McBrearty and Jablonski (2005) report on the first discovery of chimpanzee fossil remains. The described fossils are three teeth: left and right upper central incisors and an upper molar, probably M1. They may well represent a single individual, although this cannot be said for certain. The paper also hints at another, "aberrant right upper third molar", but does not describe it.
The teeth are morphologically within the range of living chimpanzees and show no strong differences from them.
The similarity in the array of fossils encountered in K3 and K3' sediments suggests that Middle Pleistocene Pan and Homo lived, or at least died, in broadly similar environmental settings. Taken together, the evidence suggests a locally wooded habitat on the shore of an alternately fresh and saline-alkaline lake, fluctuating lake levels, ephemeral nearshore fluvial channels, a nearby freshwater spring, and a semi-arid climatic regime. These conditions are not unlike those found near the shore of Lake Baringo today, although dense human populations have eliminated much of the woodland that formerly supported chimpanzees and the faunal community of which they were a part.
The paper suggests a mystery: chimpanzees were living 500,000 years ago in a place where there aren't any chimpanzees now. The conclusion seems a bit of a stretch, but for all we know may have been true:
Representatives of both Homo and Pan are present in the same stratigraphic interval of the Kapthurin Formation at sites only 1 km apart, and faunal data suggest that they occupied broadly similar environments in the Middle Pleistocene. This evidence shows that in the past chimpanzees occupied regions in which the only hominoid inhabitants were thought to have been members of the human lineage. Now that chimpanzees are known to form a component of the Middle Pleistocene fauna in the Rift Valley, it is quite possible that they remain to be recognized in other portions of the fossil record there, and that chimpanzees and hominins have been sympatric since the time of their divergence.
However, the paper doesn't probe what I think may be a bigger mystery. Today's east African subspecies of chimpanzees, Pan troglodytes schweinfurthii has a range extending into Uganda, and historically they extended yet farther east. Although this is not as far as the fossil locality, it is not that long a distance. But this comparison assumes that chimpanzees 500,000 years ago had the same range as today.
Genetic evidence suggests that they didn't--at least not the present P. t. schweinfurthii. Gagneux et al. (2001) suggest that a low level of differentiation between east African (P. t. schweinfurthii) and central African (P. t. troglodytes) chimps may indicate that their time of separation is more recent than mtDNA can adequately resolve. And east African chimpanzees are highly restricted in their mtDNA variation compared to other subspecies (Goldberg et al. 1997). Together, these points may suggest a recent colonization of the eastern chimpanzee range by central African chimpanzees. Gagneux et al. (2001:891) suggest that the divergence of these populations may have occurred between 100,000 and 300,000 years ago -- a date based on the negative evidence of the resolving power of mtDNA sequences, not positive evidence.
But if today's P. t. schweinfurthii got into east Africa only recently, then what kind of individual do these fossils represent?
Here's a hypothesis: it's the chimpanzee version of a Neandertal -- a NEANDERCHIMP! An ancient chimpanzee subpsecies that no longer exists once extended across the eastern range of the species -- and in fact, even further east into the Eastern Rift. Sometime within the past 100,000 to 300,000 years, these chimpanzees were replaced by the ancestors of P. t. schweinfurthii, who came from central Africa. And these central African chimpanzees themselves may have come from the western part of the chimpanzee range during the past 500,000 to 1,000,000 years. This hypothesis envisages the spread of modern chimpanzees from a west African source within the past half million years.
On the other hand, we might just accept that alleles have spread into east Africa without the spread of populations. The limited mtDNA diversity of east African chimpanzees might be a marker of selection, rather than population replacement. This would also account for the apparent genetic similiarities between east and central African chimpanzees, as well as the sharing of markers among the west and central African populations. It's not nearly as sexy a model as the proto-Neander-chimp idea, but it does allow this wayward Middle Pleistocene individual to be part of a population ancestral to living chimpanzees. And it does give precedence to nuclear genes (about which we know little) instead of mtDNA (about which we know much, but what we know may well be biased by selection). And it acknowledges the fact that no one has yet sampled DNA diversity along most "subspecies boundaries" in chimpanzees, so we actually don't know if the subspecies are as differentiated as the distant sampling points would indicate by themselves.
So the Neanderchimp model is not the null hypothesis -- gene flow and selection in chimpanzees would be the safe guess here. But it shows a nice symmetry with the case of humans, don't you think? And it raises the central question about this fossil individual very nicely -- what is it, and which chimpanzees are its living relatives?
What about bonobos? Do they figure into this scenario? As far as these fossils go, probably not:
Specific diagnosis of isolated teeth within Pan, however, must be approached with caution, and for this reason we assign the Kapthurin Formation specimens to Pan sp. indet. Non-metric characters that have been suggested as diagnostic criteria for P. troglodytes, such as a more quadrilateral outline shape to the upper central incisor crowns and a better expressed hypocone on the maxillary molars, seem to suggest more similarity for the Kapthurin Formation fossils to P. troglodytes than to P. paniscus, but these features are variably expressed among the living species and subspecies of Pan. Although mean tooth size is known to be significantly smaller in P. paniscus than in P. troglodytes, size ranges overlap (Table 1). Furthermore, apart from the present specimens, we lack a fossil record for the Pliocene and Pleistocene from which to assess past variability within the genus, and it is feasible that the Kapthurin Formation fossils represent members of an extinct lineage within the genus Pan.
But the speciation of bonobos and common chimpanzees occurred sometime between 500,000 and 1.5 million years ago or so, meaning that this divergence probably must be explained by the same kinds of biogeographic phenomena that explain the recent divergence of chimpanzee subspecies.
Only time will tell if the Neanderchimp model will come into vogue. But you heard it here first.
Gagneux P, Gonder MK, Goldberg TL, Morin PA. 2001. Gene flow in wild chimpanzee populations: what genetic data tell us about chimpanzee movement over time and space. Phil Trans R Soc Lond B 356:889-897.
Goldberg TL, Ruvolo M. 1997. Molecular phylogenetics and historical biogeography of east African chimpanzees. Biol J Linn Soc 61:301-324.
McBrearty S, Jablonski NG. 2005. First fossil chimpanzee. Nature 437:105-108. Full text (subscription required)