show that the first whales l.p. were fully terrestrial, and were even efficient runners. They also reveal that cetaceans are Pakicetus Diacodexis Sus more closely related to the oldest known even-toed ungulates — a group of hoofed Ambulocetus mammals that includes cows, hippos, pigs, Figure 2 Ankle bones (astragali) of hoofed 17 camels and giraffes — than to any other Pelvis and mammals and primitive whales. Phenacodus , hindlimb mammals. These conclusions are based on Dorudon a primitive ungulate, had an unspecialized solid anatomical data, and contradict the ankle bone that resembles that of Pachyaena17, previous hypotheses of both palaeontolo- Pelvis and a mesonychian ungulate from 50 million years gists and molecular biologists7–10. hindlimb Balaena ago. The double-pulleyed astragali of Pakicetus All of the mammals that existed in the early (one of the fossil cetaceans described by Tertiary — some 65–50 million years ago — Figure 1 The evolutionary route to a whale. Thewissen et al.1), Diacodexis18 (the oldest lived on land. So it has always been clear that Diacodexis13 was a primitive even-toed ungulate known even-toed ungulate), and Sus17 (the pig) aquatic cetaceans must have evolved from ter- (hoofed mammal); Pakicetus is one of the indicate a close relationship between these restrial mammals and returned to the water, terrestrial cetaceans described by Thewissen species. Bones are not shown to scale. u.p., upper and the forelimbs of recent cetaceans still have et al.1; Ambulocetus14 was amphibious; Dorudon pulley, articulates with tibia. l.p., lower pulley, the same general pattern as that of land mam- (modified from ref. 15) was a fully aquatic articulates with distal ankle bones. mals. However, modern cetaceans are highly archaeocete (early cetacean), but retained an specialized, with numerous adaptations that articulated elbow and vestigial hindlimbs; and been described. Cetaceans from the middle allow them to swim, dive and feed. Their bod- Balaena is a recent whale16. Skeletons are not Eocene, 45 million years ago, have been ies are elongated and streamlined; the shape drawn to scale. known for more than a century11, but here, of the tail is modified into a propulsive fluke; too, few limb remains were discovered, until and the limbs are much reduced in size. The What was the evolutionary path from the about 20 years ago. Since then, numerous forelimbs are flippers with rigid elbows and earliest land mammals to modern, aquatic fossils from North America, Pakistan and an increased number of phalanges (part of the whales? Pakicetids, which existed in the early Egypt have revealed that these early cetaceans finger bone), and function mainly in steering Eocene epoch, some 50 million years ago, had mobile elbows and external hindlimbs and stabilizing. The hindlimbs and pelvis are are the earliest known cetaceans. But, until with articulated knees2–7 (Fig. 1). However, internal, and have no role in swimming. now, only the skulls of these animals had they were already fully aquatic, except for Box 1Further finds from Pakistan

A report in this week’s previously known genus described Ambulocetus, life, the animals, which artiodactyls and reaffirm the Science by Gingerich et al.19 Rodhocetus. Both belonged the newly discovered would have weighed close relationship between fleshes out the picture of to the family Protocetidae, protocetids had well- between 400 and 500 artiodactyls and whales. early whale evolution with a group of extinct whales developed limbs with large kilograms, might have However, the absence of a description of two new somewhat more aquatic in hands and feet. Gingerich lived and moved like a formal phylogenetic species from 47-million- their adaptations than the et al. suggest that although large sea lions. The most analysis means that the year-old rocks in Pakistan. pakicetids described these creatures could have striking features of these researchers were unable One represents a new elsewhere in this issue by moved about on land, the protocetids are their to test the competing genus; the other, a close Thewissen and colleagues1. limbs would have astragali, which look hypotheses of whale relative, is a species of the Like the previously functioned as paddles. In remarkably like those of relationships. Henry Gee

Ambulocetus, which was amphibious — no one recent artiodactyl family is more National d’Histoire Naturelle, Paris 75005, France. much like sea lions. So there was no detailed closely related than another to cetaceans — e-mail: [email protected] information about the anatomy of the in other words, hippos are not the extant 1. Thewissen, J. G. M., Williams, E. M., Roe, L. J. & Hussain, S. T. cetaceans’ terrestrial ancestor. sister group of the cetaceans. The closest Nature 413, 277–281 (2001). 2. Gingerich, P. D., Smith, B. H. & Simons, E. L. Science 249, Recent cetaceans are very different to fossil relatives of the cetaceans were probably 154–157 (1990). other mammals, so another question that has the earliest known artiodactyls, such as 3. Gingerich, P. D. et al. Nature 368, 844–847 (1994). dogged this field is that of which group of Diacodexis (see Fig. 1). 4. Thewissen, J. G. M. et al. Science 263, 210–212 (1994). 1 5. Thewissen, J. G. M., Madar, S. I. & Hussain, S. T. Cour. mammals contains their closest relatives — Thewissen et al.’s discovery of these ter- Forschungsinst. Senckenberg 191, 1–86 (1996). which is their ‘sister group’? The cranial and restrial cetaceans is one of the most impor- 6. Uhen, M. D. in The Emergence of Whales: Evolutionary Patterns skeletal anatomy of cetaceans is highly modi- tant events in the past century of vertebrate in the Origin of Cetacea (ed. Thewissen, J. G. M.) 29–61 (Plenum, New York, 1998). fied compared with that of land mammals, palaeontology. Only a very few fossils, such 7. Geisler, J. H. & Luo, Z. in The Emergence of Whales: and fossils of early cetaceans are so rare and as these, reveal a link between two groups of Evolutionary Patterns in the Origin of Cetacea (ed. Thewissen, generally incomplete, that the affinities of vertebrates that are hugely different in terms J. G. M.) 163–212 (Plenum, New York, 1998). the group are difficult to establish. On the of shape yet closely related in terms of evolu- 8. O’Leary, M. A. & Geisler, J. H. Syst. Biol. 48, 455–490 (1999). 9. Shimamura, M. et al. Nature 388, 666–670 (1997). basis of tooth and ear morphology, palaeon- tion. When there is a drastic shift in habitat — 10.Nikaido, M., Rooney, A. P. & Okada, N. Proc. Natl Acad. Sci. tologists contend that cetaceans are most such as from land to water — the morphology USA 96, 10261–10266 (1999). closely related to the mesonychians7,8 — a of the newly adapted animals is generally so 11.Kellogg, R. A Review of the Archaeoceti (Carnegie Inst., Washington, DC, 1936). group of extinct ungulates from the early greatly modified, because of the high selec- 12. Montgelard, C., Catzeflis, F. M. & Douzery, E. Mol. Biol. Evol. Tertiary. But molecular biologists favour tive pressure, that any resemblance to the 14, 550–559 (1997). hippos — which form one of the families original ancestor is quickly obliterated. But 13.Janis, C. M., Scott, K. M. & Jacobs, L. L. (eds) in Evolution of 1 Tertiary Mammals of North America p. 360 (Cambridge Univ. of modern even-toed ungulates (artiodac- the new fossils superbly document the link Press, 1998). 9,10 tyls) — as the sister group. between modern whales and their land- 14.http://www.neoucom.edu/Depts/Anat/Ambulocet.html Thewissen and colleagues’ discovery1 based forebears, and should take their place 15.Gingerich, P. D. & Uhen, M. D. Contrib. Mus. Palaentol. Univ. allows us to address both of these problems. among other famous ‘intermediates’, such as Michigan 29, 359–401 (1966). 16.Eschricht, D. F. & Reinhardt, J. in Recent Memoirs on the The newly found fossils include several skulls the most primitive bird, Archaeopteryx, and Cetacea (ed. Flower, W. H.) 1–150 (Ray Soc., London, 1866). and postcranial bones from two early the early hominid Australopithecus. ■ 17.O’Leary, M. A. & Rose, K. D. J. Vert. Paleontol. 15, 401–430 pakicetid species — which, it seems, had the Christian de Muizon is at the Unité Mixte de (1995). 18.Schaeffer, B. Am. Mus. Nov. 1356, 1–24 (1947). head of a primitive cetacean (as indicated Recherche 8569, Centre National de la Recherche 19.Gingerich, P. D., ul Haq, M., Zalmout, I., Khan, I. H. & by the ear region) and the body of an artio- Scientifique, Laboratoire de Paléontologie, Muséum Malkani, M. S. Science 293, 2239–2242 (2001). dactyl. All the postcranial bones indicate that pakicetids were land mammals, and it is likely that they would have been thought of Surface physics as some primitive terrestrial artiodactyl if they had been found without their skulls. Many of the fossils’ features — including the A new crack at friction length of the cervical vertebrae, the relatively David A. Kessler rigid articulations of the lumbar vertebrae, and the long, slender limb bones — indicate One of the dirty little secrets of physics is that there is no generally that the animals were runners, moving with accepted explanation of the basic laws of friction. An advance in the only their digits touching the ground. theory of cracks will stimulate fresh thinking on the question. But the most eloquent information pro- vided by the fossils1 comes from the ankle riction is a ubiquitous feature of every- 285 of this issue, Gerde and Marder1 offer bones, particularly the astralagus (Fig. 2). day life. Without it, we couldn’t walk, a theory of surface cracks that may lead to a This has two pulleys, which connect to both Ftyres wouldn’t roll, and ballpoint pens better understanding of surface friction. the tibia and the more distal ankle bones and would fail to write. But what is friction, and The standard picture of friction2 (dating allow a great deal of flexibility. This type of how does it act? from as recently as the 1960s) is that the solid morphology is an adaptation for running. It The basic properties are simple to grasp. surfaces are not really planar, but are rough was once thought to be unique to artio- To move a solid object from rest on top of on a microscopic scale. The presence of these dactyls, but it is now clear that it also a solid surface, a minimum force has to be tiny surface features, or asperities as they occurred in cetaceans. So, for the first time, applied to overcome the force of friction. are known, prevents the surfaces from com- morphological evidence shows that artio- This force is proportional to the compressive ing into full contact. So the true contact area dactyls are the closest relatives of the force pushing the two surfaces together, in is much smaller than its apparent value, cetaceans (also see Box 1). this case the weight of the object. Intriguingly, and is proportional to the compressive force This means that artiodactyls and cetaceans this minimum force is independent of the between the surfaces, in much the same way form the two branches of a larger group, the area of contact between the body and the that the contact area between a car tyre and cetartiodactyls12. Thewissen and colleagues’ surface. So the friction force on a rectangular the road increases when you load your car. studies also exclude the mesonychians from solid resting on a table is the same which- Problems have arisen when physicists tried this larger group, in part because these ungu- ever face is in contact with the surface. These to confirm this picture using calculation lates do not have a double-pulleyed astra- laws have been known since the mid 1700s from first principles. The goal is to construct, galus. Mesonychians are an unspecialized and are attributed to the French physicists either analytically or on the computer, a group of primitive ungulates, and perhaps Guillaume Amontons (1663–1705) and solid body and surface from atoms with some of them should be included in other Charles Augustin de Coulomb (1736–1806). prescribed interactions, and calculate the groups of mammals. Cetaceans and some It is one of the dirty little secrets of physics friction force directly. But previous attempts mesonychians have dental similarities and that while we physicists can tell you a lot at this found that the two surfaces ride freely an elongated skull, but these features are about quarks, quasars and other exotica, on top of each other because of the mismatch probably the result of convergent evolution. there is still no universally accepted expla- between the asperities on the two surfaces, so The authors’ analysis also indicates that nation of the basic laws of friction. On page there is no friction.