American Scientist

The best thing about paleontology is the surprises.

No matter how carefully you have analyzed the fossils, no matter how insightful your understanding of the links between anatomical form and function, Mother Nature always comes up with something totally unpredicted.

Surprises certainly have been sprung by, and on, the international team of paleoanthropologists and paleontologists that looks for fossils in the remote Aramis region of Ethiopia where the Afar people live. The team is co-led by Berhane Asfaw of the Rift Valley Research Service in Addis Ababa, Ethiopia; Yonas Beyene of the Ministry of Youth, Sports and Culture in Ethiopia;the late J. Desmond Clark, formerly of the University of California, Berkeley; Giday Woldegabriel of the Los Alamos National Laboratory; and Tim White of the University of California, Berkeley. With a nice touch of delicacy, White refers to Clark as “inspiring but no longer making decisions” about the project.

On October 2, 2009, the team published in Science their analyses of a hominin (member of the human lineage) called Ardipithecus ramidus. The best representative of the species is a partial female skeleton nicknamed Ardi; she is 4.4 million years old and is certainly astonishing and noteworthy. There are parts of at least 35 other individuals in the collection, in addition to thousands of specimens of plants, invertebrates, fish and assorted nonprimate mammals from the same location.

Ardi is the oldest partial skeleton in the hominin lineage at present, but she doesn’t represent a new species. Ardipithecus ramidus has been known since the first research paper about it was published in Nature in 1994 under what might be considered an “alias”: Australopithecus ramidus.

The naming of Ardi has had some singular twists and turns. The genus into which the type specimen of Ardipithecus was initially classified was Australopithecus, meaning “southern ape,” a term created in 1925 by Raymond Dart to accommodate a face, jaw and endocast (a natural cast of the braincase) of a young juvenile hominin found in South Africa. The “Taung baby,” as it was nicknamed, was the oldest (about 2.5 million years) and first member of the human lineage to be found in Africa at that point. Skepticism from the scientific establishment at the time of discovery was rife. The hoaxed Piltdown skull, with its large brain (variously estimated between 1,000 and 1,500 cubic centimeters) and seeming antiquity, had been announced in 1911–12 and was still believed to be genuine. Thus the small brain (about 340 cubic centimeters) of the Taung baby was so unexpected that many anthropologists dismissed the fossil as merely an immature, if ancient, ape. Surely our large brain was a fundamental human characteristic extending back to the beginning of our lineage! “Surely”—but not in fact.

More than 25 years passed before the anthropological community realized what Dart had recognized from his fossil: Our ancestors had humanlike teeth long before they had a large, humanlike brain. Instead of brain size being the fundamental adaptation that separated our earliest ancestors from the apes, it was our teeth and, later discoveries showed, our bipedal locomotion. How surprising—and disappointing to those who had prided themselves on their big brains.

Since 1925, many different species of Australopithecus have been described and named, including Australopithecus afarensis, the species to which the famous partial skeleton Lucy belongs. All are bipedal hominins, although not necessarily directly ancestral to modern humans. Some apparently ate fruit, whereas others had a more fibrous diet. Some were larger, some smaller, and most species showed a striking difference in body size between males and females (a trait called sexual dimorphism). Some probably made stone tools. They lived in both East and South Africa and may have been more widely spread across the continent between 4.2 and 1.2 million years ago.

More Surprises

When Tim White, Gen Suwa of Tokyo University, and Berhane Asfaw described the fossils now known as Ardipithecus ramidus in 1994, they classified the new specimens as Australopithecus. Because the specimens showed some unusual features that differed from other hominins, they defined a new species, Au. ramidus, derived from the Afar word ramid, meaning root.

In all biological sciences, naming a new species requires that the scientists fully describe the specimens anatomically and then provide a diagnosis, stating clearly in what ways the type specimen (the object to which the name is formally attached) differs from other known species. At the time, all they had were teeth and jaws, two cranial fragments and a few arm bones—but these pieces were enough to indicate that these 4.4 million-year-old remains were something new. Au. ramidus had small canine teeth, like hominins and unlike apes. Small canine teeth that don’t project below (or above) the rest of the tooth row are a key feature separating apes from hominins. Relative to the cheek teeth, the canines of Au. ramidus were still large but were not projecting. Tellingly, the first milk molar—the tooth that falls out in humans at about age five—was small, narrow and very similar to that of a chimpanzee.

At the end of the paper, the team made a quiet pronouncement that raised a few eyebrows:

We have taken a conservative position here regarding the placement of …[these] fossils at the family and genus levels. The major anatomical/behavioural threshold between known great apes and the Hominidae is widely recognized to be bipedality and its anatomical correlates…. The anticipated recovery at Aramis of additional postcranial remains, particularly those of the lower limb and hip, may result in reassessment of these fossils at the genus and family level.

All taxonomic assessments are subject to revision if new specimens show features not previously known, so why include such a statement at all? I think that the authors had named a new species—only a new species, not a new genus—because they had limited material, none of which answered the crucial question: How did this creature move? Depending on what the rest of this fossil creature looked like, they anticipated that they might need to shift ramidus onto the ape branch of the evolutionary tree (even though the teeth looked homininlike) or—if its pelvis and legs showed evidence of two-legged locomotion—into a new hominin genus. Perhaps they had one of those nonscientific hunches that the postcranial bones would be downright surprising.

In less than nine months’ time, an extraordinary but brief corrigendum appeared in Nature, written by the same authors. The name of the fossils was revised from Australopithecus ramidus to Ardipithecus ramidus (abbreviated as Ar. ramidus to prevent confusion). The new generic name, Ardipithecus, comes from the word for “ground” or “floor” in Afar. I know of no other instance in which the naming of a new genus has been published as a correction to a previous paper. More commonly, a new genus is named as part of a detailed reanalysis of already-known material and/or with a thorough description of additional, new fossils. That was not the case here.

The major distinction between Australopithecus and Ardipithecus mentioned in the corrigendum is that the dentition of the new genus is less dominated by large cheek teeth than the former. The implied explanation for the need to revise the name at all was a bald statement that a partial skeleton, including a jaw full of teeth and many postcranial bones, had been found 50 meters from the type specimen and at the same stratigraphic level. “Analysis of this specimen has begun,” the authors deadpanned, “and will provide further features with which to characterize Ardipithecus.”

The paleoanthropological community buzzed with excitement at this revision. First of all, the note was slipped into a major scientific journal in an unusual way. Second, the paper was as clever a way of saying “We’ve just found something amazing—wait until we tell you!” as anyone could imagine.

For nearly 15 years, the Aramis team held their peace and worked fervently on analyzing and understanding their new specimens, which continued to pile up. The most spectacular was the partial skeleton of Ardi. About the only comment made in public about the locomotion of Ardipithecus was a quip made by Tim White at a conference, “If you wanted to find something that moved like these things, you’d have to go to the bar scene in Star Wars.”

In October 2009, the team published their work in yet another unconventional way. They took over most of an issue of Science with 11 papers (more than 50 published pages of text and figures) written by 47 coauthors on Ardipithecus’ anatomy, inferred behavior, geology, dating, associated animals, plants and invertebrates, and preservation. At least as many pages of additional information appeared only in the supporting online material.

The decision to put all of this work into an issue of Science was influenced by a concern to make the facts accessible. Says White, “With the click of a mouse, Science makes available—to anybody with an Internet connection—a massive amount of material and data. This is effectively monographic treatment, for free, now.” A conventional monograph is planned for the future.

In a very real sense, the biggest surprise of all was what Ardipithecus ramidus was like. It was like nothing ever seen before.

Ardipithecus may not be the elusive last common ancestor of chimps and humans, but it is far and away the best known of the oldest species. Another species in the genus, Ar. kadabba, lived at least 5.8 million years ago. Together, these new materials give us stunning insights into what the last common ancestor was like.

Practically every paleoanthropologist in the world has drawn a large Y on the blackboard for a class, labeling the lefthand branch “chimpanzees” and the righthand one “modern humans,” and putting an X on the stem just below the divergence and labeling it “last common ancestor.” Based on genetic distances between chimps and humans, that last common ancestor probably lived 6 to 7 million years ago.

One of the key points Ardipithecus has revealed is how our thinking has been influenced by the very shape of this common Y diagram. Because we humans have foolishly thought of chimpanzees and other apes as primitive, we have been misled into thinking that the last common ancestor would somehow be a sort of half-chimp and half-human. Some even fell into the trap of thinking that the last common ancestor was basically a chimp, forgetting that chimpanzees have had as long to evolve since that ancestor as we have. (Students who are a bit confused on this point are always asking, “But if we are descended from the apes, why aren’t there new humans walking out of the jungle today?”)

“We have seen the ancestor,” says Tim White succinctly, “and it is not a chimpanzee.”

The ancestor also was not Lucy. As the best-known specimen of Australopithecus afarensis—a species that lived between 3.6 and 2.9 million years ago—Lucy became iconic of early hominins. Lucy was an especially petite and diminutive female, standing about 1.1 meters tall, weighing about 29 kilograms, with a reconstructed braincase of about 380 cubic centimeters. When she was discovered in 1974, she seemed incredibly primitive and apelike, so the notion took root in many people’s minds that our earliest ancestors ought to be small, like Lucy.

In contrast, Ardi is surprisingly large, which would have had a profound impact on her ecological niche. She stood perhaps 1.2 meters tall, weighed a hefty 50 kilograms, and had a brain size similar to Lucy’s, variously estimated between 300 and 350 cubic centimeters. She was no chimpanzee, but she was within the same size range (both bodily and brainily) as modern chimpanzees and bonobos.

Like a chimp, Ardi had long arms, but her fingers were only moderate in length, and her thumbs were intermediate between a chimpanzee’s relatively short thumbs and humans’ relatively long thumbs. Like a chimp, she had a divergent big toe and long toes that could be used to grasp tree branches during climbing.

Other Comparisons

The arboreal part of Ardi’s locomotion probably closely resembled that of Proconsul, an 18-million-year-old African genus. Proconsul was not a hominin but a primitive or ancestral ape near the divergence between the monkey lineage and the lineage of apes plus humans, and probably an early ancestor of Ardipithecus.

Like Ardi, both Proconsul heseloni and a larger species P. nyanzae are known from several partial skeletons. Both had grasping hands and feet, and fore- and hindlimbs roughly equal in length, features that are very important for an animal that moves quadrupedally in trees. Proconsul did not have any strong adaptations for jumping or leaping, so these animals probably moved slowly in the trees.

Two big differences between Proconsul and Ardi are body size and sexual dimorphism. P. heseloni females weighed about 8 kilograms, with males as heavy as 10.5 kilograms; P. nyanzae ranged from 27 kilograms for females to 40.5 kilograms for males. In contrast, Ardipithecus was larger, about 50 kilograms, with little apparent male-female difference in body size. Both Proconsul species and Ardi show many adaptations for slow clambering and quadrupedal locomotion in the trees, and none for knuckle-walking or brachiating. Proconsul is much older but gives clues about the anatomy from which something like Ardipithecus probably evolved.

Another striking resemblance between Proconsul and Ardipithecus lies in their preservation. Experiments feeding baboon carcasses to cheetahs carried out by C. K. “Bob” Brain, now retired from the Transvaal Museum in South Africa, showed that cheetahs leave a very particular pattern of bones uneaten. Cheetahs leave behind exactly those parts of the skeleton preserved in the Proconsul skeleton—teeth, cranial fragments, arms complete to the hands and legs complete to the feet—because cheetahs’ relatively weak jaws cannot break into the stronger and denser bones. Vertebrae, ribs and pelvises are almost completely destroyed. On this basis, Alan Walker of Pennsylvania State University suggested that the Proconsul partial skeletons had been eaten by an extinct carnivore analogous to the living cheetah.

The partial skeleton of Ardi shows the same pattern of preservation seen in cheetah kills and in Proconsul skeletons. Were there cheetahs in Aramis 4.4 million years ago? Not as far as anyone knows. But there were several catlike carnivores in ancient Aramis, including Dinofelis, which is an excellent candidate for Ardi’s killer.

Dinofelis is a false sabertoothed cat (its canines were flattened and only moderately long compared to those of true sabertooths) that lived between 5 million and 1.5 million years ago, probably in woodland habitats such as those where Ardipithecus lived. Dinofelis weighed two to three times more than a cheetah, yet its jaws were relatively weak because of the huge gape needed to use its elongated canines effectively. Also, Ardipithecus is about twice as large as a baboon and would have had more substantial bones. Though the research team concluded that most of the hominins and other animals they found were ravaged by hyenas after death, perhaps Ardi herself was attacked by a different carnivore, which accounts for her excellent preservation.

When a Chimp Is Not a Chimp

If Ardi was living in woodlands, weighed about as much as a chimp, was about as tall as a chimp, and had arms and legs and hands and feet like a chimp’s, why was she not a chimpanzee? Let me count the ways.

First, she doesn’t have apelike teeth. She has humanlike teeth, with small canines that barely project beyond the level of the tooth row.

Second, her dental proportions are wrong. In chimps, the front teeth—incisors and canines—dominate the dentition; in humans, the opposite is true. Ardi’s teeth lie somewhere in the middle of the two.

Third, the Ardipithecus remains known so far indicate that there is little sexual dimorphism in the species in canine or body size. This is one of the oddest features about Ardipithecus. Strong sexual dimorphism is common among living Old World monkeys and apes, among fossil primates older than Ardipithecus (such as Proconsul), and among hominins in the genus Australopithecus that are more recent than Ardi. If something Proconsul-like evolved into Ardi, and later into Australopithecus, sexual dimorphism was first lost and then regained.

Fourth, although her arms and fingers are long like a chimp’s, Ardi’s wrists and hands have no adaptations for brachiating branch to branch like a young chimpanzee or for knuckle-walking on the ground like an adult chimp. Ardi walked quadrupedally in the trees, placing her hands palm-down on branches and grasping with her big toes. Because she was relatively large, Ardi was probably a slow and careful climber on large branches much of the time.

Then there is the most surprising and unchimplike feature of all. According to team member C. Owen Lovejoy of Kent State University, Ardi had a pelvis like no other primate known to science. The upper part of the pelvis in humans forms a large bowl—consisting of the short, curving iliac blades—that cradles the viscera. In a chimp, the flat, long and narrow iliac blades lie behind (or dorsal to) the viscera. In humans, muscles attached to the broad iliac blades balance the body weight over a single pivot leg during bipedal walking. These muscles are differently aligned in apes and cannot serve this function, which is why chimps stagger from side to side when they walk bipedally. Ardi’s upper pelvis was a primary adaptation to bipedal walking on the ground, much closer in shape to Lucy’s pelvis than to a chimp’s. Paradoxically, the lower part of Ardi’s pelvis is very apelike, which is important in climbing vertical trunks with long arms wrapped around the trunk, like an ape or a telephone lineman.

“Ardi isn’t a transitional biped,” says Lovejoy. “She’s a biped from a species that’s going to keep its grasping toe and climbing thigh musculature, thank you very much.” He views Ardi’s locomotion and anatomy as mosaics of those in living species.

Ardi was clearly not chimp, nor Lucy nor human. Paleoartist Jay Matternes reconstructed her as having fur everywhere but on the center of her face, which protruded more than expected for a hominin. The fur, the long arms and the divergent big toe give Ardi the gestalt of an odd chimp, but that bipedal gait transforms Ardi back into a hominin.

The lasting messages from this tour-de-force analysis are complex, but the most important lesson of all is simple. Studying the behavior and anatomy of living animals provides us with an understanding of how anatomy links form and function. Genetic comparisons reveal the branching patterns that occurred in evolution and how one living species is related to another. Only the fossils really show us the past.