This transcript was generated automatically and lightly edited for readability.

Oliver Strimpel: This is Geology Bites with Oliver Strimpel. Every half billion years or so, most of the Earth's continental lithosphere appears to have converged to form a supercontinent, a phenomenon known as the supercontinent cycle. The most recent supercontinent was Pangea, which lasted from about 330 million years ago until about 170 million years ago. Until recently, the amalgamation and subsequent breakup of supercontinents was thought to result from processes happening on or near the Earth's surface. Specifically, the idea was that the subduction of cold oceanic plates consumes the oceans lying between the continents, a process that only stops when continents bordering the subducting oceanic plates collide and a supercontinent is formed. Once assembled, the supercontinent was thought to behave like a thermal blanket that traps the heat generated in the mantle immediately beneath it, causing it to uplift, thin, and eventually break apart.

But over the past decade or two, we've learned a great deal about the mantle, mostly from seismic tomography and computer modeling. We now believe, for example, that subduction of cold oceanic plates can reach the base of the mantle, and that the subduction of the oceans that close up to form a supercontinent may interact with the deep mantle in such a way as to produce rising plumes of hot mantle material that impinge on the base of the supercontinent and foster its breakup. This has led us to overhaul our ideas about what drives the long-term motions and configurations of the continents, and even whether we should be calling it the supercontinent cycle.

Damian Nance studies tectonic activity and the large-scale dynamics of the lithosphere. Recently, he has concentrated his research on reconstructing past geographies of the continents, and particularly on whether or not a supercontinent called Pannotia formed towards the end of the Proterozoic, before Pangea but after the previous supercontinent, called Rodinia. He is Distinguished Professor Emeritus of Geological Sciences at Ohio University. Damian Nance, welcome to Geology Bites.

Damian Nance: Thank you, Oliver. It's a great pleasure to be here.

Oliver Strimpel: Could you remind us as to how we reconstruct the past positions of the continents, and thus infer the existence of supercontinents? Let's start with the most direct evidence.

Damian Nance: Well, the two most important sources of data come from paleomagnetism, which allows us to reconstruct the former position of continents at a particular time, at least with respect to latitude, and from absolute age constraints, which allow us to determine when that time was. So those are the two key areas. But reconstructing supercontinents is a bit like putting together a jigsaw puzzle. The geologic picture, if you like, on each continent must also make sense in their reconstructed positions.

Oliver Strimpel: There are also several indirect sources of evidence, or proxies, for the existence of supercontinents. Can you describe the most important of those?

Damian Nance: Yes. In fact there are a lot, but the two most important are the most obvious. Since the assembly of a supercontinent requires continents to collide, their assembly should coincide with global orogeny and the things that go with it, like granitoid magmatism and zircon age peaks. Conversely, supercontinent breakup requires continents to rift, and so it should be marked by the things that indicate rifting, such as mafic dike swarms and the emplacement of, usually hotspot-related, large igneous provinces, which are referred to as LIPs. But assembly and breakup influence a whole host of things. They should, for example, have a profound effect on sea level and climate, and therefore are likely to influence things like extinctions and major evolutionary radiations. They also have major impacts on things like the sedimentary record, the chemistry of seawater, the composition of the atmosphere; the list goes on and on.

Oliver Strimpel: Do all these proxies, then, approximately line up and agree as to when a particular supercontinent formed and then broke up?

Damian Nance: They certainly do for Pangea. They appear to for Pannotia, and those that we can use for earlier supercontinents appear to coincide. Some we lose. It's very difficult to judge sea level as you get further back in time, and things like major evolutionary radiations and extinctions become very difficult when all you're dealing with is little tiny creatures. But yes, they do.

Oliver Strimpel: In my introduction, I briefly mentioned that until recently we explained the formation and breakup of supercontinents in terms of surface processes, or near-surface processes. Can you fill out that picture a bit more?

Damian Nance: Yes. Well, you have to remember that when the idea of the supercontinent cycle was first introduced, we knew precious little about the mantle. We were of the view that supercontinents, as you mentioned, acted like thermal blankets on the underlying mantle, that they trapped mantle heat beneath them, and that that caused the supercontinent to thermally uplift and eventually break up. That process may be part of the story, but I think today we would say it is only part of the story. On the other hand, the oceans that open when a supercontinent breaks up can be expected to eventually close. They're going to subduct and eventually close, and hence they may reunite another supercontinent. So we were sort of of the view that supercontinents sowed the seeds of their own destruction, if you like, and would eventually break up, but in doing so would create new oceans that themselves had limited life expectancies and would ultimately close. But the cycle was very much viewed as a lithospheric, upper-mantle process. Everything we were talking about was relatively shallow, and we thought of it as a governing mechanism for plate tectonics. Things like what happened to subducting slabs after they subducted, and what was going on in the deep mantle, were completely unknown quantities at that time.

Oliver Strimpel: OK, so we've learned a lot more about the mantle in the last few decades, and this has indeed caused us to rethink the picture. First of all, how did we actually gain this new knowledge of the mantle? And then, what did it reveal about the mantle?

Damian Nance: As you mentioned, the breakthroughs really came from two areas: from seismic tomography and from computer modeling. For example, seismic tomography, which is really just a means of X-raying the Earth's deep interior, revealed that subducting oceanic slabs didn't simply dissipate in the upper mantle, as was generally presumed, but actually sank all the way down to the core-mantle boundary. And at the core-mantle boundary, seismic tomography revealed the existence of two thermochemical features that were antipodal to each other, which affected seismic velocities and became known as large low-shear-velocity provinces, or LLSVPs.

Oliver Strimpel: I'll interject here that while seismic tomography reveals the enormous extent of these large low-shear-velocity provinces, each of which extends several thousand kilometers across and rises about 1,000 kilometers up from the core-mantle boundary, we don't really know why the seismic shear waves slow down there. It could be a higher temperature, a different, denser composition, or some of both.

Damian Nance: And importantly, one of these LLSVPs was centered directly under the former position of Pangea, so that in itself strongly suggested a link between supercontinents and the deep mantle, which was something we certainly never considered. That connection has in fact been strengthened with the recognition that most of the hotspots that formed in the wake of Pangea's breakup actually represented the surface manifestation of mantle plumes that emanated from the edges of these LLSVPs. Again, another connection between the deep mantle and Pangea's breakup, suggesting a deep-mantle involvement which we hadn't considered.

Oliver Strimpel: How, then, do we use this knowledge to look at the mechanism of supercontinent formation, and particularly of the supercontinent cycle?

Damian Nance: It's hugely important, because it demonstrates a link between supercontinents, or the supercontinent cycle, and the deep mantle. And in doing that, it elevates the importance of the cycle significantly, from simply being a near-surface phenomenon that influences or controls plate tectonics to being a whole-mantle process. What we take from this is that the assembly of supercontinents somehow influences the mantle in a way that causes a major region of downwelling to become a major region of upwelling. So the mantle downwelling, which we would expect to be associated with the closing oceans during the assembly of a supercontinent, somehow evolves into mantle upwelling, which is evident in the rise of the mantle plumes responsible for the breakup of the supercontinent. So somehow the cycle has taken an area of major mantle downwelling and caused it to evolve into one of upwelling.

So the original idea of supercontinents acting as thermal blankets on the underlying mantle may still be part of the story, and maybe an important part of the story, but that would not involve the deep mantle. So it does not explain this relationship, and instead it is more likely that this change is linked to the subducting slabs of the oceans that close to form a supercontinent. If you think about it, as you are bringing a supercontinent together, you've got subduction in the closing oceans, and subducting slabs sinking down towards the core-mantle boundary beneath what will be the future position of the supercontinent once it assembles. So that process clearly contributes to the downwelling beneath the supercontinent. But once these subducting slabs reach the core-mantle boundary, they somehow manage to cause plumes to rise beneath the supercontinent that foster its breakup.

Now, exactly how they do this is not yet known. There are a variety of possibilities, all or none of which may be correct. They may interact, for example, with an existing LLSVP in such a way as to trigger the rise of mantle plumes at its edges. Or they may actually form an LLSVP at the core-mantle boundary, and in doing so generate edge-effect mantle plumes. Or they may simply act as thermal blankets themselves, but this time on the core, ponding heat that then rises in the form of mantle plumes.

Now, in addition to all that, once a supercontinent assembles, all of the subduction that was going on in the closing oceans beforehand stops and must move. It has to relocate, and it will relocate to the edges of the supercontinent, to the ocean surrounding the supercontinent. So imagine that you now have a supercontinent with a sort of girdle of subduction all the way around it, or most of the way around it, and a bunch of new subducting oceanic slabs now sinking down towards the core-mantle boundary as a curtain around the supercontinent. That downward motion alone might be sufficient to cause a return flow, in the form of upwelling beneath the supercontinent. So that's another way you might be able to convert this downwelling to an upwelling. But whatever the path is, the effect is to cause the mantle downwelling associated with assembly to evolve into the mantle upwelling responsible for breakup.

Oliver Strimpel: Why should the subduction, when it shuts off underneath the downwelling that has brought the supercontinent together... What's the mechanism? Why should that flip into this curtain, or girdle, around the supercontinent? Why doesn't the subduction just shut off and happen in some random place somewhere else on the planet?

Damian Nance: There is a tendency of subduction systems to jump back towards the ocean-continent margin. So there is a likelihood, and certainly modeling shows this, that the subduction system would bounce out to the margins of the supercontinent. This is certainly what happened with Pangea. It was almost entirely surrounded by subduction systems, and some of those still survive in the modern Pacific Ocean.

Oliver Strimpel: So you said that we find that there's one of these large low-shear-velocity provinces sitting underneath the former location of Pangea. I wasn't clear whether that province was thought to have predated the formation of Pangea and somehow or other caused it to form above it, or the other way round. What exactly is the connection thought to be?

Damian Nance: We're not certain about that. What we can see, however, is that there is an amazing coincidence between the position of one LLSVP and the reconstructed position of Pangea, which sits slap bang on top of it. It currently sits under Africa, which is the one piece of Pangea that hasn't moved very much since it broke up. Everything has moved away from Africa, and Africa hasn't done very much, and the current LLSVP in that hemisphere sits right under Africa.

Oliver Strimpel: So it's a kind of correlation that cries out for explanation, and we don't really have it yet.

Damian Nance: Exactly so. And when you couple that with the fact that all of these hotspots that are clearly linked to Pangea's breakup coincide with the edges of this thing, that likewise begs an explanation. We don't have one yet. We have ideas, but we certainly don't know yet. And I think that is probably where the future of this whole story will take us: into trying to resolve those kinds of issues in today's world, rather than trying to do it in the distant past.

Oliver Strimpel: What role is computer modeling playing in helping us understand what's going on here?

Damian Nance: Computer modeling confirms the idea that supercontinents form over areas of mantle downwelling, but that they preferentially do so in what is known as a degree-one structure. That is, a supercontinent will form over an area of mantle downwelling that is antipodal to an area of upwelling. So you have a mantle with a relatively simple overall convection system: a single area of downwelling and an antipodal area of upwelling. But as the supercontinent sits there, the effect of the amalgamation is to start generating these plumes at the core-mantle boundary underneath it, which of course rise and foster its breakup. But in doing so, they change the overall structure of the mantle from a degree-one structure to what is known as degree two. That is a mantle with antipodal areas of upwelling, because the original area of upwelling is still there, and you've now created another one underneath the supercontinent, antipodal to it. Between these, what you see is a girdle of downwelling, which is where subduction is going to be focused. It forms a girdle all the way around, and it's referred to as the subduction girdle.

And interestingly, that then bears on how supercontinents might assemble after they break up, because in a world like that, once your supercontinent breaks up, the dispersing continents are going to move towards an area of mantle downwelling, and that area of mantle downwelling is now this girdle. So the dispersing continents might be expected to collect over this girdle of mantle downwelling, and what they do then depends on what that girdle does in terms of movement. If it moves away from the former supercontinent by a process of rollback, for example, like we see in the western Pacific, it may continue to move until it reaches the other pole, antipodal to the former supercontinent. In that case, the continents are going to coalesce there, and you will create a new supercontinent antipodal to the former one. And in doing so, you will have closed the ocean surrounding the former supercontinent. So in the modern world, if we think about the breakup of Pangea, the effect of that would be to close the Pacific and reunite the continents as a result. That's referred to as extroversion: closure of the ocean surrounding the supercontinent.

Conversely, this girdle may collapse on itself, and the continents will then come back to more or less their former positions, and you will recreate a supercontinent in much the same position as the one you've just broken up. That process is known as introversion. In today's world, that would be like saying the next supercontinent forms as a result of the closure of the Atlantic Ocean, which would reunite the continents to form a supercontinent in much the same position that Pangea was in.

Now, a third possibility is that the girdle doesn't move, and that the supercontinent effectively assembles on the girdle somewhere. If that's the case, then the new supercontinent forms at about 90 degrees to the former one, and that's referred to as orthoversion. In today's world, that would be akin to forming a supercontinent by closing up the Arctic Ocean and closing up the Caribbean, for example, and bringing the continents together that way. So those are the three methods of reuniting a supercontinent that are often bandied about at the moment.

Oliver Strimpel: Do we know enough about the actual positions of former supercontinents, going back to Pangea, possibly Pannotia, Rodinia, maybe Columbia, and before that, to know whether they did indeed reassemble in the same place, antipodally, or at 90 degrees? In other words, which of these three theories is borne out by the data?

Damian Nance: There are certainly folks out there who have argued yes, that the relationship has been one of 90 degrees, and that we see orthoversion in the past. When we first advocated the cycle, the one we thought most likely was introversion. And the reason was not because we had any great idea of how these things happened, but because we simply looked at Pangea, and Pangea formed by the closure of oceans like the Iapetus, the Tornquist, the Rheic, et cetera. These were all oceans that formed when what we thought was the previous supercontinent, Pannotia, broke up. So from that perspective, it looked as if Pangea had formed by closing oceans produced by the breakup of the previous supercontinent, and that's introversion. But in truth, I don't think we can be specific about it.

Oliver Strimpel: The relative positions of successive supercontinents seem to be quite an important observation for resolving these large-scale mantle processes that we're talking about here, and I'm wondering whether the global distribution of these relict slab graveyards might give us some indication, or whether indeed it's the computer modeling that can come to the rescue here.

Damian Nance: I think it's more likely to be the computer modeling. You could argue, for example, that the reason we have antipodal slab graveyards, or large low-velocity provinces, is that one formed under Pangea and the other formed under the previous supercontinent and hasn't dissipated yet. And if that were the case, of course, we'd be dealing with extroversion, because they're on opposite poles. Our problem with establishing the precise positions of supercontinents is that we don't have many longitudinal constraints. Paleomagnetism will get us latitudinal constraints, but it's often rather difficult to know whether we're on this side of the world or that side of the world. So exactly positioning one supercontinent with respect to another is not an easy process.

And on top of that, we have this issue of true polar wander, which may very well impact all supercontinents. That is a feature of physics that requires the mass imbalance that is produced when you put all the continents together in one place to cause the Earth to rotate in such a fashion as to bring that mass imbalance to the equator. It's a whole-Earth rotation of the mantle and the lithosphere, detaching somewhere at the core-mantle boundary, which brings the entire mass of a supercontinent, say one that assembled at the pole... This process of true polar wander would slowly but surely bring it down to the equator, and in fact I think it brings it down quite fast. I believe the rates of true polar wander are quite a bit faster than the rates of plate motion, for example. So that means that all supercontinents, given enough time, will end up on the equator, regardless of where they formed.

Oliver Strimpel: There is a lot of debate as to whether Pannotia existed during a period between about 630 million years ago and 530 million years ago, pretty much midway between Rodinia and Pangea, and you're an advocate of its existence. What brought you to this conclusion?

Damian Nance: Well, I would certainly like to think Pannotia existed, but I may be wrong. And interestingly, whether or not it did, Pannotia is central to some of the issues we've been talking about. It's also a bit ironic, because poor old Pannotia was the first pre-Pangean supercontinent to be proposed, and when we were putting the case for the supercontinent cycle together, it seemed to us the one for which the case was strongest. But the problem is that our two key tools for establishing supercontinents, namely paleomagnetism and absolute age dating, have failed to establish its existence. Paleomagnetic data for the Ediacaran, that is, the period during which this supercontinent would have existed if it did, are very inconclusive. The data are notoriously dispersed, and that may well mean that the Earth's magnetic field was misbehaving at that time. There may have been true polar wander going on, or it may not have been a dipole, but something odd was going on with the magnetic field, which means it's difficult to use paleomagnetic data to say yea or nay. If we look at the age dating, the problem there is that the age dating suggests that the supercontinent was breaking up before it ever got together. And if that's the case, then it never really existed. And even if it did, I think even I would concede that if it did exist, it was a very short-lived supercontinent.

So why did we think it was a supercontinent? The reason is that the proxy signals are unmistakable, and these include all the ones that you'd expect: things like global orogeny, a zircon age peak, evidence of widespread rifting, and mafic dike swarms and LIPs, and changes in sea level and climate, and extinctions and radiations, and just a host of geochemical and isotopic tracers. They're all there. The problem is that while they are collectively compelling, they are individually contestable, and therefore they cannot be considered definitive. For example, you could argue that the orogenic activity and the zircon age peak, which are very real, record the assembly of Gondwana, not Pannotia; and, conversely, that the evidence of rifting, LIPs, et cetera, was just a continuation of the breakup of Rodinia and not the breakup of a subsequent supercontinent. So there is an argument against each of them that can be made, so that they're contestable. So although the proxy evidence, which was the only evidence we had at the time, was strong, it is not definitive.

But interestingly, even if Pannotia did not exist, what this proxy evidence suggests is that there is actually a far more important question going on here, and that is: is a full-blown supercontinent needed to make the supercontinent cycle work? Or might a close gathering of continents be sufficient to change mantle circulation in the way we've been talking about and cause the converging continents to reverse their convergence?

Oliver Strimpel: So the key here is to convert the downwelling that's underneath where the newly forming supercontinent is forming into an upwelling. And that's why you're really saying this is perhaps more of a deep mantle cycle rather than a supercontinent cycle, which implies a surface cycle. And the question then is, I suppose, what does it take to get that reversal? Do you actually have to close them up completely and produce that thermal blanket with no cracks or gaps in it, or is just proximity good enough?

Damian Nance: Exactly so. And if you don't need a supercontinent, then it isn't a supercontinent cycle. But although it involves the deep mantle, as you just mentioned, it is not a mantle cycle either, because the supercontinent, or this near-supercontinent, obviously plays a vital role. So it is a mantle-lithosphere cycle, but I think we're going to have to come up with a sexier title than that if we want it to catch on.

Oliver Strimpel: Why are we so concerned as to whether Pannotia actually happened or not?

Damian Nance: The existence of Pannotia is critical for a number of reasons. If we don't actually need a supercontinent, and Pannotia is an example where we didn't have one and yet we changed mantle circulation, then the question is, are there other times when this has happened? Do we have other proximate supercontinents that have affected the cycle in the past that we haven't even thought about? There are other key issues. The existence or nonexistence of Pannotia obviously affects the interval between supercontinents, and therefore it influences whether or not the cycle has experienced secular change through time. Without Pannotia, the interval between supercontinents has remained pretty much constant since the end of the Archean, at around 600 or 650 million years or so, suggesting a broadly steady-state cycle. If we put Pannotia in the picture and say it was a supercontinent, then the interval looks to be shortening, towards perhaps 300 million years or so. This would imply that the cycle is accelerating, and if that's the case, it's telling us something very important about mantle evolution.

And the other thing that it influences is a question that geologists are tantalized by, and that is: OK, we've got a cycle, so when's the next one? When does the next supercontinent assemble? Well, that depends very much on the interval between them. Without Pannotia, we can expect another supercontinent in 300 million years or so, but with Pannotia it might form in as little as 100 million years, so that's a significant difference. So there are lots of quite important questions hinging on the existence or nonexistence of Pannotia. And I think perhaps more important than whether it actually all came together to form a supercontinent is the answer to the question: did Pannotia, whether it formed or not, impact mantle circulation in such a way as to turn an area of downwelling into upwelling? If it did, then it is a supercontinent as far as the supercontinent cycle is concerned, regardless of whether or not it is a supercontinent by definition.

Oliver Strimpel: I'm struck that these ideas are based on rather small-number statistics, and some of the sources of evidence we discussed have significant uncertainties associated with them. I'm wondering what observations, or computer modeling, or other line of research you would most like to see, to help us get more confidence that we know what the long-term dynamics of the continents are and what drives them.

Damian Nance: Well, you're absolutely right, I'm afraid. The topic is plagued by the fallacy of small numbers, but we have to work with what we've got. And, as you mentioned, it's also plagued by uncertainties associated with the proxy record, which bedevil its interpretation. So there are a lot of issues here. But as to where we go from here, as I mentioned before, I think the answer lies not in scratching our heads over deep time, but in the modern world, and in the improvement of the techniques we already have for studying the thermal structure of the mantle and modeling its evolution through time. I'm fairly confident that within the next five years, certainly within the next decade, the increased resolution in seismic tomography and increased sophistication in modeling will allow us to understand the relationship between things like plate tectonics, mantle circulation, subducting slabs and mantle plumes. This will finally come together in a comprehensive picture, and once we've got that, it will provide us with a powerful new insight for interpreting the dynamics of the Earth in deep time. So once we've sorted out the present, we can go back and have a look at all this information and see if we can't make much better sense of it.

Oliver Strimpel: Damian Nance, thank you very much.

Damian Nance: Well, thank you, Oliver. It's been a real pleasure.

Oliver Strimpel: To see pictures and illustrations that support this podcast, go to geologybites.com, where you'll also find transcripts and a subject matter index of all the episodes. There you can also give me feedback, which I welcome, as well as sign up to get my emails about new episodes.