Masters of the Planet

by Ian Tattersall

  Before we look at the varied cast of contenders for the title of “most ancient hominid,” perhaps we should pause for a moment to consider just what an early hominid should look like. How would we recognize the first hominid, the earliest member of the group to which we belong to the exclusion of the apes, if we had it? The question seems straightforward, but the issue has proven to be contentious, especially since members of related lineages—such as our own and that of the chimpanzees—should logically become more similar to each other, and thus harder to distinguish, as they converge back in time toward their common ancestor. But while the characteristics that define modern groups should even in principle lose definition back in the mists of the past, attempts to recognize very early hominids have paradoxically been dominated by the search for the early occurrence of those features that mark out their descendants today.

  When the Dutch physician Eugene Dubois discovered the first truly ancient human fossil in Java in 1891, he called his new find Pithecanthropus erectus (“upright ape-man”). His choice of species name emphasized the importance he attached to the erect stature of this hominid (indicated by the structure of its thighbone) in determining its human (or at least close-to-human) status. But soon thereafter the emphasis changed, at least temporarily. Modern people are perhaps most remarkable for their large brains; and in the early years of the twentieth century, brain size expansion replaced uprightness as the key criterion for any fossil seriously considered for inclusion in the hominid family. Indeed, its big human braincase (which was matched with an ape jawbone) was the basis for recognizing the famously fraudulent English Piltdown “fossil” as a human ancestor in 1912. The fraud was only officially uncovered some 40 years later, although many scientists were suspicious of it from the start; and as time passed the Piltdown specimens became increasingly ignored, which had the effect of bringing the big-brain criterion into disfavor. In its place came a behavioral yardstick rather than an anatomical one: manual dexterity and the manufacture of stone tools became the key to human status, as the notion of “Man the Toolmaker” took hold.

  But this too had its difficulties. Eventually and inevitably, attention refocused on anatomy, and various potentially diagnostic morphological features of hominids were touted. Teeth, which are coated with the toughest biological material and thus preserve particularly well in the fossil record, received particular attention. One dental characteristic that many noticed among potential early hominid fossils was thick molar enamel—although, as we have seen, this indicator of a tough diet is also found widely among Miocene apes. Another hominid dental feature that has perennially attracted attention is the reduction in size of the canine teeth. This occurs in conjunction with the loss of honing of the large upper canine against the front premolar of the lower jaw with which it occludes. Large-bodied male apes typically have fearsome upper canine teeth with razor-sharp back edges—although in small females these teeth can be dainty. But again, a tendency toward canine reduction is not unique to hominids. It is also found in various Miocene apes, most famously the bizarre late-Miocene Oreopithecus, an insular form that additionally showed a distinct tendency toward postural uprightness. What is more, the remarkable Oreopithecus was recently reported to have had “precision-grip capability”—something else that was once thought unique to tool-making hominids.

  Part of the problem of spotting features that are unique to hominids stems from the nature of evolutionary diversification. As we look farther back into hominid history, every feature indicative of modern hominids is likely to become less distinctive—and more reminiscent of its counterparts in members of related lineages. Given this reality, it is hardly realistic to expect that we’ll ever find an anatomical “silver bullet” that will by itself tell us infallibly if an ancient fossil is a hominid or not. Every effort to do this has foundered on one technicality or another. Take, for example, the early-twentieth-century attempt of the English anatomist Sir Arthur Keith to set a “cerebral Rubicon” of 750 cubic centimeters (cc) minimum brain volume for membership in the genus Homo. Any smaller than this, Keith said, and you didn’t belong to the club. This was certainly a convenient and easily measurable criterion; and, at a time when very few hominid fossils were known, perhaps it was even a workable one. But predictably, as the hominid fossil sample increased, problems arose. Brain size is notably variable within populations (modern human brains range in size from about 1,000 to 2,000 cc, with no indication that people with larger brains are necessarily smarter), so that even in principle this standard might have admitted an ancient hominid to our genus while excluding his or her parents or offspring. Accumulating fossil finds predictably forced later authors to lower Keith’s figure several times, until it became obvious that the entire “Rubicon” idea was misguided.

  Similar objections apply to any touchstone of this kind for membership in the genus Homo or the family Hominidae. But the temptation to see matters from the “key criterion” perspective is nevertheless always there. Indeed, in recent years paleoanthropologists have come full circle back to Dubois’ view, so that the most notable common factor uniting all currently touted “earliest hominids” is the claim that each had walked bipedally on the ground. This seemingly straightforward standard for membership in our family is particularly attractive given that in the latest Miocene the eastern African forests were beginning to yield to patches of more open territory. This would have obliged at least some ape populations to spend more time on the ground (though extinction was, as always, the easier option for steadfastly arboreal types). Still, if this environmental change forced one ape lineage to stand upright, why not others? Several likely did; but only one of them can have been the hominid progenitor.

  A further confounding factor is that all of the known “very early hominid” fossils have been found in contexts indicating thickly wooded habitats, or at least mixed ones. The earliest hominids were thus not obliged to walk upright on the ground by the disappearance of their ancestral habitat. We humans have rather reductionist minds, and are beguiled by clear, straightforward explanations. But where murky Mother Nature is concerned, beware of excessively simple stories.

  THE CAST OF CHARACTERS

  Until close to the turn of this century, the known hominid fossil record extended back in time to only about three to four million years ago. But a remarkable series of finds has since turned up a variety of contenders for the mantle of “earliest” hominid that are significantly older than this. The oldest of them come from around the time that DNA studies suggest our ancestors parted company with our closest ape relatives, believed to be the chimpanzees and bonobos.

  “Toumaï” and Orrorin

  The most ancient of the “earliest hominids” on offer today is the close-to-seven-million-year-old species Sahelanthropus tchadensis, discovered in 2001 in the central-western African country of Chad (well to the west of the Rift Valley). What has so far been published of this form consists of a badly crushed cranium (informally dubbed “Toumaï”—”hope of life” in the local language) and some partial mandibles. These fossils caused a stir when discovered, because nobody had anticipated an ancestral hominid like this. What was particularly strange about Toumaï was that it combined a small (hence rather apelike) braincase with a large, flattish face that was distinctly unlike the more protruding snouts of younger fossil hominids (or apes, for that matter). Two things caused its describers to classify this form as a hominid: first, the teeth. The molars had moderately thick enamel, the canines were reduced, and there was no lower premolar honing mechanism. So far, so good; but as we’ve seen, both thick enamel and the reduced canine-premolar complex can be matched outside Hominidae. So the key finding was in the base of the crushed cranium, where the foramen magnum, the large hole through which the spinal cord exits the cranium, appeared to be shifted underneath the skull to face largely downward. This is significant in that you would expect to find this setup in an upright biped like us: a skull balanced atop an erect spine. In a quadrupedal chimpanzee
, the skull hangs on the front of a horizontal spine, so the foramen magnum has to be at the rear of the skull, facing backward. Unfortunately, though, the skull of Sahelanthropus was badly crushed, so the crucial claim about its foramen magnum was inevitably disputed.

  In response, researchers took CT-scans of the crushed skull in a medical scanning machine, and produced a computerized virtual reconstruction that eliminated the distortions. Now, no matter how high-tech the procedure is, there’s always an element of human judgment involved in making any reconstruction. But the resulting model of the pristine Sahelanthropus skull gave its creators substantial grounds for viewing Toumaï as plausibly—if not definitively—the skull of a biped. There are still some skeptics; but although the bipedality question will never be finally settled until key parts of the body skeleton of Sahelanthropus are announced, the reconstruction does appear to give this form the benefit of the doubt.

  If Toumaï was a hominid—or even if he wasn’t—what can we say about his way of life? Fossils found in the same deposits suggest that Sahelanthropus lived in an environment that was well watered, with forest in the close vicinity. This doesn’t tell us much directly, but it does say something about the kind of resources that were available to this presumed ancestor. Put this information together with its posture, its habitat, and the general form of its teeth, and it seems reasonable to suggest that Sahelanthropus was at least a part-time biped that subsisted on a fairly generalized plant-based diet that would have included fruit, leaves, nuts, seeds, and roots, and probably extended to insects and small vertebrates such as lizards. For the moment it’s probably unwise to say too much beyond that, though we’ll speculate a bit about such things as the nature of early hominid sociality in a little while.

  Almost as old as Toumaï is a form discovered in northern Kenya in 2000 (hence its nickname “Millennium Man”), and technically known as Orrorin tugenensis. Found at a number of localities dated to about six million years ago, the materials attributed to Orrorin are fragmentary, consisting of some bits of jaw and teeth and several limb bones believed (but not demonstrated) to have belonged to members of the same species. The molar teeth are thick-enameled, squarish, and not too large: all features that might be expected in an early hominid. What’s more, an upper canine is encouragingly small. But controversy has centered on the incomplete femora (thigh bones), which are unfortunately broken just where their morphology (anatomical structure) would be crucial for establishing bipedality. Still, what’s left is entirely consistent with upright locomotion. At the upper end of the body, a piece of humerus (upper arm bone) has a strong attachment area for a key climbing muscle; and one finger bone is strongly curved. Both of these features are indicative of climbing and branch-grasping, and the fossil animals found in the same area suggest that all had lived in a dryish evergreen forest environment, with a notable absence of grassland ruminants. All in all, the Orrorin fossils quite strongly support the notion that bipedal hominids were around in the slowly desiccating eastern African forests some six million years ago—a period during which DNA comparisons among humans and their relatives also suggest it’s reasonable to expect early hominids to be found.

  “Ardi”

  The third entrant in the “earliest hominid” stakes is Ardipithecus, a recently ballyhooed primate recovered from rocks in the valley of the Awash River in northern Ethiopia. In 1994 some fragments attributed to the species Ardipithecus ramidus were reported from 4.4-million-year-old deposits at a place called Aramis, from which an almost complete, if rather crushed and distorted, skeleton was published in 2009. Found eroding from the desert rocks in horribly crumbly condition, it had taken scientists over a dozen years to restore and study the skeleton. In addition, an earlier species of the same genus, Ar. kadabba, was named in 2001 from fossils found in several nearby localities between 5.2 and 5.8 million years old. Though since slightly augmented, Ar. kadabba is represented only by miscellaneous materials from sites scattered in time and space, and their association in the same species is even less secure than in the case of Orrorin.

  Most Ar. kadabba fossils are teeth and bits of jaw. The canines almost rival those of female chimpanzees in size, though they are less pointy, while the molar enamel is uncomfortably thin. Postcranial (below the neck) elements include some small pieces of arm bone, a fragment of clavicle (collarbone) and two finger bones; perhaps most interesting is one toe bone, the youngest of all the Ar. kadabba fossils at 5.2 million years. This element is strongly curved (i.e., apelike); but it nonetheless resembles its equivalents in later hominids in how it articulates with the bone behind, a feature taken as evidence for bipedality. The upper limb fossils are said to be much more apelike, arguing for a form that, as typical among early hominids, had a much more primitive structure of the upper body than of the lower part. Associated fossils suggest a wooded environment.

  The recent publication of the skeleton of Ardipithecus ramidus has given us a uniquely comprehensive glimpse of a putative early hominid. And it is indeed a strange beast. A virtual reconstruction of the badly crushed skull revealed a braincase with a volume of between 300 and 350 cc. This is about the size of a chimpanzee’s brain today, and it matched a body equivalent in bulk to a small chimpanzee’s, weighing about 110 pounds. Unlike humans, apes have small braincases with large faces protruding at the front. And despite some facial reduction, the “Ardi” skull resembles other presumed early hominids in having essentially apelike skull proportions. Its modestly sized molar teeth have enamel that is reportedly thicker than that in the Ardipithecus fragments originally discovered, its canines are much smaller than those of Ar. kadabba, and there is no premolar honing mechanism. The original Ardi fragments from Aramis had included a piece of cranial base that was said to show a somewhat forward shift and downward orientation of the foramen magnum; and, although incomplete, the new skull reportedly shows the same tendency. All in all, while the reconstructed skull of Ardi does not scream “hominid!” in every respect, its attribution in isolation to an early member of our family might not raise eyebrows excessively.

  Modern apes (especially males) have very large, pointed upper canine teeth which hone against the front premolars below. In contrast, modern humans have very reduced upper and lower canines that barely project beyond the other teeth, if at all. In determining if a fossil is that of an ape or a hominid, one thing paleoanthropologists look for is evidence of canine reduction. Here we see a side view of the teeth of a presumed male Ardipithecus ramidus (in the center), compared to a male chimpanzee (above) and a human male (who also, like many of us, lacks wisdom teeth and has an overbite). Ardipithecus shows an intermediate condition in which both the upper and lower canines are both reduced, but remain pointy and slightly projecting. Other hominids of the “very early” group show a broadly similar conformation. Illustration by Jennifer Steffey.

  But what a difference below the neck! The arm and hand bones of Ardi are those of a highly arboreal animal that was well adapted for climbing in trees. Given what we already knew about later hominids, which retain climbing features in the upper body, this came as no surprise. Perhaps more remarkably, though, these bones do not show any of the “knuckle-walking” features seen in the forearms and hands of the chimpanzees and gorillas usually reckoned to be our closest living relatives. Both extant African apes are essentially arboreal creatures (except for adult male gorillas, which are just too heavy to clamber around in most trees). When on the ground, they occasionally rear up and walk short distances on their hind limbs, making displays or even carrying objects; but all apes are basically quadrupeds while on the forest floor— and the long, slender fingers that they depend upon for grasping tree limbs would get in the way during movement on the ground, except for one thing. So when walking on all fours, both chimpanzees and gorillas curl their fingers up into a fist, bearing the weight of the front of their bodies on the outside of the first knuckles. In this way they reduce the effective length of their arms relative to their legs, and this permits
more comfortable four-legged walking while also getting those vulnerable long fingers out of harm’s way. This unusual accommodation to compressive weight-bearing, by extremities that are basically adapted to the tensile strains of arboreal life, is clearly reflected in the structure of the apes’ hands and wrists.

  But of course, apes are apes and humans are humans. So why is the absence of any hint of knuckle-walking in Ardi a worry? After all, we Homo sapiens show no structural signs of being descended from a knuckle-walking ancestor. The question arises because the molecular systematists who compare the structure of human and ape DNA agree in concluding that humans are more closely related to chimpanzees than they are to gorillas, sharing more DNA similarities. They are even prepared to hazard estimates of when gorillas split from the human/chimpanzee group, and when humans split from chimpanzees, based on the assumption of a more or less regular rate of change in the DNA molecule over time.

  Such molecular age-of-split estimates usually tend to look a little low to paleontologists: most are in the range of 5 to 7 million years ago for humans and chimpanzees, with the gorillas peeling off a couple of million years earlier. But whatever the exact times of divergence, this all means that if the common ancestor of the knuckle-walking chimpanzees and gorillas also walked that way, then so must the chimpanzee-human ancestor. In which case, knuckle-walking must have been lost in the human lineage after the chimpanzee-human split—and you might expect to find some telltale signs of a knuckle-walking past in the wrist and hand of an alleged early human ancestor such as Ardi. The absence of any such signs in Ardi makes you wonder a bit either about Ardi itself, or about our current received wisdom concerning relationships among humans and their closest living relatives.