Bruce Levell on Bias in the Sedimentary Record
Transcript
Note, transcripts are not fully edited for grammar or spelling.
Oliver Strimpel
This is Geology Bites with Oliver Strimpel. In 1788, James Hutton spotted a spectacular unconformity at Siccar Point on the east coast of Scotland. He was the first to understand that, although two sedimentary rock types were in contact with each other there, the boundary between them represented a gap in the geological record. The rocks below, which contained marine fossils, were deposited in a deep ocean during the Silurian. But only after these rocks had been folded, uplifted, and eroded, where the overlying rocks deposited in a Devonian desert. In this very clear-cut case, 80 million years are missing from the record. But the gaps in the sedimentary record are not always so obvious. Indeed, sedimentary sequences that look continuous and complete can be quite the opposite. Clearly, if we want to read the sedimentological record to make accurate inferences about the past, we need to recognize and interpret the gaps in the record, as well as the parts of the record that survive to the present day. How do depositional mechanisms and subsequent events alter or destroy parts of the record? More fundamentally, how can we tell that the record is good enough to make any solid inferences about the geological past? Bruce Levell addresses these questions by combining field work with systematic analysis based on what we know about contemporary depositional and erosional processes. Armed with an understanding of the preservational biases that can apply, he has been reexamining some widely accepted interpretations of the sedimentary record. For example, by analyzing sequences of glacially deposited rocks in Southwest Scotland, he has shown that, contrary to the widely held “Snowball Earth” hypothesis, parts of the Earth were not covered by ice, at least during the earlier of the two snowball glaciations. Bruce Levell is a Visiting Professor in the Department of Earth Sciences at the University of Oxford. Previously, he was Chief Scientist for Geology at Royal Dutch Shell. Bruce Levell, welcome to Geology Bites.
Bruce Levell
Pleased to be here, Oliver.
Oliver Strimpel
Let's start with the fundamental question I just raised. How do we know that the sedimentary record is complete enough to make solid inferences about the past?
Bruce Levell
This is a very old question that goes back to Darwin or perhaps even earlier. In the Origin of Species, he wrote:
“I look at the geological record as a history of the world imperfectly kept and written in a changing dialect. Of this history, we possess the last volume alone, relating to only two or three countries. Of this volume, only here and there a short chapter has been preserved. And of each page, only here and there a few lines.”
So, Darwin's problem was that he was being challenged to produce the intermediate forms which document the evolution of one species into another by natural selection. And his defense to not being able to produce that many examples of intermediate forms was that the record was imperfectly kept, as he wrote here. So, it's a fundamental question, and I think the answer is, we know that the record is incomplete. So, your question: “Is it complete enough?” is also very well phrased, because whether or not it's complete depends upon the question that you're asking. So, if we look at Darwin's quote and we change his metaphor a little bit, if I were to keep a diary and I were to miss the occasional entry, that might be sufficient to give you an idea of what I was doing for the year -- if you just wanted to know: What does Bruce get up to? When did he go on holiday? How much time does he spend working? But, if you wanted to know: What did he eat every day, then the entries wouldn't be complete enough to answer that question. And there might be incomplete in different ways. So, it might be that I don't write down on Friday evenings what I'm doing because I don't get round to write in my diary. So, you'd have a systematic error, or a bias in the record, that Fridays weren't preserved. So, data has been collected of the sedimentary record to ask exactly this question. This was done by a guy called Pete Sadler in the early ‘80s, and he made an extremely revealing plot which has, on the vertical axis, the accumulation rate of sediments, and, on the horizontal axis, it's got the averaging time over which the rate has been measured. So, if I measure a rate over the duration of a flood, that would be entered as maybe a week or perhaps a month on a big river, when it was in spate, and how much sediment was deposited during that event. The other extreme -- I look at sedimentary basins and I say what is the accumulation rate of sediment in, say, the Wessex Basin of southern England? And what you find on that plot is there's a linear inverse relationship between the accumulation rate and the time over which it's averaged. So, my flood deposit is accumulated relatively fast over a short period of time, maybe a meter or even two meters over a month, whereas in the sedimentary basin I'm talking about millimeters a year meters over thousands of years, maybe even 10 thousands of years in some cases. And what it says is that there are gaps at all durations in the record, and those gaps accumulate as you increase the averaging time and therefore become more important. So, the record does have the fractal gaps that Darwin postulated back in1859.
Oliver Strimpel
Are we sure that we always do know when something is missing? Might something look complete when, in fact, perhaps it isn't?
Bruce Levell
It often looks complete, when, in fact, it isn't. And that's a mistake I've made many times in my career. Typically, when you look at core data extracted from the subsurface, it's a vertical column and maybe only four inches wide, 10 centimeters wide. And I'm always surprised when I get the paleontological data, which tells me that there's time missing. There's always more missing than you think. “Mind the gap” is a very good slogan for a sedimentology.
Oliver Strimpel
What exactly would we like to learn about the past, if only we could correctly interpret the sedimentary record?
Bruce Levell
There are two main ways we use the sedimentary record to extract information. The first as a time series of information, as Darwin would have liked to. He would have liked to have had a continuous time series, albeit put together from bits of the record in different parts of the world, to make a composite time series — but, a continuous time series from the origin of life to the present day. In the modern day, with the fantastic advances in geochemistry, particularly isotope geochemistry, people are wanting to put together continuous complete records of the evolution of the Earth system. So: ocean chemistry, ocean temperature, any information about the atmosphere which you can get from, for instance, plant remains, or even trapped air in Antarctica and in the ice cores — so, continuous time series of data… And, some examples of those might be: the paleoclimate —interpolating time between points of control, where we have radiometric ages; understanding the fossil record and how it evolved; and understanding events in geological history and putting them in their time sequence. So, those could be major climate shifts; there could be major storms. There could be sea level changes, could be meteorite impacts at various scales of occurrence, from almost daily, weekly, monthly events, which are making up geological record by putting deposits down in a particular way, to these spasmodic episodic revolutionary events. And then the second way in which we use the sedimentary record is to predict essentially the geometry of sedimentary deposits, because we want to use them to extract things or to store things or to build things on. So, in that case, we want to understand the three-dimensional geometry. And, in order to understand the three-dimensional geometry, the obvious route to use is the fact that all sediments are deposited in an environment, and, that environment, if you'd have had a sort of paleo-satellite back when the rocks were being formed, would have looked like some sort of map of channels and bars and beaches and coastlines and tidal inlets and offshore fans, et cetera, et cetera. And, if we can reconstruct the depositional environment, then we can predict to a degree the three-dimensional geometry of the rock units. So, those are the kind of two main classes of use, I'd say, of the sedimentary archive.
Oliver Strimpel
OK, let's talk about the preservational bias, if you like, in the record, in the context of each of these aspects. So, first of all, what kinds of bias would tend to confound our understanding of the Paleo climate?
Bruce Levell
In order to understand the Paleo climate, we need to make a measurement which responds to temperature. The so-called proxies can be a wide variety of things. So, for example, people look at the temperature of formation of soil nodules using carbon and oxygen isotopes. You can say something about the temperature, which the soil experienced. Alternatively, you might look at changes in the rate of phytoplankton production. Or, you could measure the surface temperature of foraminifera living in the surface of the ocean. You might even be able to compare that with the temperature of foraminifera living on the bottom of the ocean, so you could get the temperature gradient in the water column. So, you want to take a series of samples which you can analyze in a lab and extract some temperature data, typically using an isotopic proxy. Now, in order to do that, you need to make a series of assumptions in order to have that as a time series of data. The first is that the record is indeed continuous at the scale at which you wish to sample it. So, if you want to reconstruct this over the time scale or 25,000 years or so, then you would need on that time scale there not to be gaps in the record. So, you could go to sedimentary sequences where, on the time scale of a few thousands of years, the record is continuous. And, the obvious one to go to is a deep water deposit, because there the preservational bias is very low: almost everything that falls down to the bottom of the ocean gets deposited, provided it doesn't get eaten, or, in the deep ocean, get dissolved. So there, your problem is less great, If you like -- you have more chance of being successful. The preservation bias is due to burrowing organisms coming across my microorganism and eating it before it has a chance to be preserved, or eating all of them from a particular time period, before they have a chance to preserve. If we come back to my soil nodule, in a fluvial sequence, the situation is entirely different. Because the soil nodule is forming outside the channel of a river. Now the channels of rivers tend to migrate. They don't stay in one place forever. They migrate sideways, and therefore they erode soil profiles, and particularly those which are closer to the level of the river are going to be destroyed, and those that are going to be preserved will be the ones, for example, on terraces, which are at a higher level away from the river itself. So you'd end up with a record which would be variably incomplete, and it might even be difficult to reconstitute, because, for instance, the terraced deposits would be topographically higher than the river channels, so they would not be contemporaneous with that particular river channel. So, you end up with a problem of sorting out the age relationships of your variable sections in which you've measured these things. And you also end up with the record that quite a lot of the sequence may be destroyed, depending on how far the river eats into the flood basin. So, those are two contrasting situations. In general, preservation bias is very strongly controlled by erosion. The most dangerous thing for a sediment is, after it's been deposited, is to be re-eroded by the next river flood or by the next storm. And it's only when it's subsided a certain increment out of harm's way (so, below the depth of the migrating river channel or below the erosional capacity of a subsequent storm), it’s been buried sufficiently deeply that it's definitively preserved. So, sediments after deposition, they’re kind of tentatively preserved, and they may spend 10, 20, 30, 40, 50, 100 thousand years in tentative preservation, and it's only subsidence which definitively preserves them by removing them from the zone of active surface currents and waves, et cetera.
Oliver Strimpel
You referred to connecting the two different time scales that we have, and indeed a major undertaking over the past decades has been the marrying of the stratigraphic time scale, which is based on the sedimentary record, with the absolute time scale, which is determined from radiometric dating. Can you quickly remind us how that is done?
Bruce Levell
Yeah. Reliable radiometric dates from sedimentary rocks are really difficult because you basically need a high temperature mineral to set the isotopic clock running and preserve the daughter products. And the best mineral that we have, the mineral of choice is clearly zircon with uranium lead dating, because it has a couple of isotopic systems when you can do internal consistency checks. And those come from felsic volcanics. And so, what you need is to find volcanic tuffs interbedded with sediments. And, ideally, those sediments should be fossil-bearing, and then you can calibrate the fossil record with the absolute age that you get from those radiometric dates. As an example, the Lower Jurassic Blue Lias of Lyme Regis, which many people know, the radiometric dates, which date part of that succession, come from Peru, because that's where Lower Jurassic ammonites are nicely interbedded with well dated tuffs in the Andes. So, what that means is that you've got a bunch of ammonites in Peru… Unfortunately, the same ammonites didn't live in Dorset. The different biogeographic provinces, so you need to connect the two groups of ammonites and their evolutionary records across two different oceans, and you do that by finding intermediate fossil groups with which both can be interbedded or via — you can find a connection. So, you might find pollen, or you might find foraminifera again, or calcareous nano-fossils, whatever it might be, that can connect those two formal provinces together, and then you end up with inferred radiometric ages for Lyme Regis. Now, you've still only got a few ages and quite a lot of rock. So, if you want to make a time series of Paleoclimate data for the Lower Jurassic in the UK, you have to interpolate between the radiometric point. And the method of choice for that is to use the rhythms which come from orbital cyclicity. So, you have the change in the eccentricity. The change in obliquity of the orbit, and the change in procession of the Earth's axis.
Oliver Strimpel
Just to explain, the Earth's orbital eccentricity is the amount by which the Earth's orbit around the sun deviates from a perfect circle, which varies with about a 100,000 year periodicity. The obliquity is the tilt of the Earth's axis of rotation with respect to the plane of the Earth's orbit around the sun. And that varies with a periodicity of about 41,000 years, and the precession is the wobble of the Earth's axis like that of a spinning top, which has about a 26,000-year period.
Bruce Levell
Yeah. And these combined to give you a series of ratios effectively between those rhythms, which you can find back in the geological record as expressed by differences in the sedimentation, if you're lucky. So, if for instance you've got difference in the rate of production of calcareous plankton, that would express itself in the limeyness of the rock, and you can measure that limeyness directly or indirectly and end up with extracting these rhythms. And then you can so-called astronomically tune the time scale so you can count the rhythms in between your radiometric dates. And you can infer an age between the radiometric age points. And, in the last edition of the geological time scale, there's been a major, major attempt to astronomically tune large parts of the stratigraphic record, and this is actually a major undertaking. But it does require you to be able to find composite sections in the record which are complete and continuous. And it can be very confusing if sedimentation rate changes because the rhythm that you're measuring is in thickness, not in time, and you have to transform it from thickness to time using the argument that you've got a complete rhythm. So, it's quite an involved exercise, and it sets very high requirements on completeness and continuity of the record, and also steadiness of deposition.
Oliver Strimpel
If now there is something that's interfered with the sedimentary record in those intervening bands between the radiometric, dated bands, does it always show up? And can we then account for that?
Bruce Levell
It's a real detective game, that ammonite zonations, in this case, can give you a clue as to where things are missing, but there are some beautiful examples, if you compare this lower Jurassic in Lyme Regis with the Lower Jurassic on the other side of southwest England and Somerset, you can see that the Lyme Regis section is extremely condensed with respect to the Somerset section. And you can actually identify particular bedding planes where stuff has to be missing based on missing ammonite zones, and I've been and looked at them and crawled all over them. Oliver, there is nothing, you find a few scattered echinoid spines and a sharp bedding plane on a calcareous limestone. And that's it. And yet the evidence is saying that several 10s to 100 thousand years is missing on that particular surface. It's really frightening. It's really frightening, because if you didn't have the more complete section, you wouldn't know that's coming back to your earlier question.
Oliver Strimpel
You also mentioned our desire to use the sedimentary record to understand the environments in which fossils lived and perished. What sort of bias affects the record there, and how can we overcome their effects?
Bruce Levell
Actually, the paleontologists have studied preservation potential in much more detail than sentimentalists, and for a longer period of time. There's a whole branch of paleontology called taphonomy. Which is the study of what happens to things when they die: the extent to which soft tissues get preserved, and also how you can reconstruct how things came to die. The biases in the record, I think the most obvious one in paleontology, is dissolution. So, the sedimentary record is incomplete, not only through erosion but also just dissolution of material. It's very common in sedimentary sequences to only find particular fossils. An example would be finding oysters and echinoids and nothing else. There's a formation we're working on in Oman at the moment, and that's the situation there. And the reason is, that those are calcitic forms and many of the animals that would have lived at the same time precipitated shells from aragonite and other aragonitic forms have subsequently been dissolved, so you've got a preservation by us towards the calcitic forms -- more robust forms, obviously, hard bodies as opposed to soft bodied organisms. And then as far as the sediments themselves are concerned, you preserve the organisms which live in the sediments which get preserved. So, the organisms which were happy living in the bottom of a fast-flowing tidal channel stand a chance of being preserved. The razor shells which were at the same time living in the tidal flats, which are not being preserved, don't get preserved. So, it comes with the territory, the sort of bulk preservation of the sedimentary record as well.
Oliver Strimpel
OK, let's talk about reconstructing discrete events that occurred in the geological past. First of all, what kind of events are you referring to here?
Bruce Levell
So, I'm thinking of the question: you have an interbedded sequence of sands and muds. For example, maybe it's a shallow marine sequence and you'd like to say something about the intensity of storms. So, I think these are individual graded beds, so they fine up once they're deposited by discrete events of flows which are waning over time and hence depositing. So, how strong were they? Can I say anything about the relative strength of storm intensity here versus there? The type of coast? The same for floods. Are these rivers characterized by very strong flows seasonal, or have you got average flows over long periods of time in a sort of temporal setting? What can we say about those kind of situations? It's a very interesting problem because, there was a classical study done on the coast of Southern California, where they asked that question of modern shallow marine sediments whose age they knew, because they'd repeatedly surveyed the shelf and they knew when the deposits had arrived. And, what they were able to show was that it was highly contingent what got deposited. So, firstly, in order for something to be recognizable in the shallow marine record, it should not be borrowed to destruction. So, if you just deposit after a storm, a graded sand bed and it stays within about 10 centimeter to the surface, then the burrowing fauna at the bottom of the sea just destroy it and it becomes unrecognizable. All the sand grains are moved around. You can't recognize it as a bed anymore. So, in order to be preserved, it has to be insulated from burrowing, which means buried below the zone in which the animals live. So, the most important thing for preservation is that something happens afterwards. It’s nothing about the event itself, but what happens subsequently. You need a rapid other event which buries it. So, it's telling you nothing about the magnitude. And then, they tracked these storm layers back into the flood basins of the rivers from which they've been derived. And they were able to show that the amount of material that was being transported to the shelf in the first place had nothing to do with the intensity of the storm or the flow in the river. It was the event which occurred, the rainfall event which occurred, when the river was already full of sediment. So, previous events had pushed sediment from the hill slopes into the river. The sediments were in the bars, the banks of the river, and the bottom of the river, and it was only when the river was actually full of sediment that the next event flushed it all out to sea. So, it was highly contingent what got preserved. It was a flood event, which happened to occur when the river was full deposited on the shelf, and then a shelf event, which happened to be preserved because something else happened later. So, the conclusion is that the record is highly contingent and actually almost impossible on a bed-by-bed basis to reconstruct. So, that's kind of disappointing in terms of what we can extract from the record, but I find it fascinating in the sense that as scientists, we are taught to see a record, a sequence of events, and interpret it deterministically, and we don't give perhaps enough credit to the chaotic interpretation where a bunch of nonlinear functions are actually combining to produce a signal which we wish to interpret deterministically, because that's how we've been educated, but in fact, there is no meaningful deterministic explanation that we can extract from it.
Oliver Strimpel
If only we were back there, we could have a time-travel machine, and then we'd know actually what happened. But in fact, you're saying the contingency both on what happened prior and what happened after interferes with what happens at any given time. So, it can just be too scrambled up — too much information loss to be able to accurately reconstruct.
Bruce Levell
Yes. People are working on this on different scales. I described it on a small scale there, but on landscape evolution scale or a larger scale, what we want to do is extract the external forcings of a particular event. So, how big was the hurricane? How big was the storm? How big was the flood? Or how long was the Ice Age? How big was the Ice Age? Or, you can scale this up to what extent instead of an input change, or did the sea level fall? These are alternative ways of making a coastline move forward into the ocean, for example. And, people talk of signal shredding. It's a term I quite like. It’s that the interplay of all the processes which are involved in bringing that sediment to the coast, the storage in the floodplain the sweeping out by subsequent floods — they're actually shredding the input signal. You're absolutely right. If you were there, someone could have observed everything that happened. But reconstructing it from the result and inferring a proximate cause — the warning signs are there from the modern studies that that is an extremely difficult exercise.
Oliver Strimpel
And again, I'm concerned about what we said earlier, which is, is there evidence in the record that the shredding has taken place, or might we just simply blithely assume that what we're seeing is a faithful record?
Bruce Levell
Well, this comes to the point of what is the null hypothesis with which you approach the geological record, and I think that's my essential philosophical point. And, I believe that before allocating an external cause, you need to rule out the chaotic interpretation. You need to rule out the autogenic interpretation. By autogenic I mean the processes which are intrinsic to a sedimentary environment. So, meanders migrate sideways and spits prograde into tidal inlets, and barriers move landwards, and all of these things are happening just by virtue of natural variability. There's nothing happening to sea level necessarily, or sediment supply, or tectonics which is controlling those. And only when you've ruled those out, are you kind of allowed to go forward and invoke an external cause. That's the way I was brought up as a sedimentology, we used to say that the sedimentology's prayer was Lord, forgive us our transgressions -- because, we would never invoke a transgression unless we had to, because there was no way out of the particular situation. And it strikes me that there's been a bit of a swing of the pendulum the other way, that people are very eager to move straight to external forcing. And there are a number of deep statistical debates as to what the null hypothesis should be when you claim to see a signal in the sedimentary record, particularly with this astronomical time-scale tuning. And it is not a simple question.
Oliver Strimpel
You also mentioned that we'd like to be able to reconstruct the environment in which the deposition we see preserved today took place. Are you talking about things like whether it was windborne or waterborne or if it was a water deposit, whether it was marine fluvial or from a lake?
Bruce Levell
Yes, that's where you'd start off. And then what we would typically need to do is to go down to the next level of detail. So, if it's within a lake, is it the beach of a lake or is it the bottom of the lake or is it a delta coming into the lake? So, what we're trying to do is reconstruct the three-dimensional shape of the sedimentary bodies that have been preserved, such that we can either store fluids in them like carbon dioxide, extract fluids like water or oil and gas, or build things on top of them, so, we'll find a solid footing for our piles, for instance, for coastal defense as sea levels rise. So, we need to know where physically these different lithesome, these different lithological bodies actually are and what shape they are. So, the trick there is to reconstruct the depositional model, so a river coming into a lake, for example, typically has low wave energy, so you see the channels protruding quite a lot into the lake. We sometimes call it a bird's foot delta, because it looks like the three toes of a bird's foot as it migrates into the lake. And bigger lakes, you would see wave reworking, and you would end up with smooth, cuspate shorelines, for instance, from the waves. And so, those two types of delta would give us different shape sand bodies, and we would like to reconstruct that. And the preservation point here is that the body that we want to reconstruct is not simply the planned view of the body when it was deposited, but what ends up being preserved over time, and that depends on this overlay of erosion due to base-level changing. For instance, in a lake, base levels drop as the lake dries up, and then they rise again, maybe on a seasonal basis. And each of those base-level lowerings will result in channel incision and erosion of previous material and lake-level rising will result in the strand plain and the beach truncating the previous delta material and wiping it out. So you have to reconstruct a 3-dimensional shape which reflects the depositional model and the preservational filter.
Oliver Strimpel
It seems that there are actually a huge number of interacting processes and chance events that can affect what gets preserved. I wonder where you see the field going and whether perhaps you think that the problem of discerning and accounting for preservational bias is amenable to artificial intelligence and machine learning. It probably depends in part on whether we have a large enough data set of records for which we know the biases that can serve as training data for a model.
Bruce Levell
That's a really good point. It strikes me that this problem that we've got of disentangling multiple interrelated processes which have feedbacks between them is one which without AI you're not going to solve. Let's put it that way. And I think there is hope there, and I'm looking forward to a literature emerging on that topic.
Oliver Strimpel
In an earlier podcast, Paul Hoffman laid out convincing evidence for the Snowball Earth hypothesis, which posits that there are two periods during the late Proterozoic during which the Earth was completely frozen over. As I mentioned in my introduction, you studied a sequence of glacial deposits in Southwest Scotland corresponding to the earlier of these two periods, so as to develop a deeper understanding of the preservational bias that might have been operating there. Can you tell us about that?
Bruce Levell
Yeah, this unit is the Port Askaig formation in Southwest Scotland, and it's Sturtian. It's the older of the two Cryogenian glaciations. And it's exceptionally thick for a Sturtian sequence. And the reason why it's exceptionally thick is that the deposits have been preserved in a basin which was starting to rift at that time, and so was very actively subsiding. And what that enables us to see is that we have a series of alternations between glacial and non-glacial depositional environments. So, we can clearly demonstrate that ice was advancing and retreating. We can demonstrate that the ice was grounded. We have lots of glacier tectonic features which indicate the advance of ice over land surface, and we have lots of periglacial features which indicate the exposure of a land surface to the air. And, we have glacial fluvial glacial lacustrine and glacial marine deposits, so we can demonstrate that there was an active hydrological cycle during that time, and that the planet wasn't completely frozen during the deposition of that material. Now, in the classical snowball, which I think is not fully believed in its hardest theoretical form anymore, it posited that the entire earth was frozen at that time. And a bit of an issue there with this particular sequence is that if the Earth was totally frozen and there was nothing happening, basin subsidence wouldn't really know about that and basins would continue to subside. There's no reason why rifting processes stopped because the earth got a little bit cold on the surface. And therefore, if there was no sediment input into a basin, because the entire earth was frozen, you would create large holes in the ground and when sedimentation restarted, you'd have to start in relatively deep water filling up those holes. So, the ice would melt, the holes would be filled, and then the sediment would arrive, perhaps as the ice melted into these holes. And that is not the sequence that we see in Southwest Scotland. So, we feel that we're adding detail to the nature of the Snowball Earth.
Oliver Strimpel
What are you working on at the moment?
Bruce Levell
I'm editing the successor to a standard sedimentology textbook that was published in 1996 by Harold Redding and coworkers, and Harold Redding was my supervisor. “It's called Sedimentary Environments: Processes and Facies”. One of the reasons why I wanted to take that up was this point that I wanted to bring an understanding of preservation to the depositional models in the next update of that book. And then another project together with a guy we both know, Henk Droste, and Mike Searle, we're working on the sediments that were deposited during the abduction of the ophiolite in Oman. So, trying to understand from the sedimentary record what we can deduce about how the continent reacted to the loading by the ophiolite slab when it ended up on top of Arabia. And that's turning out to be quite an interesting problem, because what the ophiolite slab did is it caused the carbonate platform to drown. And when the carbonate platform drowned, there was no sediment production, because it became too deep. So, in essence, we have a record of water depths being created without a sediment infill to record what actually happened in detail during the abduction process. So, it's a starved sedimentary basin, sort of an extreme example where you can't reconstruct the record because sediment supply was so little that it's not been documented.
Oliver Strimpel
Bruce Levell, thank you very much.
Bruce Levell
You're welcome. It was fun.
Oliver Strimpel
For more about Geology Bites, as well as pictures and illustrations that support this podcast, go to geologybites.com.