By the close of 2002, there were at least three contenders for the title of “earliest known human.” There was the 7 million year old Sahelanthropus tchadensis from the Djurab Desert, the 6 million year old Orrorin tugenensis from Kenya, and the 5.6 million year old Ardipithecus kadabba from northeastern Ethiopia’s Afar region. Though very different from each other, each species appeared to fit somewhere near the base of the human family tree, not long after our split with the ancestors of chimpanzees. But which of these species were truly our ancestors, and how can we be sure that we are properly reading their bones?
Separating fossil ancestors from collateral relatives is no simple task. In his 1870 address as the president of London’s Geological Society, the naturalist Thomas Henry Huxley reminded his colleagues that creatures which appeared to be perfect transitional forms between one species and another might – upon accumulation of further evidence – turn out to be cousins or uncles rather than fathers. A fossil animal with transitional features undoubtedly attested to the reality of evolution, “But the mere discovery of such a form does not, in itself, prove that evolution took place by and through it.”
Huxley learned this the hard way a few years later. During his 1870 address, he proposed that paleontologists had directly connected the modern horse Equus to at least two of its fossil predecessors – the three-toed horse Hipparion and the even older fossil equid Anchitherium. Both of these horses were found in Europe – confirming that horses were animals of the Old World – but a stunning collection of fossil horses collected by American paleontologist Othniel Charles Marsh proved Huxley wrong. Marsh’s collection confirmed that horses originated in North America, and the supposed transitional forms Huxley cited were species that had dispersed to Europe at different times. Huxley saw these fossils for himself when he visited Marsh at Yale’s Peabody Museum during an 1876 speaking tour, and he immediately changed his lecture to include Marsh’s discoveries.
The factors that led Huxley to misread the evolution of horses have also confounded the study of human origins. The most obvious is the elusiveness of the creatures that fill out the evolutionary transition. Fossil humans are very rare. The earliest humans, especially, lived in forests where their bodies were much more likely to be destroyed than quickly preserved. As a result, we are always working with an imperfect record, and therefore it is easy to take a species that appears to fill a gap in our knowledge and promote it as ancestor of something else. The belief that humans have always been relatively rare and progressed along a narrow evolutionary path has reinforced this type of argument. There was a single “main line” of human evolution leading to us, the standard trope goes, and species that don’t fit – such as the robust australopithecines – are pushed off onto a “side branch.”
But there is another problem that frustrates efforts to identify the earliest humans. Our changing understanding of horse evolution provides another instructive example. Since the late 19th century, horse evolution has been depicted as starting with a small, multi-toed, leaf-eating ancestor that – depending on the source – was called either Hyracotherium or Eohippus. Contrary to simplified depictions, though, paleontologists discovered and described many specimens of this type of mammal, creating an awful tangle of names and attributions. Paleontologist David Froehlich published a reanalysis of many of these fossils in 2002. He found that there was actually several genera of distinct, early horses hiding under the names Hyracotherium and Eohippus, and also several animals that were not horses at all.
Early in their evolutionary history, the horses, rhinos, tapirs, and other odd-toed, hoofed mammals were so similar to each other that only a seasoned expert can tell their bones apart. Separating the groups requires a relatively complete view of a group’s evolutionary history so that the distinctive, subtle traits which distinguish one group from another can be traced. In other words, the earliest members of a particular group looked very much like their close relatives among collateral lineages, and gaps in the fossil record make it extremely difficult to follow the changes in specific traits that allow paleontologists to tease out relationships. In the case of early humans, our knowledge of what happened between 4 and 8 million years ago is very incomplete, and the remains of the three species in contention for the title of “oldest human” are only fragmentary. Each of them is relevant to the question of how humans originated, but, given how sparse our knowledge of early humans is, we should take great care before declaring any of them to be ancestral to any other.
In fact, the fossils that would do the most to educate us about our origins would not be human at all. We know virtually nothing about the chimpanzee fossil record. The two living species of chimpanzee are undoubtedly our closest living relatives, and our divergence from their lineage has been projected as occurring somewhere between 4 and 6 million years ago, but, when it comes to chimpanzee evolution, there is basically no story to tell. The only confirmed chimpanzee fossils are a few teeth found in approximately 545,000 year old deposits in Kenya. Chimpanzees have been evolving for as long as we have, but we don’t have the fossils to document how they have changed.
Without a firm understanding of the chimpanzee fossil record, it is even harder to identify the earliest humans. If we knew what the earliest chimpanzees were like, we could make much stronger comparisons of the candidates for the earliest humans and detect the differences between the two lineages. Perhaps some of the fossils marked as “early humans” would even turn out to belong to early chimpanzees or other collateral branches of the great apes. We can’t solve the mysteries of human origins without understanding the evolution of our close ape relatives.
In last week’s Nature, paleoanthropologists Bernard Wood and Terry Harrison published a review paper titled “The evolutionary context of the first hominins.” It was a sequel to a similar piece Harrison printed in Science last year, and both articles were prompted by the detailed description of Ardipithecus ramidus. At 4.4 million years old, “Ardi” is not the oldest human contender yet found, but it is the most completely known human from before four million years ago and would therefore seem to bring us that much closer to an image of what the last common ancestor of chimpanzees and humans was like. Naturally, this made it easy to promote Ardipithecus as a human ancestor, but, as explained by Wood and Harrison, distinguishing early humans from other fossil apes is not as easy as previously believed.
The arguments for Sahelanthropus, Orrorin, and both species of Ardipithecus being early humans vary depending upon the parts of each species that have been found. Wood and Harrison break them down into three categories which have been drawn from expectations based upon differences between our species and living apes. Since we have small canine teeth and there is very little difference in canine teeth between the sexes, early humans are expected to have relatively small canines, too. We also walk upright, and so early humans are expected to have specializations of the hips, legs, and feet for walking bipedally, as well as a forward-oriented hole in the bottom of the skull – the foramen magnum – which indicates that the head sat atop the spinal column. Find an ape-like creature between 4 and 8 million years ago that fits at least some of these criteria, and you have a contender for an early human.
What remains unclear is whether these characteristics are good indicators of membership in the human family. Wood and Harrison suggest that they might not be. Both Oreopithecus and Ouranopithecus, approximately 8 million year old apes from Europe, had reduced canines, as did the later Gigantopithecus from Asia. The evolution of smaller canines multiple times may be related to diet, the anthropologists suggest, and so it can’t be taken as a definitive human trait.
The same goes for the position of the foramen magnum. Rather than indicating a vertical spinal column adapted to bipedal walking, Wood and Harrison state that the orientation of this hole has more to do with the length of the face and how far forward the head is positioned. Likewise, the scientists argue that characteristics of the legs and hips thought to be associated with upright walking can also be seen in fossil apes that were not ancestral to humans. Oreopithecus, in particular, has been reconstructed as a gibbon-like ape that could also walk bipedally. The Swiss paleontologist Johannes Hurzeler – who described some of the best Oreopithecus fossils during the 1950’s – even suggested that it might be a human ancestor, although this view has been discarded. The traits Oreopithecus and other apes share in common with some of the earliest proposed humans should cause us to take great care in deciding what makes a hominin and what does not.
Even though the three categories of traits highlighted by Wood and Harrison distinguish later, more specialized humans from other apes, they are not necessarily useful in detecting the first humans. Convergent evolution – when two separate lineages independently evolve the same characteristics – confounds the process. Convergent traits, if they go unrecognized, might make it seem that two species are more closely related than they truly are, and, in the case of human origins, cause us to crown false ancestors. Anyone who has studied the history of paleoanthropology knows how many times the list of our direct ancestors has been changed, and future discoveries of previously-unknown species will continue to change the picture. We have an outline of our ancestry, but the details are still subject to change. As Wood and Harrison state:
We are not advocating that researchers abandon trying to draw inferences about the phylogenetic relationships of hominins at the finest scale possible. However, we do suggest that those who present and accept these hypotheses need to be aware that such inferences, especially ones about stem taxa, are likely to be inherently prone to refutation and subsequent revision.
Despite the wonderful discoveries made in Africa over the past decade, there is still much we don’t know about the earliest humans. Even pinpointing what characteristics identify the earliest humans has become a challenge. All of the contenders for “earliest known human” have been heavily criticized, and there has been little resolution as to what they actually are.
Sahelanthropus has been the most controversial. Represented by a crushed skull, fragments of jaw, several teeth, and possibly a femur, this creature was found in the sands of Chad’s Djurab Desert, far west of the well-known fossil human sites of the east African rift valley. The remains were not found in the area’s fossil deposit, but instead they had weathered out of the 7 million year old rock and may have even been moved by people sometime before their discovery. Naturally, the fact that a skull was found made this ape an instant celebrity – skulls look much better in magazines and on book covers than fragments of the rest of the skeleton – but the small canines and forward-oriented foramen magnum are not enough to designate Sahelanthropus as a human. Milford Wolpoff, Johan Hawks, Brigitte Senut, and Martin Pickford fired off a brief critique of the fossil in 2002, and, joined by James Ahern, provided a more detailed reassessment in 2006, finding that Sahelanthropus probably was not a human but instead an ape with human-like traits. Though the new review by Wood and Harrison serves as a prominent reminder, these earlier studies should have made anthropologists think twice about the traits traditionally thought to identify the first humans.
A lack of material plagues the identification of Orrorin, too. Found among 6 million year old fossil sites in Kenya, Orrorin would seem to be in the right place at the right time to be one of the first humans, but a few teeth, jaw fragments, broken femurs, part of a humerus, and portions of the digits are all that have been found. While a 2008 study by Brian Richmond and William Jungers found that the femur was very similar to that of australopithecines and may mean that Orrorin was bipedal, the other parts of the skeleton that would confirm this are missing, and, as with Sahelanthropus, the traits said to identify Orrorin as a human may not be diagnostic. Martin Pickford – one of the scientists who described Orrorin – has proposed that there may have been a direct line from the 6 million year old ape to the first members of our own genus (Homo) to the exclusion of “Lucy” and all her australopithecine kin, but at best this is unhindered speculation that has no scientific support.
Then, of course, there’s Ardipithecus. At this moment, there are two recognized species – the 5.6 million year old Ar. kadabba and the 4.4 million year old Ar. ramidus. As described in a special 2009 issue of Science, the younger species is the better known of the two, but together they have been proposed by Tim White and colleagues to be part of a single ladder of early humans stretching from around the last common ancestors with chimpanzees to the early australopithecines. (White has been especially critical of the idea that there was any branching early on in early human evolution, preferring a linear model of one species shading into another.) Of course, not everyone is agreed on this point. Harrison explicitly criticized the identification of Ardipithecus as an early human in his 2010 commentary and his new review with Wood, calling attention to the various “human” traits in Ardipithecus that have been seen in other apes, like Oreopithecus. White blasted this critique in news reports about the Nature review, stating that it offered nothing new and was a failure of peer-review. Who is right and who is wrong will rely on future discoveries and analysis, but, so far, Ardipithecus appears to be the best candidate for the earliest known human and representative of the type of human from which the australopithecines evolved. The question is whether this is because the evidence is that good, or simply because Ardipithecus ramidus is the currently the most completely-known species?
Such controversy aside, though, these fossils are and will remain important to our attempts to understand where we came from. Even if every single one turns out to be a close evolutionary cousin rather than an early human, that would still tell us quite a bit about ape evolution and what our ancestors might have been like between 4 and 6 million years ago. We may never be able to tell whether Ardipithecus ramidus was one of our ancestors, for instance, but the transitional features in Ardi’s skeleton have allowed paleoanthropologists to better calibrate their image of what early humans – or early chimpanzees closely related to us – might have been like. Even though modern chimpanzees are often used as models for the first humans and our last common ancestor with the chimpanzee lineage, the description of Ardipithecus ramidus has shown that the fossil apes around the root of our lineage were unique creatures not quite like any apes alive today. The modern great apes – orangutans, gorillas, chimpanzees, and us – are just the last remaining twigs of what was once a more widespread and diverse group. As discoveries continue to be made, our expectations about ape and human history change.
Our relationships to Sahelanthropus, Orrorin, and Ardipithecus remain unresolved. Doubts remain as to which of them were humans, and, even if some of them turn out to be hominins, we can’t confirm which species may have been ancestral to us. If anything, discoveries made by paleoanthropologists since the late 20th century have confirmed that the human family tree is quite bushy, and not every fossil species was directly ancestral to us. We see this during the latter half of our evolutionary history – from about 3 million years on – and I think future efforts by paleoanthropologists will uncover additional early species. It would be foolish to assume that we have now named all the players and can now go about fitting them into this or that role in the human evolutionary drama. Many of the most controversial fossils have only been found in the last 20 years, and we only know them from scraps. I can only imagine the things future paleoanthropologists will laugh about when they look back on what we thought we knew about our own prehistory.
[For more, see John Hawks’ post and Katherine Harmon’s Scientific American article about the review by Wood and Harrison. Ann Gibbons’ book The First Human is also essential reading for those interested in this debate.]
Top Image: The reconstructed skull of Lucy (Australopithecus afarensis) next to a sculpture of a Homo sapiens skull at Yale’s Peabody Museum.
Brunet, M., Guy, F., Pilbeam, D., Mackaye, H., Likius, A., Ahounta, D., Beauvilain, A., Blondel, C., Bocherens, H., Boisserie, J., De Bonis, L., Coppens, Y., Dejax, J., Denys, C., Duringer, P., Eisenmann, V., Fanone, G., Fronty, P., Geraads, D., Lehmann, T., Lihoreau, F., Louchart, A., Mahamat, A., Merceron, G., Mouchelin, G., Otero, O., Campomanes, P., De Leon, M., Rage, J., Sapanet, M., Schuster, M., Sudre, J., Tassy, P., Valentin, X., Vignaud, P., Viriot, L., Zazzo, A., & Zollikofer, C. (2002). A new hominid from the Upper Miocene of Chad, Central Africa Nature, 418 (6894), 145-151 DOI: 10.1038/nature00879
FROEHLICH, D. (2002). Quo vadis eohippus? The systematics and taxonomy of the early Eocene equids (Perissodactyla) Zoological Journal of the Linnean Society, 134 (2), 141-256 DOI: 10.1046/j.1096-3642.2002.00005.x
Haile-Selassie, Y. (2001). Late Miocene hominids from the Middle Awash, Ethiopia Nature, 412 (6843), 178-181 DOI: 10.1038/35084063
Harrison, T. (2010). Apes Among the Tangled Branches of Human Origins Science, 327 (5965), 532-534 DOI: 10.1126/science.1184703
Huxley, T.H. (1870) The anniversary address of the president. Quarterly Journal of the Geological Society, London, XXVI, xxix–lxiv.
McBrearty, S., & Jablonski, N. (2005). First fossil chimpanzee Nature, 437 (7055), 105-108 DOI: 10.1038/nature04008
Senut, B. (2001). First hominid from the Miocene (Lukeino Formation, Kenya)Premier hominidé du Miocène (formation de Lukeino, Kenya). Comptes Rendus de l’Académie des Sciences – Series IIA – Earth and Planetary Science, 332 (2), 137-144 DOI: 10.1016/S1251-8050(01)01529-4
White, T., Asfaw, B., Beyene, Y., Haile-Selassie, Y., Lovejoy, C., Suwa, G., & WoldeGabriel, G. (2009). Ardipithecus ramidus and the Paleobiology of Early Hominids Science, 326 (5949), 64-64 DOI: 10.1126/science.1175802
Wolpoff, M., Senut, B., Pickford, M., & Hawks, J. (2002). Palaeoanthropology (communication arising): Sahelanthropus or ‘Sahelpithecus’? Nature, 419 (6907), 581-582 DOI: 10.1038/419581a
MILFORD H. WOLPOFF, JOHN HAWKS, BRIGITTE SENUT, MARTIN PICKFORD, JAMES AHERN (2006). An Ape or the Ape: Is the Toumaï Cranium TM 266 a Hominid? PaleoAnthropology, 36-50
Wood, B., & Harrison, T. (2011). The evolutionary context of the first hominins Nature, 470 (7334), 347-352 DOI: 10.1038/nature09709