Menu Close

Putting meat on the bones of Au. sediba, our oldest ancestor

We now know the exact age of a species that confounds scientists. Lee Berger/University of the Witwaterstrand

Since its discovery in August 2008, the site of Malapa in Johannesburg has yielded more than 220 bones of early hominins representing at least six individuals, including the remains of babies, juveniles and adults.

And now there’s something else, something very significant.

As lead geologist on the team that found MH1 and MH2 in caves at the Cradle of Humankind World Heritage Site, I’m pleased to announce we have now accurately dated these individuals, and can reveal several findings that cast doubt on long-held theories about human evolution.

What we’ve discovered

Our findings (through detailed descriptions of the brain, hand, pelvis and foot of the fossils) place the fossils as an intermediate form between earlier Australopithicines (hominid family originating between 10 and 5 million years ago) and later species of the genus Homo, to which modern humans belong.

Evidence for this, published today in Science, is based on the remains of two individuals from the site – MH1 (a juvenile male) and MH2 (an adult female), who we think could possibly be mother and son. These superbly preserved fossils were first announced to the world in April last year.

At the time, their age was estimated at between 1.78 and 1.95 million years old. But we can now say with some certainty they are 1.977 million years old.

Courtesy Duke University.

This date places the occurrence of Au. sediba at Malapa briefly after the last occurrence of Au. africanus (represented by Mrs. Ples, a skull found in South Africa in 1947) at about 2.03 Ma – a unit of time equal to one million years ), and well before (approximately 0.2 Ma) the first definite occurrence of members of the genus Homo, in the shape of H. erectus.

Unfortunate endings

The site where the fossils were discovered is the infill of a de-roofed cave that was about 30-50 metres deep 2 million years ago.

The individuals appear to have fallen, along with other animals, into a deep cavern, landing up on the floor for a few days or weeks.

The bodies were then washed into an underground lake or pool, probably pushed there by a large rainstorm. They didn’t travel far, maybe a few metres, where they were solidified into the rock, as if thrown into quick-setting concrete.

There is evidence in the deposits that the cave experienced a major collapse shortly after the fossils were buried, and it’s possible the hominins wandered into the cave for shelter and fell into an unexpected shaft that had formed after the cave floor had collapsed.

The hominin skeletons were found with bones either in partial articulation or in close anatomical association, which suggests the bodies were only partially decomposed at the time of deposition in the lower chamber.

This, in turn, suggests they died very close in time to each other, either at the same time, or hours, days or weeks apart.

The hominins are thought to have fallen into a deep cavern.


The juvenile MH1 is around 10–13 years old in human developmental terms. He was probably a bit younger in actual age (perhaps as young as eight or nine) as he is likely to have matured faster than humans.

Mrs Ples? That’s Australopithecus africanus to you. mharrsch
The age estimate is based on modern human standards by which the eruption stages of the teeth are evaluated and the degree of development of the growth centres of the bones.

Based on the extreme wear of her teeth, MH2 is probably at least in her late twenties or early thirties but it’s very difficult to determine the age of an adult at death because her bones would have finished growing.

The estimated adult brain size based on the MH1 juvenile is approximately 440 cm3, which is slightly below the average for Au. afarensis (best known by a skeleton known as Lucy) and Au. africanus (Mrs. Ples’ species).

This suggests that, like australopiths, sediba gave birth to small-brained babies (based on the relationship of adult to neonatal brain size in chimpanzees, australopiths are thought to have given birth to babies with brains about 153 – 201 cm3 in size).

Sediba’s brain

The study of the brain shows a surprising mix of characteristics. The overall shape of the MH1 endocast (a cast made of the inside of a cranial cavity, reconstructed from a high-resolution synchrotron scan, obtained at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France) groups closely with all humans.

Given how small sediba‘s brain is (around the size of a medium-sized grapefruit), these results are consistent with a model of gradual neural (brain) reorganisation in the front part of the brain.

One of the major discoveries announced in Science is that the shape and form of sediba’s brain is not consistent with a model of gradual brain enlargement, which is what has been hypothesised previously for the transition from Australopithecus to Homo.

Sediba’s hand

The hand of MH2 represents the earliest, most complete fossil hominin hand post-dating the appearance of stone tools in the paleoanthropological record.

Almost all other fossil hominin hand bones prior to Neandertals are isolated bones that are not anatomically associated (i.e. do not belong to the same individual) and are not clearly affiliated with specific hominin species.

For the first time, the Au. sediba hand allows us to evaluate the functional morphology of the hand overall.

Sediba’s hands give evolutionary clues. Duke University
Au. sediba has many of the features that have been associated with human-like precision grip and the ability to make stone tools, including a relatively long thumb that would facilitate thumb-to-finger grips.

Importantly, Au. sediba has more features related to tool-making than the OH 7 hand that was used to originally define the “handy man” species, Homo habilis, which lived approximately 1.75 million years ago.

Au. sediba also retains morphology that suggests the hand was still capable of powerful flexion needed for climbing in trees.

Taken together, this indicates sediba still used its hand for tree climbing, but was also capable of human-like precision grips.

In comparision with the hand of Homo habilis, Au. sediba makes a better candidate for an early tool-making hominin hand and the condition from which the later Homo hand evolved.

MH2’s hand was not found in direct association with tools, but this probably reflects the fact the skeletons were washed into their burial position by a mudflow away from their original location of death.

Sediba’s pelvis

The pelvis (hip bone) of Au. sediba again represents a combination of more primitive and derived morphologies. It is short and broad like a human pelvis, creating more of a bowl shape than in australopiths.

But it retains some features of earlier hominins, such as the size of the joint that links the sacrum with the vertebral column and the length of the front part of the pelvis.

Australopithecus afarensis, or Lucy. Carlos Lorenzo

Parts of the pelvis are indistinguishable from that of humans, and it has a sigmoid shape (s-shape) along the top of the blades.

It has been surprising to discover such an advanced pelvis in such a small-brained creature.

It was previously thought that the evolution of the larger brains in early Homo were the reason the human pelvis is shaped differently to early hominins (such as Lucy, Au. afarensis) which are broader, flatter and more flaring.

Au. sediba, with its small brain, proves that at least in this lineage, this theory is wrong, and will probably be refuted in the entire human lineage.

In other words, there is now a need to look at explanations other than brain size for the origins of modern pelvic shape.

Foot and ankle

Of all the evolutionary specialisations that define humans, the foot is thought to be one of the most important.

It’s pivotal in allowing the evolution of arguably the most critical defining character of the hominins: walking upright consistently or bipedalism – and playing soccer.

Parts of both the MH1 and MH2’s ankles have been found so far, and the female MH2 ankle is one of the most complete ankles ever found.

The ankle joint is mostly human-like in form and inferred function, and there is some evidence for a human-like arch and Achilles tendon.

But Au. sediba is more ape-like. It possesses a more slender heel and a more robust shin bone than expected.

This suggests Au. sediba may have practised a unique form of bipedalism, and would have almost certainly climbed trees.

With parts of the heel being more primitive than earlier hominins such as Lucy, it may be that Au. sediba did not descend from this lineage.

No ankle has ever been described with so many primitive and advanced features, and if the bones had not been found stuck together, they may have been described as belonging to different species.

Dating the fossils

When the fossils were announced last April, the top of the fossil deposits was not exposed.

But late last year a flowstone (a limestone layer) was found that caps the deposits containing Au. sediba.

This flowstone provided a absolute age date using Uranium-Lead radiogenic isotope techniques, which was the same age as the flowstone underlying MH2.

The new date, in combination with a reverse magnetic polarity for the flowstone, meant that the normal magnetic polarity preserved by the sediments containing the fossils could not have formed during the originally postulated time period of between 1.78 and 1.95 million years ago.

Instead the fossils had to be laid down during a short-duration magnetic excursion (a significant change in the Earth’s magnetic field) which occurred 1.977 million years ago and which is estimated to have lasted for about 3,000 years, making the age constraint for the Au. sediba fossils remarkably accurate, considering the age of the fossils.

Interestingly the fossils were deposited during a period of magnetic instability on Earth, which raises the question of whether this instability may in some way have caused major changes in the environment.

We are not sure of the answer to this, but there may be severe changes in weather patterns as well as exposure to increased solar radiation as the Earth’s magnetic field fluctuates.


Some scientists feel that Lucy’s species, Au. afarensis, gave rise to Au. africanus (Mrs. Ples) and in today’s Science papers, it’s suggested that Au. africanus or something earlier, gave rise to Au. sediba.

Who came first, habilis (above) or sediba? mharrsch
There is some evidence, based upon the both primitive and advanced foot and ankle of sediba, that it does not descend from Au. africanus, but comes from some as yet unidentified lineage of early hominin.

And the very advanced nature of sediba’s hand suggests it may not give rise to Homo habilis, which although later in time, has a more primitive hand.

There is broad acceptance of the species Au. sediba among scientists as something previously unknown to science.

It is turning out to be one of the most intriguing hominins yet discovered, and it certainly shows a mosaic of features shared by both earlier and later hominins.

Our study indicates that Australopithecus sediba may be the best candidate yet to be the ancestor of Homo erectus.

Malapa is already one of the richest early hominin sites ever discovered – even though excavations have not commenced yet.

When they do, later this year, we expect to make even more remarkable finds at the site.

Want to write?

Write an article and join a growing community of more than 174,400 academics and researchers from 4,802 institutions.

Register now