Darryl Granger

Dating Sterkfontein Australopithecus to 3.4-3.7 million years

Cosmogenic nuclide dating shows that Australopithecus-bearing cave breccias at Sterkfontein were deposited from about 3.4-3.7 million years ago, considerably older than previously supposed. Newly recognized stratigraphic unconformities at the site reconcile these older ages with previous work by showing that the faunal assemblage is mixed with a much younger member, and that dated flowstones are younger than the breccia in which they crystallized.

FULL TRANSCRIPT 

I am a geologist and geochronologist and I specialize in cosmogenic nuclides, and I want to tell you about the work that I've been doing off and on for the last 20 years or so at Sterkfontein dating the Australopithecus bearing deposits. And this is of course work that I've done as part of an interdisciplinary team. The main point that I want to get across to you today is the importance of stratigraphy. We've learned this week are people have talked this week about how important stratigraphy is for fossils to understand the deposition history, to put your fossils into their context. Stratigraphy is equally important in geochronology. We have to know what it is that we're dating. And this work at Sterkfontein, I think is a fantastic example of that. As we learned earlier this week, and I think all of the early discoveries of Hominins in Africa all came from the South Jose’ was just talking about Paranthropus that was discovered in 1938. 

I didn't list that up on here, but the Taung child in 1924 and TM1511, the first adult Australopithecus at Sterkfontein. And these were associated with lime mines that were stimulated by the need for lime and the gold fields. So, all of these sites are early mining sites. At the time, they were really difficult to date, and they remain really difficult to date. This is because in contrast to what we've heard about Turkana, there are no volcanoes in South Africa. We don't have these ashes that are smeared all over the landscape that we can use for geochronology. We can go back to some of the early geologic descriptions. Here's a picture of the site in 1936, and they still hold true today. So Houghton visited from the geological survey in 1949 and he wrote that the open working at the mine there leads down into a cavernous opening down into the cave, and part of which reddish bone-bearing breccia extends from the floor to the roof, the breccia being veined with white calcite streaks and showing some prominent, almost horizontal but irregular and thick veins of the same material. And this is important, which have undoubtedly been formed subsequent to the deposition of the breccia. So all the way back in 1949, it was recognized that there are two deposits here. There was the breccia deposited with the fossils, and then there was the later intrusion or the later deposition of calcite flow stone. 

Sterkfontein, of course, is very well known for Australopithecus. It remains the single most prolific source of Australopithecus. There have been 600 fossils, more than 600 fossils recovered from the site. We know there are at least 87 individuals recovered from Member 4. So, this air photo here shows the Member 4 workings, the surface workings of the site. Sterkfontein is in a low dolomite hill. The site used to be a cave. It's been de roofed by erosion, and we can see Member 4 and the type site there. And then Member 5 being excavated a little further to the west. There are also still caves underneath there and those caves are fossiliferous. So, we have the Silberberg Grotto where the little foot skeleton was recovered, STW 573. And then we also have a few fossils recovered from Jacovec Cavern, a separate room inside the cave. 

Now these fossils have traditionally been dated to between about 2 and 2.5 million years ago on a variety of lines of evidence. This is primarily and originally on faunal correlations. So there are young fauna there. We have Equus, Antidorcas Springbok and there's at least one juvenile pig. And we all know the pigs don't lie. There's also been dating of the flow stones with uranium lead and with paleomagnetism. I've been working there, and I use cosmogenic nuclide dating and consistently my ages have been older than that. So, what I want to tell you today is about the dating method, and I want to give you a stratigraphic interpretation that I think reconciles the dates of the flow stones, the fauna and the cosmogenics. 

So, I know that most of you are not familiar with cosmogenic nuclide dating. So I want to spend two slides going over that. Hopefully it won't bore you. I've brought in here today. You can see this flashing light up here. This is actually a neon detector and every time it flashes, there is a cosmic ray muon passing through a little piece of plastic about that big, a little scintillator that's inside there. And that just I want to help you visualize the shower of cosmic rays that are passing through us, through our bodies, through the rocks and causing nuclear reactions all the time. So cosmic rays are these really super energetic particles. They're coming from outer space. They're accelerated by supernova explosions. They hit the top of the atmosphere and they start a nuclear cascade, the shower of secondary particles that comes down and causes nuclear reactions. 

So it's shown in cartoon form right here. As it passes through the atmosphere, these nuclear reactions do things like turn Nitrogen into Carbon 14. This is the basis for radiocarbon dating. It also can convert Nitrogen and Oxygen into radioactive Beryllium, Beryllium 10 that then gets rained down onto the surface. This was the basis for dating Sahelanthropus. A lot of you have asked me about this week. I want to talk about a different application looking at the cosmic rays as they pass through mineral grains, through rocks down on the surface and they have enough energy as they pass through to convert some of the silicon into radioactive aluminum and some of the oxygen into radioactive beryllium. 

The longer a rock is exposed at the surface, every rock, the longer it's exposed at the surface, the higher the concentration of these radioactive cosmogenic isotopes. So, this cartoon up here on the top right illustrates Aluminum 26 in pink and Beryllium 10 in blue. As a rock is exposed, they accumulate these nuclides. If you now take that rock and bury it in a cave, all that burying it in a cave, you shield it from cosmic rays, all the Aluminum 26 and Beryllium 10, you started with decays away over time, the aluminum decays faster than Beryllium. So we can use the ratio of those two to date when the rock went into the cave back to about 5 million years. So this is the basis of what I call simple burial dating. We can date a bag of sand, but the problem with a bag of sand is that just as Thure was talking about with the dating of volcanics, is that okay, well maybe it's a mixture. 

Maybe you've got some old material and some young material in there. So, to address that, we came up with isochron dating method. So we date individual rocks. If you go pick up five or 10 rocks from the ground surface, they should all fall. If we plot Aluminum 26 versus Beryllium 10, they should all fall on that black curve there. If they've been in a cave for a million years, they all fall on the red curve and so forth. So we can regress a line through there, and this allows us to mathematically solve for a lot of variables. So it's a much more reliable dating method. 

Just to demonstrate the reliability, we've had isochron methods now for about a dozen years and we've applied it all over the world and it's proven very, very reliable. A couple of examples. There's a site in China where we have two basalt flows and a gravel in between them. The basalt state to 1.2 and 1.4 million years. By Ar/Ar, we get 1.36 million years plus a minus 80,000. So it agrees very well. There was just a paper published last year by Toshiyuki Fugioka and others working with Ignacio de la Torre. And at Olduvai bed two, there they actually dated the stone artifacts. There are quartzite artifacts in this artifact layer, and they form an isochron that gives an age of 1.5 million years plus or minus a quarter million, exactly in between the bracketing ages of the ashes. 

So now I want to talk about my work at Sterkfontein. So this cross section up here shows the inner part of the cave, the Silberberg Grotto, where we can see Member 1. It's a talus cone inside the cave Member 1 is sterile. It's only formed by cave breakdown as soon as the cave broke open to the surface and started admitting dead animals and sediment. Then we call that Member 2 and Member 2 forms this cap on the debris cone right there. At the time, there would've been at least a 20 meter tall vertical shaft, so this would've been a death pit. Any animal that got dropped in there or tripped and fell would've fallen 20 meters landed on top of this rubble pile and Little Foot actually fell down and rolled partway down the debris cone and it was found at this little red star right there. 

So we did simple burial dating. I was working there in the early two 2000s, published in 2003. This was one of the very first, the third or fourth set of burial dates ever published. So it was a brand new method. We got an age of about 4 million years. I can't say that the community embraced these new ages. It's a brand new dating method. It gives a surprisingly old result. A lot of you in the room were there, and especially after 2006 when Joanne Walker in 2010, when Robin Pickering dated the flow stone at the site to 2.2 million years, almost 2 million years younger than the breccia age. And I will confess, I believed at the time that there was a problem with my burial dates. This was simple burial dating. It was a bag of sand. And I said, well, it's probably a mixture. It's probably a mixture of some old stuff and some young stuff together. 

But even at the time it was recognized that the flowstone is younger than the skeleton. So, if you look at the stratigraphy here, you can see that part of the skeleton is below the flow stone and part of it is above the flowstone. The flowstone must be younger than the skeleton. And actually, if you look in more detail, the flowstone, the breccia was deposited with the skeleton. There was a cavity dissolved underneath it, the breccia collapsed and then the flowstone came in there. So there's a whole sequence of events you can see in the block there. On the bottom right, the flow stone actually encases the bones. So it very definitely postdates the deposition. In 2014, a couple of things happened that prompted me to go back to Sterkfontein and try to figure this out. Was it a mixture or not? One of these was that Lauren Brussel published a beautiful, beautiful stratigraphy of the site shown in the top right there, and he demonstrated conclusively all of these flowstones are intrusive. 

They formed in cavities that were dissolved in Member 2. Here you can see in fact in the picture there a calcified piece of the breccia that's enclosed within the flowstone. We had invented the isochron method by then. So I could actually test the method, use the method to test whether there's mixing, and then we had a major breakthrough in measurement technology. So, we could measure Aluminum 26 much better than I could before. So these were actually the very first samples that I measured with the new instrumentation. And you can see the isochron here, we regress. So, laying through the data and the age is 3.67 million years plus or minus 160,000. It agrees within measurement uncertainty with these early ages that I got of about 4 million years. 

So again, this is a surprising result, right? It shouldn't be this old and this led to a lot of criticism. What has to be my favorite criticism of all time came from Fred Grine, who at the time said that he had a dead squirrel in his backyard, right? You remember this, and I don't know if it's true or not, but you buried the dead squirrel. And if somebody finds that just because the squirrel is in 10,000 year old soil doesn't mean that the squirrel is 10,000 years old, which is of course perfectly true, but it does ignore the careful stratigraphy there. We know that the bones were deposited in a stratigraphic sequence. Jan Kramers also published a criticism in 2017, again, not questioning whether that the sediment was 3.7 million years old but are the fossils 3.7. And in his scenario, well, maybe you had an older cave that no longer exists that used to be up there full of sediment but not fossils. 

And then the fossils came in a different entrance but no sediment, and then they happened to mix together. So this is an unlikely scenario. Why would you have sediment? But no fossils mixing with fossils and no sediment. But nonetheless, it introduced doubt in the community and it's almost unfalsifiable. So, we decided to go back and date number four where most of Australopithecus fossils have been found. Now there's been a lot of work on Member 4, including some nice geochronology by Robin Pickering and others. So here's an image from Pickering and Kramers. What Robin did is she dated flowstones primarily from bore holes that were drilled around the periphery of the site. And you can see the bore holes here. And she interpreted the stratigraphy to show that the Member 4 dated between 2-2.6 million years ago. 

I admire Robin, I respect her very much in her geochronology, but I do disagree with her interpretation here, and I'll show you why. She had a few assumptions. First of all, the flowstone is horizontally continuous, and you can correlate it from core to core. Second, it's contemporaneous with the breccia. So, you had breccia deposition flowstone, breccia, flowstone, and then third, there's a low angle dip down to the southwest. So the site, she found boulders in the northeast and gravel in the southwest. So it came in from that direction and that was important. Some of the issues that we have here are first of all, the capping flowstone, the one that gives the minimum age at the top of the deposit. Here's a cross section here of the site. Member 4 is in purple. The capping flowstone is over here. And I want you to see they're separated by two meters of dolomite. 

They are not in stratigraphic contact, and there was a wall of rock that separates the two. They were probably never in stratigraphic contact. So, this actually has nothing to do with Member 4. The second thing, those of you who have hiked in the mountains know if you're on a talus cone, the boulders are at the bottom. They're distal because boulders roll and gravel falls into the cracks. So you have gravel close to the entrance and boulders far away from the entrance. That's completely backwards. And if you just go out on the ground and you look, here's a nice picture from about 1980, you can see the stratigraphy is intact in Member 4, and you can see it's dipping steeply down to the northeast, 180 degrees different. And then finally, every flow stone we've looked at has been intrusive, just like Houghton showed in 1949. So we collected samples here. 

This is the type site, the discovery site of TM1511. A lot of fossils have been recovered here. We collected from right next to where the 431 partial skeleton was found. But I want you to focus on the cross section. This is Lauren's beautiful work here. Again, Member 4 is in purple, but the top of the type site here is in orange. That's Member 5. It looks like in the cross section that they interfinger that you had Member 5 and Member 4 being deposited on top of each other in their fingering relationship. But that's not the case. What happened is Member 4 was deposited and then after a period of time it was partially dissolved by water coming into the cave, dissolved cavities, and then Member 5 came in a separate entrance and filled in those cavities. So it's an intrusive relationship and we can see that they have different compositions. 

Member 5 has a redder color. It doesn't have the dolomite blocks in it. It has a different entrance. It came in a totally different part of the cave. In fact, there's a picture of me sitting there at the top of the section and you can see my feet down here on red colored sediment, that's Member 5. Behind my head over there is the gray colored sediment in Member 4. Once you learn how to recognize it, it's not that hard. So our dating results, we sampled there. This is a north south, or sorry, an east west cross section. The type site is right here. We also collected in fossil cavern, which excavated by the miners and Jacovec. They all come out the same. The isochron is shown here, 3.4 million years, plus or minus 100,000 years more or less. All the ages are consistent. 

They all range from 3.4 to 3.6, just a little bit younger than Little Foot further down in the cave and a million years older than the flowstones. So this is important. This is actually the borehole that was dated by Robin and the flowstone right here in the middle. That's this one right here. She thought that that was the bottom of Member 4 because the dip was backwards, and it projected actually below the type site. You can see with our correct dip here that it actually projects to the top of the type site. It has nothing to do with the bottom of Member 4. 

The pigs don't lie though. We still have young fauna, we've got Equus. There's no doubt there's a young suid. There's some spring buck there. If there is even one Equus bone at Member 4,  I don't know why, but my date can't be right. But there's not the young fauna from the cave. They all come from early excavations. They don't all come. Most of them come from early excavations done by blasting, done by miners. And there was no provenance done at the time. So, there I believe mixed between Member 4 and Member 5. There was one fossil, just one bone attributed to Equus from deep within Member 4. It's right here. And what you can see is if you look at the top panel, it looks nothing like a modern zebra, right? This was an ankle bone, but it does look almost exactly like above it an ankle recovered from Makapansgat about 3 million years old. So, this is not an Equus. 

Also, it's important to recognize that there was a lot of dissolution going on when trees grew on the site, they decalcified the sediment and that allowed vertical mixing of the fossils up and down from meters near that boundary. So, it's really difficult to say from historical records whether a fossil came from Member 4 or Member 5. And if you look at the cross section here, what you'll see is up at the top of the type site, this is the type site at the top, you have almost a 50/50 mixture of Member 4 and Member 5. They didn't recognize that when they were excavating whatsoever. If you even go back to some of Tobias and Hughes excavations, they recorded the location of the excavation, but they didn't record the stratigraphic details. So, it may be impossible to know from these upper parts whether they came from Member 4 or Member 5. But what we can say is that we are sure based on this stratigraphy, that some of the fossils that came from the top here must be from Member five. They must be, it must be there. 

How old is Member 5? We dated it further west based on this manuport here to 2.2 million plus or minus 0.2. I have one unpublished date from the top of the type section. That's also about 2.2 million years. Has a huge uncertainty. I need to go back and redate that. But it's consistent with this younger age. We also know that there is Equus at the site. It must be younger than 2.3. And there is a probable Paranthropus femur that Ron Clark has identified that came from the top of that site. So, we believe the best explanation is that all of the young fauna come from Member 5 that we know existed at the site. And this has been biasing the faunal comparison since Elizabeth Verba’s fantastic work from 1970s. So, the implications we've put Australopithecus a million years older, they've got, I leave the phylogeny up to you guys out here, like the wonderful talk we saw yesterday. 

This separates Australopithecus from Paranthropus robustus and early Homo by a considerable period of time. But it puts it as part of this mid-Pliocene diversification that Johannes Haileselassie was telling us about on the first day on Monday, when we see the Burtele Foot, we see deyiremeda, we are contemporaneous with afarensis, with late anamensis. And so, it's part of this diversification in the mid-Pliocene. It also shows that the genus Australopithecus was not endemic to East Africa. We had a broad geographic spread at that time as well. And then finally, we need to go back to the other South African sites because they're all faunally compared to each other. Could they be older as well? And then finally, I just want to close acknowledging the importance of Bob Brain's work. Bob was one of the true pioneers of cave stratigraphy and cave taphonomy and has been extremely influential in all of this work in South Africa. Particularly the work that we've done at Swartkrans and Sterkfontein. It couldn't have been done without his pioneering work. And I actually have a copy of his book that stays almost permanently next to me on my desk. Thank you.

View all Geology and Geochronology videos