Age of hominids from Sterkfontein

A recent spate of articles has carried on a debate about the age of the Sterkfontein hominids. Sterkfontein is a complicated site, including several distinct caverns and deposition layers, called members. The dating of these layers is a serious problem because of their complex stratigraphy and the lack of volcanics that could be subjected to radiometric dating. Until recently the only insights into the age of the fossils came from uranium-series dating and paleomagnetic analysis of calcite deposits in the caves.

The Sterkfontein deposits are divided into six members, and hominid have been recovered from Member 5, Member 4, and Member 2. Most of the hominid remains assigned to Australopithecus africanus come from Member 4, which was long thought to date to between 2.8 million and 2.6 million years. Before Member 5 was deposited, there was erosion on the top of Member 4, and the two are separated by an unknown period of time. This deposit is generally thought to be less than 2 million years in age, perhaps extending as recently as 1.4 million years (Kuman and Clarke 2000). In recent years, excavations lower in the deposit, including the Jacovec cavern and the Silberberg grotto, have produced hominid fossils attributable to Member 2. These were initially believed to be around 3.5 million years old.

The most important fossils from Member 2 belong to the specimen StW 573. The foot bones of this skeleton were initially found in a dump of breccia outside the cave (Clarke and Tobias 1995). The origin of the bones was traced to Member 2, and they were matched to the broken end of a tibia still in situ in the Silberberg grotto. The skeleton is now known to be largely complete, including a skull and mandible, forelimb and hindlimb elements, and much else. It appears to be considerably more complete than the "Lucy" skeleton from Hadar, AL 288-1, or any other australopithecine, but it is not yet fully excavated from the overlying breccia and flowstone. The idea that this skeleton might date to 3.5 million years was potentially very important. At this date, it would be a contemporary of A. afarensis from Laetoli and Maka (it would be earlier than the Hadar deposits). It is not clear yet whether StW 573 anatomically resembles A. afarensis or is more similar to later South African hominids, but this would certainly be an important question to answer from the respect of early hominid phylogeny.

Making Sterkfontein later

McKee (1996) suggested that Member 2 was likely immediately earlier than Member 4. His argument was that the fauna of Member 2 were all found in Member 4, but several species were absent from Makapansgat Member 3 and 4, which date to between 3.2 and 2.9 million years. He proposed that this could be explained by the chance lack of these species at Makapansgat, but viewed that possibility as less likely than the hypothesis that the species appeared after Makapansgat Member 4, to be found in the later Sterkfontein deposits.

Clarke and Tobias (1996) responded to this argument by noting the long stratigraphy of Member 3 between the Member 2 and 4 sequences, with several flowstones that must have taken a long time to deposit. They note that although Makapansgat does not preserve all the Member 2 fauna, the species that are absent are known from other African sites prior to 3.5 million years, and therefore are not of use in dating the deposits. The exception is one baboon species, Papio izodi, which is known only from Member 4 and Taung, and may therefore be rare enough to be absent from other sites.

Berger and colleagues (2002) argued that the entire Sterkfontein sequence is substantially later than had previously been thought. They base their argument on biostratigraphic and paleomagnetic considerations. They have a number of reasons for this:

  1. The presence of Equus in the deposit, which is not radiometrically dated in Africa earlier than 2.36 million years ago.
  2. In addition to Equus, several other taxa are found in Member 4 that do not have secure radiometric dates above 2.5 million years anywhere in Africa.
  3. A later date for Member 4 would suggest that the sequence of magnetic samples from the site should be displaced earlier by a reversal cycle. This would place the top of Member 2 within the Olduvai subchron, and the StW 573 hominid would then date to between 2.15 and 3.04 million years ago. If this is displaced by another cycle more recently, StW 573 would date to as recently as 1.07 to 1.95 million years.

As far as Equus, Kuman and Clarke (2000) are at pains to show that it actually may not occur in Member 4. According to them, only one equine tooth has been excavated from Member 4 in situ, with the remaining bones taken from fill that may derive from Member 5. They argue that the one tooth is insufficient evidence of the presence of the genus, considering the possibility of erosion from later deposits.

Making Sterkfontein earlier

Partridge and colleagues (2003) dated the Sterkfontein Member 2 deposits by using the radioactive decay of cosmogenic isotopes. These are created when cosmic rays from outer space interact with the elements in quartz grains near the earth's surface. In particular, aluminum-26 and beryllium-10 accumulate in quartz grains at a predictable ratio. These two isotopes have different half-lifes (26Al = 1.02 million years, 10Be = 1.93 million years), which means that once the quartz grain is buried and no longer exposed to cosmic rays, the ratio of the two isotopes changes.

Sediments near the StW 573 specimen gave a date estimate of 4.17 million years, while the orange breccia in the Jacovec Cavern gave an estimate of around 4.02 million years. These date estimates are substantially earlier than were previously estimated for these localities at the site.

It was not possible to date Member 4 in this way, because it is shallow enough that cosmic rays can still affect the quartz grains used for dating.

Partridge et al. (2003) do not present a response to Berger et al. (2002), except to note that their earlier dating "is unsustainable on stratigraphic and faunal as well as on paleomagnetic grounds" (612, note 12). In any event, there seems to be no strong biostratigraphic reason to place Member 2 at either an earlier or later date; the preserved fauna is not specific as to age.

Member 5 stratigraphy

Kuman and Clarke (2000) review the stratigraphy of Member 5. The most important hominid specimen that has been attributed to Member 5 is StW 53, a nearly complete skull that has been variably attributed to A. africanus or Homo habilis. Kuman and Clarke (2000) show that the skull derives from an area that likely is intermediate in age between Members 4 and 5 proper. They call this area the "StW 53 Infill." No artifacts derive from this area. The authors argue that the infill is likely more recent than Member 4 because of the presence in the deposit of Theropithecus oswaldi, a species found in the later Swartkrans Members 1--3, and associated with drier open grassland habitats. On this basis, they place the StW 53 Infill between 2 million years ago and 2.4 million years, which marks the earliest appearance of T. oswaldi in East Africa.

According to Kuman and Clarke (2000), Member 5 can be divided by the presence of two distinct tool industries. The Oldowan Infill dates to between around 2 million and 1.7 million years ago, and preserves 3245 excavated artifacts (Field 1999). The paleoenvironment seems to indicate a grassland. The later phase is referred to the Acheulean because of the presence of bifaces, and is placed between 1.7 and 1.4 million years ago. Like the earlier Oldowan infill, the Acheulean infill represents a predominantly grassland fauna, similar to Swartkrans.

Kuman and Clarke (2000) provide a list of hominid fossils with their probable associations in the stratigraphy. They also discuss the taxonomy of the fossils and their resemblances with elements of the earlier Member 4 and Swartkrans remains.

More on Sterkfontein

More on Makapansgat


Berger LR, Lacruz R, de Ruiter DJ. 2002. Brief communication: Revised age estimates of Australopithecus-bearing deposits at Sterkfontein, South Africa. Am J Phys Anthropol 119:192-197.

Clarke RJ, Tobias PV. 1995. Sterkfontein Member 2 foot bones of the oldest South African hominid. Science 269:521-524.

Clarke RJ, Tobias PV. 1996. Faunal evidence and Sterkfontein Member 2 foot bones of early hominid. Science 271:1301-1302.

Field AS. 1999. An analytic and comparative study of the Earlier Stone Age archaeology of the Sterkfontein Valley. Masters thesis, University of the Witswatersrand.

Kuman K, Clarke RJ. 2000. Stratigraphy, artefact industries and hominid associations for Sterkfontein Member 5. J Hum Evol 38:827-847.

McKee JK. 1996. Faunal evidence and Sterkfontein Member 2 foot bones of early hominid. Science 271:1301.

Partridge TC, Granger DE, Caffee MW, Clarke RJ. 2003. Lower Pliocene hominid remains from Sterkfontein. Science 300:607-612.