Denise F. Su

Paleoenvironments of Australopithecus afarensis and implications for hominin evolution

Australopithecus afarensis was a long-lived and widespread species, found at sites in Kenya, Ethiopia, and Tanzania that range in age from 3-3.8 Ma. This talk will look at the habitat diversity of A. afarensis and its implications for hominin evolution.

FULL TRANSCRIPT

Thank you so much for inviting me, Lawrence, and Fred, and of course Alicia for doing all the organization. So I don't, unlike many of you, I actually don't know Richard, I didn't know him, never met him. But as I sat in the audience over the last few days and listened to all of you talk about him and what your memories about what he's done, it actually occurred to me that Richard and the Leakey family in general has had a huge impact on my life and my career and why I'm standing here today talking to you about Australopithecus afarensis. I actually, funnily enough, first my curiosity for human origin study was sparked when I was four, when I saw a photo of Marian Lewis at Olduvai, and that was really my first contact, and it was something that I retained interest in. And then I was going to go off and be a doctor. 

I think I'm okay to make this joke right now, but all good Asian daughters, I was going to go be a doctor and then I was in college, and I took introduction to biological anthropology because I thought, well, I just really, I should do something for my social science requirement that I would enjoy. I want to do something in human evolution. So, I did that was, too bad Augustine is not here because he was my professor. He was a newly minted PhD. And what really, really cemented the rest of my questions and the things I want to do was actually going to the Koobi Fora Field School where I found my very first fossil, primate proximal radius. Still remember that. And just that sense of excitement of discovery. 

And this was a while ago, so this is not the TBI Origin School, but the opportunity that I got as a 20 year old to go off into Kenya and actually find parts of our origin. I mean, I can think a lot of you say in the audience understand that, but that was it. I came back, I told my parents, I'm not going to med school, sorry. And it really was it. I have incredibly supportive parents obviously, so that really, even though I never met Richard, what he has done in establishing these infrastructures really made a difference. It made a huge difference. And I just think about, as I hear all of you talking about TBI and all the infrastructure building, how many more curious four year olds will be inspired and really continue this on? So perhaps that should have been my ending rather than my introduction. That sounded really inspiring. 

But that has taken me here thinking about paleoenvironments and A. afarensis, and I've spent a lot of time actually thinking about this. And a lot of this work is also my colleague and Kay Reed's work on afarensis. We actually have known Kay Reed for a really long time since I was a graduate student. And so just a lot of our interests overlap. And so, we've been really thinking about this seriously together in terms of what does this mean for afarensis. A. afarensis has come up long-lived species. So it covers this wide range in terms of both geography and chronology. It ranges all the way from Tanzania into Ethiopia. 

And one of the most remarkable things about afarensis is that not just that it's wide ranging both in time and space, but that there's quite a lot of it for an early hominin. I have arguments with my sister all the time, she's a doctor, she's like, how can you say all that with a sample size of one? What is wrong with you? But there's quite a lot of it, right? Hundreds of specimens. And they are from actually quite different. They're from different localities too, right? It's not just a lot of specimens from one point in time or space. It's a lot of specimens in different places at different times. So that's why I actually spent so much time thinking about afarensis and have done for quite a long time. And I actually started working on afarensis when I was a grad student at Laetoli. 

And I'll talk a little bit about Laetoli because it's just a really cool weird place compared to the other afarensis sites. So, because you have different afarensis localities at different places during different times, I think that we can start thinking about, well, what were they doing in those places? What are these different populations doing? How did their local habitats affect them? And I think at its core, that's sort of, for me, certainly that's what I'm really interested in is I leave the species designations to the systemists and taxonomist. I am really, really interested in what were they doing and how does that affect our understanding of our evolution? 

And so we can sort of ask two pretty basic questions under this. First is how do the ecological conditions of different sites compare to each other? I mean, are they the same? Are they different? Intermediates? overlaps? Like what's going on? And then the second is that, well, what did this mean for these local populations of afarensis? If there are differences, how did that impact them? And so, we took a look at all the different sites for afarensis has been said to be, some of them have more than other sites that's grayed out. Awash are not included in the analysis I'm going to be talking about because we don't actually have published faunal lists for them. But afarensis has been reported from there. And you can see from this map how disparate the distribution actually is. And they do cover quite a wide range of time too. Hadar is the youngest, Laetoli is the oldest, and then with Woranso-Mille sort of somewhere in between. And then there's a couple of other older sites as well. And so, the way that we did this, I mean we're thinking about a lot of different methodologies, but the ones I want to talk about here today is really looking at these afarensis sites within a context of modern mammal communities. We're all paleontologists in this room. So for us, our starting point is always what we see around us. This helps us understand what the processes and relationships are. 

That to me is really what we're trying to get at, understanding these processes that then allow us to make interpretations of our past. And so, this is a map of 206, thereabouts, I think modern mammal communities from all over Africa that Kay Reed and her lab over the last 10 years, 15 years have compiled in terms of their follow list and the functional traits that Tara talked about. So thank you, Tara. Actually, that means I don't have to go into a lot of detail and then what their functional traits were. And so this is just a map showing you all of the different modern communities that are included. And obviously they span quite a large geographic distance. And as Darryl also mentioned that there are certain groups that we just didn't include because well, rodents are usually not because there's just so much. So there's differences in collections in paleontology sites with rodents, they preserve slightly differently. Their taphonomic history is different. And so in this case, we actually focused on just herbivores in the sense that non carnivores basically, and the ones that are bigger than 500 grams in body size. 

And we took the functional traits, the ones that have to do with substrate use and diet. And we ran them through this canonical correspondence analysis. And these are all the modern sites. And what you see are sort of these, I should have done the X axis first. So, X-axis first, ignore the Y-axis for now the X-axis where it's really the variation of dimensions really due to rainfall. And so the ones on the far right basically are rainforest sites, a high annual rainfall, low seasonality. The one on the left are highly seasonal, low rainfall. And on the Y axis, it really seems to be capturing how much water there is on the surface around the year round. And what that means really is are there large permanent river systems, lake systems? Because the relationship between rainfall and water availability in terms of what happens on the vegetation is actually really, really important. 

If you have a lot of rainfall, then it matters a little bit less whether or not you have these big water systems on the landscape in terms of trees because they need a lot of water. Whereas if you don't have a lot of rainfall year round, then having that large river system is actually really, really, really important. And we know from looking at modern ecosystems that in these really sort of semi-arid habitats environments where there's not a lot of rainfalls, high seasonality, those water systems are really important in terms of organizing the vegetation on the surface and what's there, how is it distributed on a landscape. 

And so this is something that I do want to point out because a lot of times when we say someplace is very dry, we're sort of thinking really dry as in all-encompassing dry, but it may not actually be dry, maybe dry in terms of rainfall, but it may not actually really be dry if there is. And from an animal's perspective, I mean if there's actually this large water system there that then with all the vegetation that comes along with that water system, so basically surface water ameliorates any sort of impacts of seasonal aridity, what about the fossils? Where are they? Well, here they are, these are the afarensis sites. It's really interesting when I look at this now because I remember when I first started using community structure as a way to understand patterns in faunal communities is that the fossil sites all fall outside of the modern communities. 

And I wasn't the only person who noticed. I mean, this is in many, many papers where all look fossils are different. So does this mean that really structurally speaking, they're just different? And so what does this mean for us? So there was actually quite a lot of, maybe not necessarily in publication discussions about that. There is a lot of at the bar discussions about that. But interestingly, this is falling, the fossil sites are actually falling within these modern communities. And partly there is something about, I'm working on it, what Kay and I are working on this right now, but I think part of this is because we're only using herbivores. So carnivores aren't included in this. There is something about our record of carnivores, there's a lot more, if you just look at species lists, there's a lot more small carnivores and modern communities than there are in any fossil site. Well almost any fossil site. 

And so there's something about, there's also a body size issue here. And so there's a lot more work to be done here that we're thinking through and we're working on in terms of thinking about certain taphonomic factors that's involved that really seems to inherently differentiate than the fossil and the modern communities. So let me just sort of point out where the specific fossil sites, there's Woranso-Mille, there's a Hadar localities, there's Laetoli all by lonesome, what it was before I included the Turkana Basin insights. Now it's not biased lonesome anymore. It always was. And so what this suggests if we sort of go with this interpretation based on the modern communities, is that there is a difference in terms of the, not so much a seasonality. There's a little spread, but they're not like all over. But in terms of the surface water that's available at each of these sites and that it certainly seems like from this that, okay, Woranso-Mille overall may have had sort of whether it's, it's over time or space had environments associated with a lot more water then. So, as you go up and maybe drier, but I think at this point, just thinking about, okay, so these sites have differences in terms of their water systems and so what that means in terms of the vegetation that comes along with it. And of course, that then means that there's a difference in how the fauna might interact with each other, and that includes the Hominin that they are part of that faunal community. And so what did this mean for each of these populations of afarensis? 

And so this is our attempt at looking at, I'm amongst friends, so I'm showing this. This is, this is just trying in terms of thinking about how do we use these sort of modern climatic parameters that's tied to these faunal communities to, can we get at something about the abiotic factors in the past? And there are some really phenomena. I'm so glad that I was here all so much that's being done today. So, all of you who talked anything about abiotic factors reconstructing them, you should probably expect an email from me. So this is based on a fauna and the rainfall reconstruction, and this is just a regressing for the modern communities, a rainfall against the dimension one. And so then once we have the regression, this is very simple, as I keep on telling Andrew, I am not a mathy person, so I just do the really simple things. And so this is just from a regression, the R square value is not great. 

And it does look like when Kevin did his talk and he had a slide I think of a rainfall, this is definitely sort of lower compared to what he has. So maybe let's not take a look necessarily the numbers, they do feel a little low and it's just kind of gut feeling. Nothing less publishable, or less, they're probably a little low, but just maybe looking at sort of patterns rather than the specific number. And so this is just looking at the mean annual precipitation for each of fossil localities as in their distance from what the average is for the sites. And these are ordered from oldest to youngest and the south to north, south to north? yes. 

And so what you see here is that the older sites all seem like they're wetter compared to the younger sites that there was more sort rainfall in the older sites compared to the younger sites. I'm not really sure what to make of that yet. So that's kind of the broad picture, but I like to sort of go a little bit into three sites, Woranso-Mille, Hadar, and Laetoli mostly because these are three sites that they form kind of the core sites where we have afarensis specimens from. And in Afar, region one is in Tanzania. And so, I think a lot of that, we look at bovids a lot when we think about reconstruction paleoenvironmental conditions because bovids tend to be relatively habitat specific. They're everywhere in African Plio-Pleistocene every single site has them. They're often the most common ones. Not always, but often the most common element of the fauna. 

And so, it's kind of a rough look. And yes, there is definitely, this is a very sort of indicator species. We assume that they're doing certain things based on their modern counterparts. And certainly isotope studies in the last decade or more have shown us really that it's not a one-to-one thing. And so I will put that caveat in there and there's some really, really interesting patterns that I'm excited about. But in this case, just broadly, the kinds of things that they were doing and eating. And so, what we find in terms of just the bovid abundances at each of these three sites, and you can see that there is a real difference. And this is another, again, this a correspondence analysis and there's Laetoli sort of off on the left by itself, and that it is really closely associated with these bovid taxa that are generally thought of as being much more open taxa. And then, and the Woranso-Mille and Hadar on the right. And Hadar is definitely much more associated with the redundancy. I mean, there's a lot of redundancies from certain levels of Hadar, like a lot. So it's not so surprising that it would be pulled that way. And then Woranso-Mille really is looking like there's a much bigger sort of wooded component to it. 

So this is a kind of summary of what's been published about afarensis habitats for these areas. And what I want to point you to is that is about the water here. When you look at this, Woranso-Mille, Hadar, Dikika, all have these large permanent river or lake systems at different times; Laetoli does not. I mean, I could do a whole talk on Laetoli, but I opted not to. Probably some of you have heard it probably a gazillion times. At Laetoli it really is just a local tree. It's a seasonal river system. There are permanent springs there. There were permanent springs there. There still are permanent springs there and probably some ponds when during a really heavy rainy season. But that's it. And so even though all four of them, everyone says mosaics, right? That's like the hated word these days. They're all mosaics. Of course they're mosaics in some ways perhaps that shouldn't be so surprising. And I also hate it when people say, oh, that's not surprising. 

But the fact that you have these really large permanent rivers or lakes at these other afarensis sites and not at Laetoli, to change something about the vegetation. And you can see in those two photographs there it is what kind of difference you could get in terms of vegetation when there is or is not water. So this is stylized representation of the major, I guess components of the habitat at these three sites. Hadar, Woranso-Mille, large river or a lake, none at Laetoli. Lots of trees at Woranso-Mille, lots of edaphic grasslands at Hadar and lots of bush and shrub at Laetoli. Now they all have afarensis, but the difference in terms of the relative abundance of afarensis are those three sites is quite striking. Woranso-Mille and Hadar, they're composed of about 2%, of large mammal fauna. And definitely we have to think about collection, different collection strategies and taphonomic biases. 

And then at Laetoli, it's about 0.2%. So it's quite different. Now, Tara brought up deposition environments, and it's an excellent point, right? Because Laetoli is a completely different deposition environment from Woranso-Mille, Hadar. Woranso-Mille, Hadar are fluvial lacustrine environments, like a lot of Plio-Pleistocene sites. Laetoli is sub-aerial airflow tufts (volcanic ash). So in many ways it may be capturing, it's basically volcanic ash comes down. If you can't leave, you're dead. And if you're small enough to be totally buried, then you probably have a pretty good chance of being preserved and put into the fossil record. So that's actually, there's the agent of accumulation preservation is very different between them. I think that that actually does have a lot to do with what we're saying. Anyways, that's kind of just take a look at modern primates for just a moment. 

This look at chimpanzee density and basically chimp's a great animal really to look at too, compared to us large bodied ape. And they also range in a lot of different habitats from rainforest to quite open woodlands. And so what we see is that their density and those different habitats are really, really different too, because obviously Rainforest has more of the things that they need and that they eat. And so, this is my hypothesis that Woranso-Mille, Hadar then represents more optimal habitats for these early hominins compared to Laetoli, that the Laetoli fundamentally, even though they could live there was probably not the great place for them to be able to thrive and do really, really well. That's more of a marginal environment. 

The next step in thinking about this actually is then we've heard a lot about Hominin diversity, and I know I said I would leave the diversity issue to the taxonomists and systemists. However it matters in terms of thinking about how are these large body apes using the landscape together, which shouldn't be such a surprise to us because I mean lots of closely related species do that. I mean, right, there are sites where gorillas and chimpanzee are actually an overlap as well. But thinking about that, I think that's of the next thing that we're working on. But when you think about, take a look at this. At Hadar there's only one species, Woranso-Mille, certainly at least two, whether or not we can argue about whether our deyiremeda is a valid species or not. But certainly, Burtele Foot is doing something fundamentally different from afarensis. 

So why would that be? And this is actually a project that it's a really, really, really big project composed of lots and lots of scientists. I don't remember off the top of my head how many, but really looking at Woranso-Mille, Hadar they're not that far apart. They're only about what, 30 miles apart from each other, but actually quite different fauna, certainly different diversity of hominin. And so just looking at, I was so excited about Tara’s talk because just really thinking about, okay, from the very beginning ground up, how has this tectonics affected all this, which then landscape is affected obviously, and then hydrological systems and it comes along with that, the vegetation system and then the animals including the hominins. So with that, I'd like to thank you all very much for your attention and thank you again for your invitation.

View All Paleoecology videos