Dating the Sterkfontein Fossils

See allHide authors and affiliations

Science  01 Aug 2003:
Vol. 301, Issue 5633, pp. 596-597
DOI: 10.1126/science.301.5633.596

In an article (“Great age suggested for South African hominids,” News of the Week, 25 April, p. 562) published in conjunction with our announcement of new absolute dates and early hominid specimens from Sterkfontein (“Lower Pliocene hominid remains from Sterkfontein,” T. C. Partridge et al., Research Article, 25 April, p. 607), Ann Gibbons writes, “co-author Clarke says that Little Foot does not resemble A. anamensis in crucial aspects. And he says that the new Jacovec fossils, which [co-author] Granger dated to 4 million years ago, may represent two types of australopithecines—suggesting a diversity of 4-million-year-old hominids.”

These statements by Gibbons are incorrect. Clarke told Gibbons that it was not possible at this stage to compare the australopithecines at Jacovec Cavern with Australopithecus anamensis because the most diagnostic part of A. anamensis, the mandible, was not represented in the Jacovec fossil sample. Furthermore, the Little Foot skeleton, including the mandible, is still embedded in breccia and thus cannot yet be compared directly to A. anamensis or any other Australopithecus species.

In the text of our paper, we observed that there are two forms of femur represented in the Member 4 breccia of Sterkfontein and that one of these, the long-necked form, is present in Jacovec. We state, “This suggests that two different forms of Australopithecus might be represented by the Sterkfontein femurs.” We did not say, and do not even consider, that two forms are currently represented in the Jacovec sample.

We take this opportunity to correct other misconceptions about the dating. The stratigraphy at Sterkfontein is not particularly complex and is well documented in numerous surface and underground exposures, as well as by five stratigraphic boreholes [see (2, 5) in our Research Article]. The validity of our results is, however, constrained by local relationships and not by overall stratigraphy. The hominid-bearing deposits that we dated were the first to accumulate after the cave opened to the surface via two vertical shafts. The sampling sites are in deep chambers whose dolomite roofs remain intact and exceed 20 m in thickness. These deposits overlie a basal sequence (Member 1) made up of roof-collapse blocks and insoluble components of the dolomite bedrock (mainly manganese and iron oxides) that accumulated as a fine residue during phreatic solution (i.e., below the water table); no extracavernous elements have been found within Member 1. Careful examination of the surface of the Sterkfontein hill, which consists largely of outcropping dolomite bedrock, as well as the shaft conduits themselves, has failed to reveal any infill deposits that could have contaminated soil entering the cave from the hill slopes outside. The shafts themselves are dolomite-walled and vertical and are not a source of quartz particles. It must also be pointed out that contamination from infills at depths shallower than 20 m would result in a reduction in apparent age, as such deposits would, regardless of their true age, have accumulated recent 26Al and 10Be.

The sequence within the Silberberg Grotto that contains the skeleton is well layered and includes four thick and laterally persistent layers of hard calcite flowstone that occur within a vertical interval of 3 m; after each of the intervening soil layers fell from above, it was sealed off by the next flowstone, precluding vertical mixing. The skeleton is sandwiched between the second and third flowstones, both of which were intact when our excavations began.


The article should have said that there were two types of australopithecines at Sterkfontein rather than in the Jacovec caverns, and I regret the error.

On the question of hominid diversity in Africa, I asked co-author Ron Clarke if the skeleton called Little Foot (Stw 573) was a member of Australopithecus anamensis, which lived in East Africa at the same time, and he replied in a telephone interview: “No, we cannot say it is like anamensis.” He added that he did not have the “right parts” to settle the taxonomic question, but then continued, “This was another indication that there were many species of australopithecines ranging all over Africa, just as there were multiple species of Miocene apes. There were multiple species of ape-men.”

I gave all the authors an opportunity to check facts in the final draft of the article. Granger made minor corrections on the dating, Partridge did not respond to e-mail or a telephone message for 4 days, which was after deadline, and Clarke was unavailable.

Navigate This Article