New Look at Human Evolution Copyright 2003 Scientific American, Inc

ORIGINS early hominid fossils from AFRICA The year was 1965. Bryan Patterson, a paleoanthropologist ern humans than the one other Australo- Yet Patterson’s fossil would eventu- from Harvard University, unearthed a pithecus humerus known at the time. ally help establish the existence of a new fragment of a fossil arm bone at a site And yet the age of the Kanapoi fossil species of Australopithecus—the oldest called Kanapoi in northern Kenya. He proved somewhat surprising. Although yet to be identified—and push back the and his colleagues knew it would be hard the techniques for dating the rocks where origins of upright walking to more than to make a great deal of anatomical or the fossil was uncovered were still fairly four million years ago. But to see how evolutionary sense out of a small piece of rudimentary, the group working in Ken- this happened, we need to trace the steps elbow joint. Nevertheless, they did rec- ya was able to show that the bone was that paleoanthropologists have taken in ognize some features reminiscent of a probably older than the various Austra- constructing an outline for the story of ) species of early hominid (a hominid is lopithecus specimens that had previous- hominid evolution. right any upright-walking primate) known as ly been found. Despite this unusual result, and Australopithecus, first discovered 40 however, the significance of Patterson’s An Evolving Story years earlier in South Africa by Raymond discovery was not to be confirmed for an- SCIENTISTS CLASSIFY the immediate center Dart of the University of the Witwater- other 30 years. In the interim, researchers ancestors of the genus Homo (which in- srand. In most details, however, Patterson identified the remains of so many impor- cludes our own species, Homo sapiens) and his team considered the fragment of tant early hominids that the humerus in the genus Australopithecus. For sev- arm bone to be more like those of mod- from Kanapoi was rather forgotten. eral decades it was believed that these ancient hominids first inhabited the AUSTRALOPITHECUS earth at least three and a half million ANAMENSIS (right) lived roughly four million years ago. The specimens found in South years ago. Only a few Africa by Dart and others indicated that anamensis fossils have there were at least two types of Austra- been found—the ones lopithecus—A. africanus and A. robus- shown at the left tus. The leg bones of both species sug- include a jawbone and ); ALAN WALKER; © NATIONAL MUSEUMS OF KENYA ( part of the front of the gested that they had the striding, bipedal left face (left), parts of an locomotion that is a hallmark of humans arm bone (center) and among living mammals. (The upright fragments of a lower leg posture of these creatures was vividly bone (right)—and thus confirmed in 1978 at the Laetoli site in researchers cannot determine much about Tanzania, where a team led by archae- ); ROBERT CAMPBELL ( the species’ physical ologist Mary Leakey discovered a spec- appearance. But tacular series of footprints made 3.6 mil- scientists have lion years ago by three Australopithecus illustration established that individuals as they walked across wet anamensis walked upright, making it the volcanic ash.) Both A. africanus and A. earliest bipedal creature robustus were relatively small-brained

MATT MAHURIN ( yet to be discovered. and had canine teeth that differed from

14 SCIENTIFIC AMERICAN UpdatedNEW LOOK from AT the HUMAN June 1997EVOLUTION issue COPYRIGHT 2003 SCIENTIFIC AMERICAN, INC. A new species of Australopithecus, the ancestor of Homo, pushes back the origins of bipedalism to some four million years ago By Meave Leakey and Alan Walker

those of modern apes in that they hard- of hominid bones and teeth discovered ly projected past the rest of the tooth at Laetoli, as well as a large and very im- row. The younger of the two species, A. portant collection of specimens from the robustus, had bizarre adaptations for Hadar region of Ethiopia (including the chewing—huge molar and premolar famous “Lucy” skeleton). The group teeth combined with bony crests on the named the new species afarensis. Radio- skull where powerful chewing muscles metric dating revealed that the species would have been attached. had lived between 3.6 and 2.9 million Paleoanthropologists identiﬁed more years ago, making it the oldest Aus- species of Australopithecus over the next tralopithecus known at the time. several decades. In 1959 Mary Leakey This early species is probably the best unearthed a skull from yet another East studied of all the Australopithecus rec- African species closely related to robus- ognized so far, and it is certainly the one tus. Skulls of these species uncovered that has generated the most controversy during the past 45 years in the north- over the past 30 years. The debates have eastern part of Africa, in Ethiopia and ranged over many issues: whether the Kenya, differed considerably from those afarensis fossils were truly distinct from found in South Africa; as a result, re- the africanus fossils from South Africa; searchers think that two separate robus- whether there was one or several species tus-like species—a northern one and a at Hadar; whether the Tanzanian and southern one—existed. Ethiopian fossils were of the same spe- In 1978 Donald C. Johanson, now at cies; and whether the fossils had been the Institute of Human Origins at Ari- dated correctly. zona State University, along with his col- But the most divisive debate con- leagues, identiﬁed still another species of cerns the issue of how extensively the Australopithecus. Johanson and his bipedal afarensis climbed in trees. Fossils team had been studying a small number of afarensis include various bone and

www.sciam.com SCIENTIFIC AMERICAN 15 COPYRIGHT 2003 SCIENTIFIC AMERICAN, INC. joint structures typical of tree climbers. BONOBO Some scientists argue that such characteristics indicate that these hominids CHIMPANZEE must have spent at least some time in the HOMO trees. But others view these features as Ardipithecus ? simply evolutionary baggage, left over Orrorin ramidus Au. afarensis from arboreal ancestors. Underlying this Au. robustus Au. africanus discussion is the question of where Aus- tralopithecus lived—in forests or on the Sahelanthropus open savanna. Au. anamensis Au. aethiopicus By the beginning of the 1990s, researchers knew a fair amount about the Au. boisei various species of Australopithecus and how each had adapted to its environ- 6 MYR 5 MYR 4 MYR 3 MYR 2 MYR 1 MYR mental niche. A description of any one of AGO AGO AGO AGO AGO AGO the species would mention that the creatures were bipedal and that they had ape- FAMILY TREE of the hominid Australopithecus (red) includes a number of species that lived between roughly 4 million and 1.25 million years (Myr) ago. Just over 2 Myr ago a new genus, Homo (which size brains and large, thickly enameled includes our own species, H. sapiens), evolved from one of the species of Australopithecus. teeth in strong jaws, with nonprojecting canines. Males were typically larger than 1982, expeditions run by the National River dominated the Turkana area for females, and individuals grew and ma- Museums of Kenya to the Lake Turkana much of the Pliocene (roughly 5.3 to 1.8 tured rapidly. But the origins of Aus- basin in northern Kenya began finding million years ago) and the early Pleis- tralopithecus were only hinted at, because hominid fossils nearly four million years tocene (1.8 to 0.7 million years ago). Only the gap between the earliest well-known old. But because these fossils were main- infrequently was a lake present in the species in the group (afarensis, from ly isolated teeth—no jawbones or skulls area at all. Instead, for most of the past about 3.6 million years ago) and the pos- were preserved—very little could be said four million years, an extensive river sys- tulated time of the last common ancestor about them except that they resembled tem flowed across the broad floodplain, of chimpanzees and humans (about six the remains of afarensis from Laetoli. proceeding to the Indian Ocean without million years ago, according to molecular But our excavations at an unusual site, dumping its sediments into a lake. evidence) was still very great. Fossil just inland from Allia Bay on the east The Allia Bay fossils are located in hunters had unearthed only a few older side of Lake Turkana [see maps on page one of the channels of this ancient river fragments of bone, tooth and jaw from 18], yielded more complete fossils. system. Most of the fossils collected the intervening 1.5 million years to indi- The site at Allia Bay is a bone bed, from Allia Bay are rolled and weathered cate the anatomy and course of evolution where millions of fragments of weath- bones and teeth of aquatic animals— of the earliest hominids. ered tooth and bone from a wide variety fish, crocodiles, hippopotamuses and the of animals, including hominids, spill out like—that were damaged during trans- Filling the Gap of the hillside. Exposed at the top of the port down the river from some distance DISCOVERIES IN KENYA over the hill lies a layer of hardened volcanic ash away. But some of the fossils are much past several years have filled in some of called the Moiti Tuff, which has been better preserved; these come from the the missing interval between 3.5 million dated radiometrically to just over 3.9 animals that lived on or near the river- and 5 million years ago. Beginning in million years old. The fossil fragments banks. Among these creatures are sever- lie several meters below the tuff, indi- al different species of leaf-eating mon- MEAVE LEAKEY and ALAN WALKER, to- cating that the remains are older than keys, related to modern colobus mon- gether with Leakey’s husband, Richard, the tuff. We do not yet understand fully keys, as well as antelopes whose living have collaborated for many years on the why so many fossils are concentrated in relatives favor closely wooded areas. discovery and analysis of early hominid this spot, but we can be certain that they Reasonably well preserved hominid fos- fossils from Kenya. Meave Leakey is a were deposited by the precursor of the sils can also be found here, suggesting researcher and former head of the divi- present-day Omo River. that, at least occasionally, early homi- THE AUTHORS sion of paleontology at the National Mu- Today the Omo drains the Ethiopian nids inhabited a riparian habitat. seums of Kenya in Nairobi. Walker is highlands located to the north, emptying Where do these Australopithecus Evan Pugh Professor of Anthropology into Lake Turkana, which has no outlet. fossils fit in the evolutionary history of and Biology at Pennsylvania State Uni- But this has not always been so. Our col- hominids? The jaws and teeth from Al- versity. He is a MacArthur Fellow and a leagues Frank Brown of the University of lia Bay, as well as a nearly complete ra- member of the American Academy of Utah and Craig Feibel of Rutgers Uni- dius (the outside bone of the forearm)

Arts and Sciences. versity have shown that the ancient Omo from the nearby sediments of Sibilot just SLIM FILMS

MANDIBLE

The human jaw widens at the back of the mouth

The jawbones in anamensis and chimpanzees are U-shaped

TIBIA

The top of the tibia, near the In the tibias of anamensis knee, is and humans, the top of the somewhat bone is wider because of T-shaped in the extra spongy bone chimpanzee tissue present, which serves as a shock absorber Primates such as in bipedal creatures chimpanzees that walk on their knuckles have a HUMERUS deep, oval hollow at the bottom of the ) humerus where the Human and humerus and the human anamensis ( ulna lock in place, bones lack this making the elbow feature, joint more stable suggesting that, like humans, anamensis did not walk on its

Photo Researchers, Inc. knuckles

FOSSILS from anamensis (center) share a number of features in common their interrelationships and thereby piece together the course of with both humans (right) and modern chimpanzees (left). Scientists hominid evolution since the lineages of chimpanzees and humans use the similarities and differences among these species to determine split some five or six million years ago.

to the north, show an interesting mix- story began. One of us (Leakey) has the layers of sediment, also carried out

and anamensis); VIDEO SURGERY ture of characteristics. Some of the traits mounted expeditions from the National by Feibel, reveal that the fossils here are primitive ones—that is, they are an- Museums of Kenya to explore the sedi- have been preserved by deposits from a cestral features thought to be present be- ments located southwest of Lake Turka- river ancestral to the present-day Kerio

chimpanzee fore the split occurred between the chim- na and to document the faunas present River, which once flowed into the Tur- panzee and human lineages. Yet these during the earliest stages of the basin’s kana basin and emptied into an ancient bones also share characteristics seen in history. Kanapoi, virtually unexplored lake that we call Lonyumun. This lake later hominids and are therefore said to since Patterson’s day, has proved to be reached its maximum size about 4.1 mil- have more advanced features. As our one of the most rewarding sites in the lion years ago and thereafter shrank as team continues to unearth more bones Turkana region. it filled with sediments. and teeth at Allia Bay, these new fossils A series of deep erosion gullies, known Excavations at Kanapoi have pri- add to our knowledge of the wide range as badlands, has exposed the sediments at marily yielded the remains of carnivore of traits present in early hominids. Kanapoi. Fossil hunting is difficult here, meals, so the fossils are rather fragmen- Across Lake Turkana, some 145 kilo- though, because of a carapace of lava tary. But workers at the site have also re- meters (about 90 miles) south of Allia pebbles and gravel that makes it hard to covered two nearly complete lower jaws,

ALAN WALKER; © NATIONAL MUSEUMS OF KENYA ( Bay, lies the site of Kanapoi, where our spot small bones and teeth. Studies of one complete upper jaw and lower face,

www.sciam.com SCIENTIFIC AMERICAN 17 COPYRIGHT 2003 SCIENTIFIC AMERICAN, INC. the upper and lower thirds of a tibia, bits conﬁdent in both the age of the fossils to a changed diet—possibly much hard- of skull and several sets of isolated teeth. and Brown’s and Feibel’s understanding er food—even though its jaws and some After careful study of the fossils from of the history of the lake basin. skull features were still very apelike. We both Allia Bay and Kanapoi—including A major question in paleoanthro- also know that anamensis had only a Patterson’s fragment of an arm bone— pology today is how the anatomical mo- tiny external ear canal. In this regard, it we felt that in details of anatomy, these saic of the early hominids evolved. By is more like chimpanzees and unlike all specimens were different enough from comparing the nearly contemporaneous later hominids, including humans, previously known hominids to warrant Allia Bay and Kanapoi collections of which have large external ear canals. designating a new species. So in 1995, in anamensis, we can piece together a fair- (The size of the external canal is unre- collaboration with both Feibel and Ian ly accurate picture of certain aspects of lated to the size of the ﬂeshy ear.) McDougall of the Australian National the species, even though we have not yet The most informative bone of all the University, we named this new species uncovered a complete skull. ones we have uncovered from this new Australopithecus anamensis, drawing on The jaws of anamensis are primi- hominid is the nearly complete tibia—the the Turkana word for “lake” (anam) to tive—the sides sit close together and par- larger of the two bones in the lower leg. refer to both the present and ancient lakes. allel to each other (as in modern apes), The tibia is revealing because of its im- To establish the age of these fossils, rather than widening at the back of the portant role in weight bearing: the tibia we relied on the extensive efforts of mouth (as in later hominids, including of a biped is distinctly different from the Brown, Feibel and McDougall, who have humans). In its lower jaw, anamensis is tibia of an animal that walks on all four been investigating the paleogeographic also chimplike in terms of the shape of legs. In size and practically all details of history of the entire lake basin. If their the region where the left and right sides the knee and ankle joints, the tibia found study of the basin’s development is cor- of the jaw meet (technically known as at Kanapoi closely resembles the one rect, the anamensis fossils should be be- the mandibular symphysis). from the fully bipedal afarensis found at tween 4.2 and 3.9 million years old. Mc- Teeth from anamensis, however, ap- Hadar, even though the latter specimen Dougall has determined the age of the pear more advanced. The enamel is rel- is almost a million years younger. so-called Kanapoi Tuff—the layer of vol- atively thick, as it is in all other species Fossils of other animals collected at canic ash that covers most of the fossils of Australopithecus; in contrast, the Kanapoi point to a somewhat different at this site—to be just over four million tooth enamel of African great apes is paleoecological scenario from the setting years old. Now that he has successfully much thinner. The thickened enamel across the lake at Allia Bay. The chan- ascertained the age of the tuff, we are suggests anamensis had already adapted nels of the river that laid down the sediments at Kanapoi were probably lined 3.9 MILLION 4.2 MILLION with narrow stretches of forest that grew YEARS AGO YEARS AGO OMO RIVER OMO RIVER close to the riverbanks in otherwise open country. Researchers have recovered the remains of the same spiral-horned ante- lope found at Allia Bay that very likely MODERN lived in dense thickets. But open-coun- LAKE TURKANA try antelopes and hartebeest appear to have lived at Kanapoi as well, suggesting that more open savanna prevailed away ALLIA BAY from the rivers. These results offer equi- MODERN vocal evidence regarding the preferred LAKE TURKANA habitat of anamensis: we know that bushland was present at both sites that have yielded fossils of the species, but LAKE there are clear signs of more diverse LONYUMUN habitats at Kanapoi.

KANAPOI An Even Older Hominid? AT ABOUT THE SAME TIME that we were ﬁnding new hominids at Allia Bay KERIO RIVER and Kanapoi, a team led by our colleague Tim D. White of the University of Cali- TURKANA BASIN was home to anamensis roughly four million years ago. Around 3.9 million years ago a river sprawled across the basin (left). The fossil site Allia Bay sat within the strip of forest (green) fornia at Berkeley discovered fossil hom- that lined this river. Some 4.2 million years ago a large lake filled the basin (right); a second site, inids in Ethiopia that are even older than

Kanapoi, was located on a river delta that fed into the lake. anamensis. In 1992 and 1993 White led SLIM FILMS; SOURCE: FRANK BROWN AND CRAIG FEIBEL (1991)

18 SCIENTIFIC AMERICAN NEW LOOK AT HUMAN EVOLUTION COPYRIGHT 2003 SCIENTIFIC AMERICAN, INC. an expedition to the Middle Awash area en the human fossil record back close to of Ethiopia, where his team uncovered the time of the chimp-human split. More hominid fossils at a site known as Ara- recently, White and his group have found mis. The group’s finds include isolated parts of a single Ardipithecus skeleton in teeth, a piece of a baby’s mandible (the the Middle Awash region. As White and lower jaw), fragments from an adult’s his team extract these exciting new fos- skull and some arm bones, all of which sils from the enclosing stone, reconstruct have been dated to around 4.4 million them and prepare them for study, the pa- years ago. In 1994, together with his col- leoanthropological community eagerly leagues Berhane Asfaw of the Paleoan- anticipates the publication of the group’s thropology Laboratory in Addis Ababa analysis of these astonishing finds. and Gen Suwa of the University of To- But even pending White’s results, kyo, White gave these fossils a new name: new fossil discoveries are offering other Australopithecus ramidus. In 1995 the surprises. A team led by Michel Brunet of group renamed the fossils, moving them the University of Poitiers has found frag- to a new genus, Ardipithecus. Earlier fos- ments of Australopithecus fossils in sils of this genus have now been found Chad. Surprisingly, these fossils were re- dating back to 5.8 million years ago. covered far from either eastern or south- Other fossils buried near the hominids, ern Africa, the only areas where Aus- FOSSIL HUNTER Alan Walker ( foreground) and such as seeds and the bones of forest tralopithecus had appeared. The Chad two colleagues excavate the bone bed at Allia monkeys and antelopes, strongly imply sites lie 2,500 kilometers west of the Bay, where several anamensis fossils have been recovered. The bone bed appears as a dark band that these hominids, too, lived in a western part of the Rift Valley, thus ex- about 18 inches thick at the top of the trench. closed-canopy woodland. tending the range of Australopithecus This new species represents the most well into the center of Africa. The significance of these exciting dis- primitive hominid known—a link be- These fossils debunk a hypothesis coveries is now the center of an active tween the African apes and Australo- about human evolution postulated by debate. pithecus. Many of the Ardipithecus ram- Dutch primatologist Adriaan Kortlandt The fossils of anamensis that we have idus fossils display similarities to the and expounded in Scientific American by identified should also provide some an- anatomy of the modern African great Yves Coppens of the College of France swers in the long-standing debate over apes, such as thin dental enamel and [see “East Side Story: The Origin of Hu- whether early Australopithecus species strongly built arm bones. In other fea- mankind,” May 1994]. This idea was lived in wooded areas or on the open sa- tures, though—such as the opening at that the formation of Africa’s Rift Valley vanna. The outcome of this discussion the base of the skull, technically known subdivided a single ancient species, iso- has important implications: for many as the foramen magnum, through which lating the ancestors of hominids on the years, paleoanthropologists have accept- the spinal cord connects to the brain— east side from the ancestors of modern ed that upright-walking behavior origi- the fossils resemble later hominids. apes on the west side. nated on the savanna, where it most like- Describing early hominids as either Brunet’s latest discovery, an impor- ly provided benefits such as keeping the primitive or more advanced is a complex tant cranium older than six million years, hot sun off the back or freeing hands for issue. Scientists now have almost deci- is also from Chad and shows that early carrying food. Yet our evidence suggests sive molecular evidence that humans hominids were probably present across that the earliest bipedal hominid known and chimpanzees once had a common much of the continent. This cranium, to date lived at least part of the time in ancestor and that this lineage had previ- which the team called Sahelanthropus wooded areas. The discoveries of the ously split from gorillas. This is why we tchadensis, together with fragmentary past several years represent a remarkable often use the two living species of chim- jaws and limb bones from about six mil- spurt in the sometimes painfully slow panzee (Pan troglodytes and P. panis- lion years ago in Kenya [see “An Ances- process of uncovering human evolution- cus) to illustrate ancestral traits. But we tor to Call Our Own,” on page 4], are ary past. But clearly there is still much must remember that since their last even older than the Ardipithecus fossils. more to learn. common ancestor with humans, chimpanzees have had exactly the same MORE TO EXPLORE amount of time to evolve as humans Australopithecus ramidus, a New Species of Early Hominid from Aramis, Ethiopia. Tim D. White,

National Geographic Image Collection have. Determining which features were Gen Suwa and Berhane Asfaw in Nature, Vol. 371, pages 306–312; September 22, 1994. New Four-Million-Year-Old Hominid Species from Kanapoi and Allia Bay, Kenya. Meave G. Leakey, present in the last common ancestor of Craig S. Feibel, Ian McDougall and Alan Walker in Nature, Vol. 376, pages 565–571; August 17, 1995. humans and chimpanzees is not easy. From Lucy to Language. Donald C. Johanson and Blake Edgar. Simon & Schuster, 1996. But Ardipithecus, with its numerous The Earliest Known Australopithecus, A. anamensis. C. V. Ward, M. G. Leakey and A. Walker in

KENNETH GARRETT chimplike features, appears to have tak- Journal of Human Evolution, Vol. 41, pages 255–368; 2001.