An interesting story has been making the rounds in the science press today – the unveiling of new hominid fossils, by Lee Berger, of the University of Witwatersrand in South Africa. In two related articles in Science, Berger and colleagues have tentatively classified these as belonging to a new species: Australopithecus sediba.

The fossils represent two individuals – a juvenile male and an adult female, and there is some evidence that these two individuals were associated in life (perhaps mother and son). The skeletons are remarkably complete, even in comparison to such well known specimens as Lucy. An almost complete skull, mandibles, part of a pelvis, and a complete collar bone are included, as well as limb bones, fragmented ribs, and some vertebrae.

Fossils of two individuals (juvenile male on left, adult female on right) of Australopithecus sediba. From Berger, et al "Australopithecus sediba: A New Species of Homo-Like Australopith from South Africa" Science, April 9, 2010.

The fossils date from between 1.8 to 1.95 million years old, which makes them very exciting, since fossils from this period are rare. This is also the period when the genus Homo differentiated from australopithecines, and may therefore cast some light on the otherwise hazy ancestry of Homo.

The original paper in Science (full text available with free registration at their site) describes the fossils in great detail, but the interesting features seem to be:

Skull: A. sediba is different from earlier australopithecines in being less prognathous, having a generally thinner and lighter jaw, smaller teeth. These are all characteristics trending towards Homo. In contrast, the cranial capacity has been estimated to be about 430 cc, which is smaller than the lowest currently accepted range of early Homo (510 cc).

Postcranial: The rest of the skeleton is much like earlier australopithecines, with two significant differences. First, the legs are quite long, making A. sediba somewhat taller than earlier australopithecines (estimated height about 4.5 feet). Second, the pelvis appears to be more adapted for walking. Again, both of these traits seem to foreshadow Homo. In contrast, the arms are long and australopithecine-like, as is much of the rest of the skeleton.

Australopithecus sediba skull. From Berger, et al "Australopithecus sediba: A New Species of Homo-Like Australopith from South Africa", Science, April 9, 2010.

Given this mix of australopithecine and early Homo traits, it was not clear whether these fossils represent a late australopithecine or an early Homo. The authors have chosen to classify them as a new species of Australopithecus, and they go over their reasons in great detail in the paper. I am not qualified to comment on their merit, but from a quick glance at various commentaries on the paper it seems that the anthropological community is divided on the issue.
So why is this issue important and what does it all mean? Well, first we need to understand the context of this discovery. The larger context is the appearance of Homo.  This is not well understood. The earliest member of the genus Homo is generally considered to be Homo habilis. According to the fossil record, H. habilis appeared about 2.3 million years ago, and is hence older than these fossils. However, because of the fragmentary nature of the early remains, it is difficult to be sure that the early remains do in fact represent Homo. The better known and more complete skulls, such as OH 7, OH 24, KNM ER 1805 and KNM ER 1813, are all about 1.7 to 1.9 million years old. For a while, anthropologists did not even agree whether H. habilis represented a separate species, preferring to classify skulls as either australopithecine, or those of H. erectus. These days, H. habilis is more accepted, but because of the wide variation in H. habilis fossils, many people believe that they represent more than a single species. H. rudolfensis is possibly a second ancient species of Homo, with one skull (KNM ER 1470) being dated to about 1.9 million years old.

Secondly, there is the matter of the persistence of H. habilis. According to the fossil record, H. habilis persisted as late as 1.4 million years ago. This means that for about half a million years, H. habilis co-existed with H. erectus, which is generally reckoned to have descended from H. habilis. Of course, this can be explained by punctuated equilibrium (some population of H. habilis, due to local conditions, rapidly evolved into H. erectus, but other populations of H. habilis continued to exist contemporaneously for a long time). However, if there were in fact habitats for H. habilis to survive so long, it weakens the argument for rapid speciation into H. erectus. Again, this is not a definitive argument against the descent of H. erectus from H. habilis; most anthropologists do in fact believe that H. erectus is descended from H. habilis. It’s just one of those complications that needs to be better understood, and it increases the complexity of the landscape so far as different hominid species are concerned, in which humans involved.

Another point relevant to this discussion: H. habilis had a cranial capacity about half that of modern man, but much larger than australopithecines. Typical numbers are about 650-700 cc for adults, and the low end is about 510 cc. A. sediba has a cranial capacity of about 430 cc, which puts it below the range for H. habilis. This was one of the reasons why the authors preferred to classify the new fossils as australopithecine, rather than human (in anthropology, “human” typically refers to members of the genus Homo).

So what do these fossils say about human evolution? The answer is not clear, but several possibilities can be raised:

  1. A. sediba is the direct ancestor of humans. This seems to be one of the possibilities that the authors lean towards, though by no means do they assert this as fact. To support this, they make certain claims and offer some reasoning. The claim is that previous to the discovery of these fossils, the best candidate for the ancestor of Homo would be A. africanus. Berger has made this claim before, though it is not widely accepted by other anthropologists. Since A. sediba seems to be intermediate between A. africanus and H. habilis in terms of physical characteristics, the argument could be made that the line of descent is A. africanus to A. sediba to H. habilis. The same theory of punctuated equilibrium could be used to explain the persistence of A. sediba to 1.8 million years ago, even though H. habilis appeared 2.3 million years ago. The chronology would then make sense: A. africanus (3.0 to 2.4 million years ago), A. sediba (? to 1.8 million years ago), and H. habilis (2.3 to 1.4 million years ago). This would mean that A. sediba originated somewhere between 2.4 to 2.3 million years ago, which is quite possible, though of course, the only specimens we have are these two fossils dated about 1.8 million years old.
  2. A. sediba is a cousin instead of an ancestor. In this scenario, both A. sediba and H. habilis shared a common ancestor (some other species of Australopithecus). Similarities between A. sediba and H. habilis (such as the more human-like pelvis and skull) could be explained by either positing a more human-like australopithecine ancestor, which is yet to be discovered, or by convergence. After all, the climate and vegetation were changing, grasslands were becoming more common, and these changes might very well have affected two related species which shared the same habitat.
  3. A. sediba is really not an australopithecine, but some ancestral species of Homo. In favor of this theory are a few things – the relatively young age (we know that H. habilis and probably H. erectus were already around at the time of these fossils), the previously mentioned wide variation in H. habilis specimens (which makes it more acceptable to think that this is just yet another variant), and the many human-like characteristics of these new fossils. Against that is the fact that the cranial capacity is smaller than we had previously accepted as the lower limit for Homo, and that the skeleton retains more primitive characteristics than are seen in Homo.

So at this point, it’s hard to say what the true situation is. If we classify these fossils as Homo, it could easily start a debate over whether earlier fossils are correctly classified. As I mentioned earlier, some of the older H. habilis fossils are quite fragmentary, and there have been arguments made in the past that some of them should really be considered australopithecines.

One problem in anthropology is that because of the scarcity of fossils, a lot of classifications are based on “type specimens” rather than any solid statistics. There is a large range of variability in any population (consider humans today, for example). In the absence of a sufficient number of fossils to quantify the extent of variability, anthropologists pick certain fossils with well-defined characteristics as “type specimens”, or “typical” of a certain species. Then when new fossils are found, their classification becomes a matter of relating them to known type specimens, setting up a chain of inferences. If you knock out a link in that chain (for example, declaring some early H. habilis specimen to be australopithecine instead), then there is a cascading effect on the classification of many other specimens, which were in part classified based on some similarity to these fossils.

In short, if we classify the new fossils as yet another variant of H. habilis, we will need to do some rethinking about other early specimens of H. habilis. Rethinking is always good, so this is not a problem. In fact, I am sure that right now there are many anthropologists busily thinking and writing away on just such issues.

Better answers will have to await more fossils. In the meantime, we have added to the richness and complexity of human evolution, and specifically to the period around 1.8 to 2.0 million years ago, which happens to be relatively fossil-scarce. We now have yet another species wandering around Africa at this time period, which is a critical period for the emergence of H. erectus, and the branch of the evolutionary tree leading to us.

EDIT [4/10/2010]: Scientific American has a story up quoting Donald Johanson (the discoverer of Lucy), in which he opines that (1) these fossils have been misclassified as Australopithecus, they are really a new species of Homo, and (2) he thinks it unlikely that this species descended from Australopithecus africanus.