I explained in an earlier blog post the significance of the sampling effort that was undertaken to understand the pattern of isotopic values, and how this changed over time, of precipitation coming from Superstorm Sandy as it made its landfall and slowly died over the interior of North America.
I ended my sampling effort on Saturday night after collecting a total of nine samples, one every twelve hours since about the time Sandy made landfall on Monday night, the 19th of October. There was only one span of time – on Halloween – when it did not rain sufficiently for me to collect a sample.
These nine fine samples are now on their way to the University of Utah where their isotopic values will be measured. But, see, I’m also an isotope geochemist. And I also have a water analyzer in my lab. And I might be just a tad impatient.
So I analyzed the waters before I sent them off.
Let’s think back on what I said before, about Rayleigh Distillation. So if a cloud rains, the isotopically heavier water (mass 19 or 20) is more likely to fall (because it’s heavier) than the more common, lighter (mass 18) water. So the rain is isotopically heavier than the cloud. After the rain has fallen, the cloud is isotopically lighter than it was before.
So, what happens when that cloud rains again?
When a big storm (like Sandy) moves inland, the rain causes the cloud to get lighter and lighter. And since the cloud water is getting lighter and lighter, so does the rain coming from the cloud, though it is always heavier than the cloud itself. This leaves a tell-tale pattern of heavier isotopes near the coastlines where the storm first came on land, to lighter and lighter isotopes further inland.
So what pattern would you expect if you did all your sampling in one place and a storm simply passed over? What if a storm parked over your house and rained for days and days? What would that look like?
Think about it. I’ll give you a few minutes. I need a glass of water.
Keep thinking. I need to check my e-mail.
Well, it would stand to reason, that unless – somehow – heavy water vapor was getting back into the cloud, the isotopic values would get lighter and lighter over time.
So, one might predict that the rainwater that I collected would get lighter and lighter over time.
The pattern we expected to see was completely borne out for the first three collections, from when Sandy made landfall, to when the center of the storm was supposed to be over the Rochester area, where the samples were being collected.
But then what happened? The values start to increase again. Any ideas?
Well, for one thing, Sandy was supposed to pass over Rochester on Halloween, but it didn’t. The bulk of the storm passed to the south. In fact, it didn’t rain at all on Halloween (which made trick-or-treating possible!). Superstorm Sandy swung south and then west of Rochester before becoming too diffuse to know where the core of the storm was.
Something happened. Something changed.
Well, maybe some heavier isotopes did make it into the vapor mass. Perhaps it was the arctic front that was swooping down from the north as Sandy struck from the east that brought the isotopically heavier rain. It definitely cooled off. It was snowing occasionally during those last two sampling intervals. I suppose it’s also possible that the storm picked up some moisture from the Great Lakes as well.
Again, this is the beauty of the larger project and sampling effort. With only one sample site, we can’t be sure. But once we have all the data from the 100+ sampling sites, we’ll be able to map in detail what was happening. It will be obvious of secondary vapor masses (clouds, storms) joined up with the remnants of Sandy. We’ll be able to tell where and when that occurred.
It’ll be a while before all those samples are gone through and analyzed. I sent my own samples off to Dr. Bowen, so he can re-analyze them using his own instrument and add the data to his huge database. In the meantime, I have this one tiny subsample of the data and a lot of excitement for what will be discovered when the entire data set is complete!
Earlier this week, Hurricane Sandy (an anomalous late-season hurricane) made landfall in the United States near Atlantic City, NJ (also anomalously far North). Because of the timing of Sandy (near Halloween), and it’s coincidence with another strong system moving across North America from the West, the weather event was given the moniker “Frankenstorm”.
This storm was a big deal, and my heart goes out to everyone adversely affected by its aftermath. My own heart broke with each image the popped up on my Twitter-feed that night. Yet there were some heartwarming stories, and certainly some good will come from this unfortunate event.
Much of the discussion of Sandy revolved around how unusual it was and how it might be related to global warming. I even got a call from a local journalist wondering if I would be willing to comment on that. (I said no, because it’s really outside of my realm of expertise, but hopefully might be contacted later regarding ancient episodes of global warming which really are my specialty.) There are plenty of web resources on the topic, which cover that question better than I can. This is one of my favorites.
This is all interesting, but is not why I was kind of excited about Sandy (in the way only a geochemist can be). For me, Sandy provides an opportunity to verify what we think we can learn about ancient weather patterns using chemical tracers in rocks. That is, Sandy is a natural isotopic experiment. I’m not the only person who thought this. Gabriel Bowen of the University of Utah thought of it first. I’ll explain below.
Before you get upset about the term ‘isotope,’ remember that all atoms are isotopes and that not all isotopes are radioactive. Most atoms are ‘stable’ meaning that they don’t undergo radioactive decay. It’s just that the term ‘isotope’ makes people think of nuclear reactors and meltdowns (and somehow Homer Simpson).
So then, what do I mean by an isotopic experiment? I’ll save the details of how isotopes work for a later blog post, and just start with a simpler story of just water. Different isotopes have different masses, or weights. Most water molecules have a weight of 16 atomic mass units. Let’s just say most water has a mass of 18. Some water molecules have a mass of 19, where one of the hydrogen atoms is ‘heavy’ (but stable) and some molecules have a mass of 20, where the oxygen atom is ‘heavy’ (but also stable).
When the mass of the molecule is heavier than most (19 or 20 versus 18) the molecule is, well, heavy! That means that if water evaporates, the lighter (mass 18) molecules evaporate first, because they’re lighter, leaving the heavier water (mass 19 and 20) behind in the puddle. This seems very common-sense, and it is. Vapor that evaporates from puddle is lighter than the water that remains in the puddle and, in fact, the remaining water gets heavier. This process is called fractionation.
Now, if we have a bunch of water vapor, like a cloud for example, and the vapor condenses, the heavier water condenses first and falls as rain (because it’s heavier). The rain is heavier than the vapor in the cloud and the cloud’s water gets lighter and lighter as it rains more. Again, this is fractionation.
When we’re talking about isotopes, we use this crazy delta notation. If we want to say something about the oxygen isotopes in water we use δ18O. For hydrogen, we use δD or δ2H. The number we report is really a ratio, but we tack on the permil symbol (‰) to make the numbers easy to talk about (again, this is something to talk about later). What’s important is that if the delta value is more positive, that means that the water is heavier. If the delta value is more negative, the water is lighter. Everything is measured relative to ocean water which has been assigned a delta value of zero for both hydrogen and oxygen. δ18O = 0‰ and δD = 0‰ for ocean water.
A hurricane, like Sandy, gets all its water from the evaporation of the ocean – so the clouds forming over the ocean will have delta values more negative than zero. As long as the storm is over the ocean rain from the hurricane and falls back on the ocean and new water evaporates keeping the isotopic value of the clouds stable. But once the storm moves over land, the addition of new water vapor from the ocean stops, but lots of water is lost as rain.
The result is that as a storm moves across the landscape, the isotopic value of the cloud gets lighter and lighter over time. The precipitation coming from the cloud also gets lighter and lighter over time, though it’s always heavier than the cloud it came from.
This is called Rayleigh Distillation, and is one of the basic concepts in isotope geochemistry. It seems pretty straight forward and reasonable, and has been used as the basis of isotopic interpretation for many years. But it’s been difficult to test… Until now. With electronic messaging and, more importantly, social media, it is now possible to recruit a fleet of people of a broad geographic area with only a few hours notice to collect rain samples that can then be measured for their isotopic values. We can finally ground-truth this important hypothesis!
This was tried for the first time with a storm called “Snowzilla” (now less creatively called the ‘Groundhog Day Storm’) that happened in 2011. Snow fractionates from clouds just like rain does, so would be expected to show a similar isotopic pattern as rain water. When this huge storm that hit much of the eastern United States, and Gabriel Bowen, then at Purdue University, put out a call for people to collect snow samples and send them to him. The results are detailed here.
Looking at the figure, we see that the isotopic values shift from more positive in the southeast to more negative in the northwest. From this, it’s easy to see that the vapor moved in from the Gulf of Mexico and Atlantic Ocean.
What might we expect to see from Sandy? Well, this time when the call went out, Dr. Bowen asked participants to collect samples over specified time intervals and to record those times, meaning that it will be possible to make an isotope movie and perhaps watch Sandy move across the continent.
So… Why does this matter? Oxygen isotopes from rain can be preserved in rocks. As rain water is exposed to carbon dioxide and percolates through the soil, it forms carbonate (CO32-)which is then bound into carbonate minerals like calcite. This calcite can form little nodules in the soil or a calcrete layer. The oxygen in the carbonate records the oxygen in the water (with a little more fractionation). Later – as in millions of years later – geoscientists like me can analyze the oxygen from the carbonate and get back to the original distribution of oxygen isotopes in the rain water. From there, we can then figure out ancient air-flow patterns around the world.
With this knowledge, we can start answering other questions. How does the uplift of high mountains (like the Himalayas) affect global air flow? What happens to air circulation when climate changes rapidly, whether it be warming or cooling? We can address these questions and more, which might help us understand what the future might bring if projections of warming bear out.
In the meantime, I’m a participant in the project myself and am still collecting waters. Sandy’s not quite dead, though her destructiveness is well past. We’ll see what the data tell when all is said and done!
Here they are: the sample set from my house. I’m done sampling, so the analyses can begin!
What is a mass spectrometer? I was just asked this question. It gave me pause.
You know, most of us forget that what is completely familiar to us in our daily lives might be utterly foreign to 99% of the world. For example, I’m a vertebrate paleontologist and know many, many paleontologists. Sometimes, I think the whole world is teeming with paleontologists. But when I think about it, there’s maybe 5-10 thousand people in the world that can call themselves a vertebrate paleontologist. Still a big number, but when compared to the world’s population of ~7 billion, or the population of New York City at ~8 million, it’s quite possible that I may be the only paleontologist that many of my non-paleontology friends know and may ever meet.
I suspect that there are more mass spectrometer technicians in the world than paleontologists, just because there are so many different kinds of mass spectrometers and zillions of applications for mass spectrometry. Nevertheless, most people’s exposure to mass spectrometry comes from watching episodes of CSI, where (naturally) the television show gets it mostly wrong. (Seriously, you don’t just turn on a mass spectrometer in the morning and expect to get results in a few hours. I switched ours on yesterday and I’m extremely hopeful that I can start running analyses tomorrow!)
So, then, what’s a mass spectrometer? Breaking down the name itself is a good start.
Mass: This is the science kind of mass, not a religious ceremony. Mass is generally equated with ‘weight’ or ‘size.’ ‘Mass’ in science-ese is actually more specific than that, but this works. We’re basically considering something in terms of its size or weight.
Spectrometer: Well, the ‘spectro-’ part is the same as spectrum – a range. Just like a spectrum of colors: red, orange, yellow, green, blue, violet. The ‘meter’ part just says that a measurement is being made. We’re measuring a range of something.
Since it’s a MASS spectrometer, we’re measuring a range of sizes or weights.
OK, but measuring weights of what?
Now here’s the fun part, and why I say there are so many kinds of mass spectrometers. We’re usually looking at the weights of components of some material. It may be an unknown material, and we want to know what it is. Or it may be a known material, but we want to look for impurities or (potentially) for its origin.
Some mass spectrometers are set up to look for heavy elements like strontium or uranium and measure their abundance. Others look at organic compounds like fats or waxes to determine, for example, how much unsaturated fat versus saturated fat there is. The one I work with is highly sensitive and can only be used with ‘light’ elements like carbon, oxygen, and nitrogen (and occasionally hydrogen – but I hate hydrogen… we won’t go there!)
All mass spectrometers have the same general components: A means to get the sample into the instrument (an inlet system or peripheral device), a means to separate the masses, and a means to measure the different masses.
A common instrument you might see on a TV show like CSI is a gas-chromatograph mass spectrometer (or GC-MS, seriously, that’s a mouthful!). The inlet system is a combustion chamber (a furnace) where the samples are burned, causing the original molecules to break down into smaller molecules that are now gaseous (rather than a solid). These molecules are separated, by mass, using a chromatographic column, which is essentially just a really, really long narrow tube. The smaller molecules flow faster down this tube than the bigger ones. At the end of the tube is a collector of sorts, which basically counts how many molecules of each size pass through the tube and we get a spectrum of the different sizes of molecular fragments that came from our original sample. The pattern of molecule sizes and amounts is characteristic of a particular material.
Another common mass spectrometer is a quadrupole mass spectrometer. This is used for the heavy elements, and makes measurements atom by atom (not whole or fragmentary molecules). We had one running here at the University of Rochester for a while. The inlet system on it was experimental, but fun. A laser was shot at the sample, forming a fine dust which was then carried into the inlet system in Argon gas. There it went into a plasma torch and was burned up and the gas went into the mass spectrometer. This system has the fancy name of Laser Ablation Inductively Coupled Plasma Mass Spectrometry, or LA ICP-MS. Changing voltages on the four metal rods for which the quadrupole instrument gets its name is how the different masses are selected. A collector is at the end of these rods, which measures how many of our specified atoms got through.
The instrument that I manage is called an isotope ratio mass spectrometer (IRMS). There are several different peripheral devices attached to ours, one of which has a furnace like on the GC-MS, and another that has a series of vials with a moving needle. The peripheral devices are where the solid samples are converted to gas. In the first, samples are burned up and converted only to carbon dioxide and nitrogen gasses. In the other, the solid samples placed in the vials are reacted with phosphoric acid to make carbon dioxide gas, which is what we measure. (And I inject that acid drop-by-drop, vial-by-vial. So when you see me say that I’m dropping acid, that’s what I’m doing!). These gasses go into the mass spectrometer and are ionized by an electron beam (3000 Volts!!) after which they fly away from the electron beam toward the collectors. The different masses are separated by a strong magnet and a voltage is measured by the collector cups.
What’s different about what I do is that I’m only looking at one molecule at a time, usually carbon dioxide. But I’m looking at isotopes. Not radioactive isotopes, but stable ones. Isotopes are atoms of the same element, but with different masses (or weights). Every atom is an isotope. Some are just unstable.
Carbon dioxide has carbon and oxygen. Carbon has two stable isotopes: Carbon-12 and Carbon-13 (and one unstable, radioactive isotope, Carbon-14). Oxygen has two important isotopes: Oxygen-16 and Oxygen-18.
Some math: carbon dioxide = one carbon plus two oxygens. Most carbon dioxide is composed only of carbon-12 and oxygen-16. Take those numbers and add them up: 12 (the carbon) plus 16 (one oxygen) plus 16 (the other oxygen) equals 44 – the total mass of ‘light’ carbon dioxide.
Let’s say the carbon is the ‘heavy’ isotope instead (Carbon-13). Math again: 13 (the ‘heavy’ carbon) plus 16 (one oxygen) plus 16 (the other oxygen) equals 45 – the mass of carbon dioxide with heavy carbon.
What if one of the oxygens are heavy? 12 (the carbon) plus 16 (one oxygen) plus 18 (the ‘heavy’ oxygen) equals 46 – the mass of carbon dioxide with one heavy oxygen.
Obviously, there are other combinations possible, but these are rare and we don’t worry about them. What’s important is that carbon dioxide comes in three masses: 44, 45, and 46. An IRMS can separate these out and we can measure them.
Subtle differences between the relative amounts of heavy and light isotopes of oxygen and carbon (and nitrogen and hydrogen) can tell us a lot about the origins and history of the sample that we’re analyzing. For some examples from my own research, look in my blog under “stable isotopes”
As I was driving home from work yesterday, I was pondering what the next great bit of science would be that I should publish. I started thinking about this project that has been back-burnered for a while.
Projects in the sciences get back-burnered for many reasons. This particular one has been set aside as I wait for results from other colleagues from other institutions. This happens, and is a common occurrence in the sciences. But as I was driving, I realized that part of the project is complete and can be its own paper in the absence of the contributions from the others for the greater project.
Co-authors, get ready. There’s a manuscript coming together by ME!
So, what’s it about? And why is the title of this post “Rodents of Unusual Size”? ROUSes don’t exist anyway, so what am I worried about?
Well, there are some big rodents out there. The largest modern rodent is the Capypara (or Carpincho), which roam around in South America. These rodents average about 50 kg (110 pounds), so they are fairly large. But in the ancient past, South America hosted even larger species of rodents, including Arazamys, Isostylomys, and Josephoartigasia. The latter, is thought to have potentially weighed 1000kg (2200 pounds)! Now that’s a rodent of unusual size!!
A common research question that I answer with my type of research is “what did the animal eat?” I can get at this using geochemical analysis of tooth enamel. The larger project that my colleagues and I are working on seeks to answer the question, “What did these giant fossil rodents eat?”
The obvious answer, of course is, “Anything it wants!” But we want to be a bit more specific. So how do we do this? By studying the isotope geochemistry of tooth enamel.
Diet recorded in tooth enamel
We joke in isotope geochemistry that “You are what you eat, plus a few permil.” When I’m analyzing samples, I’m comparing the amount of Carbon-13 (’heavy’ carbon, but not the radioactive stuff, Carbon-14) relative to the amount of Carbon-12 (the common carbon in the world). Slightly less than 99% of all carbon atoms in the universe are Carbon-12. Around 1% of all carbon atoms are Carbon-13. (And whatever is left is the radioactive Carbon-14). A mass spectrometer can measure the relative amounts of Carbon-12 to Carbon-13 and gives us a number, called a ‘delta value’ in units of ‘permil’ (‰).
We write this like: δ13C = -14‰ (said “delta 13-C equals minus fourteen permil”)
Depending upon what you’ve just measured the isotopes from, this delta value can be interpreted in a number of ways. For diet and tooth enamel, it goes like this:
Plants, in general, use one of two types of photosynthesis. These two types are called C3 and C4. C3 plants are typically trees and bushes (or occasionally grasses) that live in cooler moist environments. C4 plants are typically plants especially grasses that live in arid environments. (This is an over-generalization, of course, but is usually our first assumption.)
Luckily for us, C3 and C4 plants have different δ13C values. C3 plants are usually about -27‰; C4 plants usually around -13‰.
Now, let’s say an animal comes along and eats these plants. You are what you eat, they say. Plus a few permil… In the case of mammal tooth enamel and plants, it’s plus 14‰. So a bison grazing on C4 grasses has a tooth enamel δ13C of about 1‰. A horse that prefers to eat the bushes with have a tooth enamel δ13C of about -13‰.
The difference in δ13C in tooth enamel reflects a difference in diet. In general, we assume that animals that show a C4 diet (tooth enamel δ13C around 1‰) probably were grazing (grass-eaters) and those that show a C3 diet (tooth enamel δ13C around -13‰) were probably browsing (leaf-eaters). Of course, there are animals that do some browsing and some grazing (horses in particular). We can tease out the relative amounts of grazing and browsing in a single animal too.
So the plan is to look at the tooth enamel of the giant rodents Arazamys, Isostylomys, and Josephoartigasia and figure out if they were browsing or grazing. We might assume that they were grazing, since some of the largest land mammals are also grazers (like elephants), but they might also be browsing, just to eat enough food to fuel such a giant body!
It’s always a good idea to ground-truth your assumptions whenever you have the opportunity. There are lots of assumptions that go into inferring that an animal is either a browser or a grazer when there is only isotopic data to look at. We decided it would be worthwhile to examine the isotopes in modern giant rodents to see if our predictions and assumptions are borne out. Since capybaras are the largest modern rodents, we decided to study them.
Capybaras are known to be grazers. We can sit and watch them graze on grass in an environment where there is lots of C4 grass to be eaten. We also know that there are some C3 grasses in the places where capybaras live, but we might assume that since the majority of grasses are C4, then the majority of the capybara’s diet is C4 as well. Thus we predict that the tooth enamel δ13C from a capybara would be around 1‰.
Well, guess what?
Capybaras selectively eat the C3 grasses. Their tooth enamel only reflects a C3 diet! But we didn’t know this until we ran the isotopes! We ran a couple hundred samples, so we know it’s not an error. This was completely unexpected. Seems like we have a problem, right?
Well, really, it’s not the end of the world. It is what it is. This is how science works. We know that capybaras are grazers, but if all we had to go on was tooth enamel, we’d get it wrong.
But we have other things. We have the shape of the teeth themselves. Long and rootless teeth are common in animals that eat an abrasive diet – and is a common characteristic among grazing animals. Have you ever looked at a horse tooth? Most rodents, including the capybara, have these long and rootless teeth.
We can also look at microwear on the surface of the tooth. An abrasive diet (actually, any diet) will scratch and wear the tooth surface, leaving tell-tale marks that we can observe using microscopes. Specific types of marks are associated with different diets: grazing, browsing, fruit-eating, etc. This is the realm of my colleagues. They are looking at the microwear on the teeth of the giant fossil rodents. Hopefully, they’ll get on that soon. I ought to start bugging them.
What does an ROUS eat?
The isotopic analyses from the fossil giant rodents are done. But in the light of what we learned from the capybaras, the interpretation is sketchy. I can’t say more than that right now. Until we have the microwear data, all we can say is “Huh.”
In the meantime, though, the conclusions of the capybara study are important and need to be published, since they kind of shake up some of our basic assumptions for interpreting diets from carbon in tooth enamel. Now all I gotta do is decide which journal. Hmm.
Modern sloths are curious beasts. Generally fairly small, tree-dwelling critters, they’re notorious for their slowness. But they come from a grand tradition of great size. Until the big extinction of large mammals that occurred about 10,000 years ago, there roamed across the land giant ground sloths that would have made most people run in terror.
These giant sloths coexisted with great beasts like mammoths and woolly rhinos and saber-toothed tigers. They didn’t live in the trees; they were far too big. Instead, they moved about on the ground, using their huge claws to rake leaves from trees to eat.
All this is romantic, but seriously, if giant sloths were as slow as their modern cohorts, wouldn’t they have just been gobbled up by the saber-tooth tigers and the dire wolves?
Well, that’s a good question. How can it be answered?
Modern sloths are slow because they have low metabolic rates. Their diets consist of foods of poor nutritive value, so they balance this by sticking high in the trees and taking their time to get around. The low metabolic rate is reflected by having a low body temperature. Most mammals (like us, or horses and cattle) keep their bodies at 37-39°C. Modern sloths (and other low-metabolic-rate mammals) keep theirs at around 32°C.
So all we need to do is measure the body temperature of a giant sloth! Oh, wait. They’re extinct. Dang.
Geochemistry to the rescue!
Almost all of my research revolves around the geochemical analysis of fossilized teeth in mammals, to make inferences about their biology, and the environments in which they lived. To do this, I measure the relative amounts of stable isotopes (not the radioactive ones!) of carbon and oxygen from tooth enamel. The methods I use are (relatively) straightforward, and have been used actively for decades. The relative amounts of the different isotopes of oxygen and carbon can be related to temperature – and here’s our foot in the door to get at body temperature.
It can be complicated though, especially for oxygen, and until recently we couldn’t easily distinguish temperature changes from things like changes in the amount of precipitation. We also could only look at changes in environmental temperature, rather than body temperature. (Sigh.)
That changed a few years back with the development of a new method of temperature determination called “clumped isotope” paleothermometry or just delta-47 (Δ47). As it happens, the heavy isotopes of carbon and oxygen can exist together (clump) in a single molecule of carbon dioxide, CO2 (which is what we measure with the mass spectrometer). This carbon dioxide comes from carbonate (CO3) which comes from the tooth enamel. How often the heavy carbon and heavy oxygen clump in a molecule is directly related to the temperature at which the molecule formed. In the case of mammals, this is the temperature of the mammal’s body.
So all we have to do is count how many carbon dioxide molecules have both the heavy carbon and the heavy oxygen (= clumped isotopes) and we can measure body temperature!
It sounds simple, it’s really not, but only because there aren’t that many molecules with the clumps, so we need a lot of material and tons of analytical time to get it done. This makes it expensive and it’s hard to get materials because you basically have to destroy most of a tooth. Museums don’t like to lend you specimens that you’re going to destroy. I don’t blame them, really.
We’ve been fortunate, however. One museum has recognized the importance of this study: We really do need to know the metabolic rates of giant sloths if we want to understand their biology and behavior. We were lent teeth from two species of giant sloth, as well as teeth from a horse and a bison from the same cave locality that the sloths came from. We know body temperature in horses and bison, so we can use those results for comparison.
We’re also lucky that the clumped isotope method is so new, that the few labs that are capable of running these analyses are eager to try different things. Right now, we’re not having to pay for the analyses, though we do plan to see if we can get funding to pay for more analyses later.
Cool! Let’s do it!
But wait. There’s another problem. You see, sloths don’t have tooth enamel.
We use enamel from fossil teeth because it’s really hard and resistant to alteration during the process of fossilization. If the material we want to measure the isotopes from has been altered, we may be measuring something besides the body temperature signal – and that could be anything!
Sloth teeth are made entirely of dentine (which we have in our teeth, too, underneath the enamel). Sloths have two layers of dentine, a harder outer layer equivalent to enamel and a softer inner layer like our dentine. We’ve decided to measure the clumped isotopes from both the inner and outer dentine layers (assuming that the outer one is less likely to be altered, because it is much harder). We’re also measuring the clumped isotopes from the enamel and dentine of the horse and bison. This is how we’re going to determine if there is any alteration of the dentine in the sloth. If the sloth outer dentine gives the same temperature as the dentine in horse and bison, we have to be suspicious that it represents some alteration value and not really body temperature (and then all this work is for naught!).
Where we are.
Well, the preliminary data are in. They weren’t what I expected, but I’m not a sloth expert, so I’ll wait for my colleagues to chime in.
This is the first installment of my attempt to convert a scientific paper (my own) into plain language that is accessible to everyone. Feel free to ask questions in the comments. I’ll respond there, or with additional blog posts.
Climate change at the Paleocene-Eocene boundary: New insights from mollusks and organic carbon in the Hanna Basin of Wyoming.
There is a lot of interest in climate change these days, especially global warming. Especially if that global warming can be blamed on increasing amounts of carbon dioxide in the atmosphere. The problem is that it’s hard to know if the trend toward warmer temperatures (at least the global average) is due to natural cycles of the Earth or due to increases in atmospheric carbon dioxide because of the burning of fossil fuels by us, or if there is even a relationship between increasing carbon dioxide and warming (maybe increases in both are coincidence, but not not due to some causal relationship).
This paper doesn’t make any arguments to support or refute any ideas about modern global warming. However, it is relevant because it explores a past episode of rapid global warming. This ancient event took place about 55 million years ago. Global average temperatures might have increased by as much as 10 degrees, and did so relatively rapidly (over about 10,000 years). It is suspected that this rapid warming was due to the release of massive amounts of carbon dioxide into the atmosphere.
So the warming of 55 million years ago seems similar to modern warming in being rapid (though not as rapid as in the modern scenario) and being potentially blamed on increased carbon dioxide in the atmosphere. In this paper, we assume that the warming at 55 million years ago did happen, lasted about 150 thousand years, then things cooled back down to more-or-less where they had been before. For the sake of this paper, it doesn’t matter what caused the warming, only that it was.
This warming event began at the boundary between two epochs on the geologic time scale: the Paleocene and the Eocene. We call this event the Paleocene-Eocene Thermal Maximum, or the PETM.
The Paleocene-Eocene boundary is actually defined based upon the onset of warming, as identified by a big change in the relative amounts of two isotopes of carbon (13-C and 12-C) in the atmosphere, and consequently in all organic material that was deposited at that time. How we measure these amounts and what the actual numbers mean are the topic of another paper or blog post. What’s important is that these relative amounts, or isotopic ratios, are presented in what we call the ‘delta notation’ (like δ13C, δ15N, and δ18O) in units of permil (‰). When delta values are more negative, there’s relatively more 12-C in a sample; when delta values are more positive; there’s more 13-C in a sample. The PETM, then, is recognized by a negative carbon isotope excursion (CIE), where the delta values suddenly drop by three to five permil. The PETM ends when carbon delta values go back to what they had been before the CIE started.
Much of what’s known about the climate change at the PETM, and the Earth’s subsequent recovery, is known from cores of rock and sediment collected from the ocean floors. Naturally, we’re interested in what would happen to us – those of us stuck on land. In the Hanna Basin, in south-central Wyoming, there is a sequence of rocks that began to be deposited before the PETM started, and continued to be deposited during the PETM and after the PETM. These rocks were deposited on land and are sediments from lakes and floodplains. In these lakes and small rivers were living lots of organisms, in particular, mussels. There was also a lot of organic material being deposited – so much so that now it is represented by many thick coal seams that are actively mined.
This sets up a scenario where we can use the organic carbon (from the coals) to identify the CIE, and therefore the Paleocene-Eocene boundary and the PETM in a terrestrial rock sequence. Then, we can look at the fossil mussels, and other things, to examine the environmental changes that happened during and after the PETM. The main questions, and ones that are relevant to modern concerns about climate change, are:
1) after the warming ended, did the environment go back to its original state or was it forever changed?
2) what effect did climate change have on the organisms that lived through it?
FINDING THE PETM
First, let’s look at the rocks. The Hanna Formation, the rock unit I’m studying, is about three kilometers thick (or about two miles). The part we care about is in the top half. The bulk of the Hanna Formation is composed of sediments deposited on floodplains, with little shallow streams that wound around (called fluvial). There are two parts of the Hanna Formation that have lake beds in them (called lacustrine), cleverly called the upper and lower lacustrine units (ULU and LLU). The focus of this study is on the upper and lower lacustrine units and some fluvial rocks in between them. I had reason to suspect, when I started this study that the Paleocene-Eocene boundary lies between the lacustrine units. This is borne out in this paper.
It turns out that is wasn’t very easy to identify the CIE (and therefore the Paleocene-Eocene boundary) in the Hanna Formation. The delta values from the coals and other organic materials jump around a lot, probably because the organic carbon I was looking at comes from lots of different types of plants, all of which are slightly different isotopically. One conclusion of this study is that we need to do more ‘compound-specific’ work. That is to say, if we can isolate specific organic molecules and analyze them separately from everything else, that should make the carbon isotope values less variable. Unfortunately, the type of instrument and laboratory that’s needed to do that isn’t present here at the University of Rochester. I’m working on it.
Nevertheless, in general where the values are more negative than -26‰, you’re in the CIE. To help make it more clear, I used a three-point running average of the raw carbon isotope data. This tends to smooth out the line, while keeping the big jumps visible. The first major jump into more negative values occurs at about 2500 meters, which coincides with estimations made using mammal fossils and fossil mollusks by others who have worked in the Hanna Formation before.
When I compared this overall pattern with other published patterns of carbon isotope variability (some from ocean cores and some from terrestrial sections), things matched up pretty nicely. Using pattern matching, I placed the top of the CIE (and the end of the PETM) at about 2650 meters, which is in the lower part of the upper lacustrine unit. This means that the 150 thousand years of the PETM are represented by about 150 meters of rock in the Hanna Formation, or that a meter of rock was laid down every 100 thousand years. This is actually reasonable – no one in the geological sciences is bothered by this rate of deposition.
Now that the CIE is identified, I could begin to address the environmental changes that might have occured during that period of warming. I approach this in two ways:
1) Looking at isotopes of nitrogen – which gives us information about the organisms from which the organic matter is coming (e.g. we can distinguish between a stagnant pond or a lively lake).
2) Looking at isotopes of carbon and oxygen in the mussels that have been collected – which can give us information about annual changes in the environment that the mussels lived in.
Most of the organic molecules that go into coal also have nitrogen in them, though not as much nitrogen as carbon. Usually, when looking at fossil organic carbon, the amount of carbon is so low that there essentially is no measurable nitrogen in the samples. In the case of the Hanna Formation, though, we have coal, which is basically ALL organic carbon-bearing molecules. That means that there’s some hope of finding measurable nitrogen, and that’s what I did.
So really, this part of the study was basically done for giggles – just to see if I could do it. And once I had data, well, I had to interpret it.
There are two ways to think about nitrogen. One is to simply compare how much nitrogen there is relative to carbon (C/N ratios). A second is to look at the ratios of two isotopes of nitrogen, 14-N and 15-N. C/N ratios give us information about the origin of the organic molecules (from algae or land plants, for example) and the isotopic ratios tell us about status of lake, whether it be full of actively photosynthesizing plants or if it is stagnant.
Using the combination of C/N ratios and nitrogen isotopes, it seems that for the most part the organic carbon in the lakes of the Hanna Formation is dominated by land plants. So these are leaves and litter that were washed into the lakes. One interesting isotopic data point sits at the bottom of the CIE. From this point, it seems that there was might have been drying of the lake at the beginning of the PETM. That would make sense, assuming that warming could cause greater evaporation.
The work with the mussels is actually been the topic of two undergraduate senior theses that I’ve advised. They’ve been doing some great work to look at the annual changes in isotopes by collecting multiple samples from single shells, following growth lines, to put together a picture of environmental changes that happened during the individual animals’ lives. I don’t say much about that work in this paper. That’ll be published later. What I do talk about is trends. I’ve taken the averages from individual shells and used those to look at how the isotopes of carbon and oxygen from the shells change over time. I also talk about how carbon and oxygen isotopes change relative to each other within a single shell.
So, how are carbon and oxygen in the shells of mussels, you ask? Mollusk shells are made of calcium carbonate (CaCO3) which contains one carbon and three oxygen atoms. We collect powdered bits of the shells by using a dental drill and take this powder and put it into the mass spectrometer. The calcium carbonate is converted to carbon dioxide (which is easily measured by the mass spectrometer) by reacting the powders with acid. You put acid on the calcium carbonate, it fizzes, making carbon dioxide, which is drawn into the mass spectrometer and – wango! – we have carbon and oxygen data.
Carbon in mussel shells is thought to be derived mostly from carbon dioxide that has been dissolved in the water, and so should track the isotopic value of atmospheric carbon dioxide. Atmospheric carbon dioxide, as mentioned earlier, gets more negative during the CIE then returns to the pre-CIE values. The average values from the shells seem to follow this trend, so there’s no surprises.
But now we’re talking about yet another isotope: Oxygen. The isotopes that we measure are 16-O and 18-O. Isotopes of oxygen are a big topic of discussion when dealing with climate change. That’s because oxygen is an important component of water, and water is an important component of climates. For example, climates can be described as arid or humid. There can be rainy seasons or monsoons. Precipitation can take the form of rain or snow. All of these processes affect the isotopes of oxygen in water. Temperature also affect oxygen isotopes. Unfortunately, isotopes of oxygen in water are a very complex system, and would best be discussed in a separate blog post. What is important is that there isn’t any obvious trend in the average values of oxygen from the shells over time, either.
Since oxygen is affected by climate, one would expect that there should be some change if there’s been a significant climate change. However, because oxygen is so complicated, the changes in oxygen isotopes by changing one part of climate (the time of year when it rains, for example), might be offset by other changes (like a change in average temperature). The fact that there isn’t any obvious trend or change in oxygen isotopes over time doesn’t mean that there wasn’t any change in climate.
And we do see a difference when we compare the variations in carbon and oxygen within a single shell. Carbon and oxygen in shells from the lower lacustrine unit (before the PETM) tend to change in opposite directions (or are negatively correlated). When oxygen isotopic values get more positive, carbon isotopic values get more negative. Shells in the upper lacustrine unit show the opposite pattern. Carbon and oxygen values change in the same direction (are positively correlated), so when oxygen gets more positive, so does carbon. From this, it’s possible to infer that before the PETM there was a lot of vegetation and photosynthesis going on around the lakes, whereas during the PETM, photosynthesis might have slowed down during the warmer months and bacteria might have dominated life in the water. This seems reasonable since in the upper lacustrine unit there are also huge fossilized bacterial mats called stromatolites.
So that’s what this paper is about. We see some evidence of environmental change due to warming at the Paleocene-Eocene boundary. Particularly, we have the one really positive nitrogen isotopic value near the base of the CIE and we see a change in the relationships between carbon and oxygen isotopes in individual mussel shells during the PETM as compared to pre-PETM.
One thing that hopefully is obvious however: there is more work to be done.
The work with nitrogen isotopes started with a shot in the dark. More samples should be analyzed. There’s definitely more work to be done there.
My students’ work on the mussel shells will greatly contribute to this as well. Since I wrote this paper, there have been more shells collected and more samples analyzed. That work needs to be wrapped up and published soon.
We really need to do that ‘compound-specific’ work I mentioned earlier to help clarify the CIE. It’ll also help clear up what sorts of plants were around at that time, so we can better interpret the nitrogen data.
Also, though not discussed at all in the paper, is the fact that there are paleobotanists out there looking at fossil leaves and pollen. There’s a story to be told there, and someone’s getting a Ph.D. For their efforts. I can’t wait until that work is done!
Have you ever looked at an elephant molar? I mean, really looked?
It’s a pretty funky thing to observe. They’re usually comprised of a series of plates – nothing like the teeth we’re more familiar with (i.e. our own) – with roots hanging down, seemingly one root per plate. Well, that’s kind of bizarre. To make matters even stranger, elephant teeth aren’t all in use at the same time. That is to say, usually it’s one or two teeth in use at one time in the mouth in each of the four quadrants (upper right, upper left, lower right, lower left) , while over the span of the elephant’s life there are a total of six or so adult teeth in each quadrant. They come in one at a time, conveyor-belt style, from back to front, falling out the front when they’re too worn to be of any use any more. Aged elephants can die from starvation, when the last of their molars (the M3) falls out and they have no teeth left.
We more-or-less know how the more ‘normal’ teeth of mammals form, like our own teeth. They start to mineralize at the crown (the chewing end), and lengthen incrementally toward the roots. The roots themselves are the last part to form (and in animals with ever-growing teeth, the root never forms, the tooth just keeps growing). People like me take advantage of this pattern of tooth growth because it records the body chemistry of the mammal over the period of time that the tooth mineralized. Sometimes this represents several months to years (and that’s the topic of another blog post). But how does an elephant tooth, comprised of all these plates, form?
Well, since that’s what I’m working on RIGHT NOW, I thought maybe I’d put out some information on what we (being me, and some of the folks at the Mammoth Site in Hot Springs, South Dakota) think might be the case. Then maybe I can tell you about how we’re planning to answer that question.
So there I was, at the Mammoth Site during the summer of 2011. I was given the opportunity to work with the molars of a mammoth, whose skull, unfortunately, had taken a bad fall, but had thereby released its teeth. I was presented with an M3 (still mineralizing) and an M2 (in wear).
Mammoth teeth are a lot like modern elephant teeth. They grow in plates and have roots to accompany each plate. If one were to draw a schematic of the tooth in cross section (which conveniently, I have), it would look like this:
So here’s the crazy thing. I started clearing the cementum off of the M2 to expose the dentine (as seen in the picture below), and discovered that it is not one-root-per-plate, but that the roots, once closed, span between two plates (see the schematic above).
Our collective reaction was a resounding “Huh?” So we cleared all the cementum off of one side of the tooth to make sure we weren’t losing our minds. And this is what we saw:
OK, so we’re still not much closer to understanding how mammoth – or elephant – teeth mineralize. But, using this M2, newly cleared of its cementum covering, we have begun the process of more detailed analysis. In the picture above, I’ve pointed out some sample pits for isotopic analysis. Well, remember what I said about the incremental growth of most ordinary mammal teeth. From crown to root, I said. Well, if that is true for this tooth, then we should see annual changes in the isotopic (geochemical) composition of the plate enamel as we move from crown to root. We see this in all other mammal teeth (and, like I said, I’ll explain that better in a later post). A correlary to that, is that we should see the SAME pattern from crown to root on EVERY plate if the mineralization pattern is always crown to root.
But what if it’s different? Since these teeth come in conveyor-belt fashion, and begin to wear at one end first, then maybe they mineralize one plate at at time, starting at the front. In that case, we wouldn’t expect to see any geochemical change from crown to root on any given plate, but might see a change from plate to plate. And of course, there’s the third hypothesis, that the mineralization front is at some angle with respect to the plates – maybe aligned with the grinding surface itself.
How do we do this? Lots and lots of isotopic samples. When that photo (above) was taken, I had only sampled plate 5. Now I’ve sampled plates 5, 6, 7, and 8, and I continue to sample whenever I can (usually while watching episodes of the Tudors on Netflix). Analyses are on-going. The results so far are interesting, but I can’t say much more. Stay tuned…
Added August 30, 2012:
The conclusions of this study will be presented at the Society of Vertebrate Annual Meeting, in Raleigh, South Carolina on October 18, 2102:
Penny Higgins, Olga Potapova, and Larry Agenbroad: MINERALIZATION OF MAMMOTH MOLARS
This recent post reminds me that I ought to be using this blog for good, not just for shameless self-promotion (although self-promotion can be fun). I need to promote science, its importance and utility. And, of course, how fun it is!
So I think I’ll start adding blog entries about my current research, what it is, and why it matters. Anyone interested?
Current projects include:
Body temperature in giant ground sloths. (You can do that?)
Paleobiology and dietary preferences of giant (1000 kg) rodents in South America. (Yes, Rodents of Unusual Size do exist!)
Tooth mineralization patterns and their relationship to diet in notoungulates (extinct endemic mammals from South America).
Continental environmental change associated with rapid global warming at the Paleocene-Eocene boundary (55 million years ago).
Late Cretaceous vertebrates from Axel Heiberg Island. (yeah, in the Arctic)
Less-is-more: Using bulk isotopic analysis from tooth enamel of fossil mammals to predict yearly patterns of temperature and precipitation.
Mid-Paleocene mammals and reptiles, and species turnover due to climate change 60-ish million years ago
If you’re interested in any of these things, let me know, and I’ll write about them. I can also write about day-to-day life as an non-tenure-track isotope geochemist, in a rigorous research-heavy earth-sciences department.