Kim Hannula is a 40-ish geology professor at a public liberal arts college in the Rockies. Her New Year's resolution is to reduce stress by changing her rheology, or maybe by walking to work and looking at the pretty mountains.
Gerta Keller has a new paper published by the Journal of the Geological Society (London). Keller has a long history of arguing that volcanism at the Deccan Traps (a huge flood basalt in India) is a better explanation for the extinction of (non-avian) dinosaurs than the Chicxulub impact was. I haven't read the paper (and given that I've got limited time and I don't teach paleontology, stratigraphy, or historical geology, I probably won't read it myself). However, Suvrat Kher at Reporting on a Revolution has, and describes the paper as providing testable hypotheses, although it doesn't disprove the most compelling aspects of the impact theory (the iridium layer, in particular).
If you're not familiar with the Chicxulub impact story, Ethan Siegel at Starts with a Bang has a good overview, prompted by a story in Universe Today about Keller's paper.
And I would stop there, except that I want to take issue with Ethan's characterization of Keller:
But this has been known for years; why write about it now? This is topical because of a story by my buddy Ian, writer for AstroEngine.com and Universe Today, where he reports on a naysayer who doesn't believe this story. It's true that there are a few unexplained mysteries about this theory, such as how uniform the K-T layer actually is (i.e., how much sedimentation/overturning there was), whether the Chicxulub impact is the one that actually created it, and whether a few hundred-thousand years is a reasonable settling time for the debris. But the evidence here is really overwhelming and the explanation is very, very simple and compelling. That, and a consensus of scientific experts disagree with her conclusions, and support the view presented here.Much like with global climate change, evolution, dark matter, gravity and the big bang, there are unexplained details here. However, these theories -- overall -- are solid, sound, valid, and by far the best scientific, natural explanations available. So don't be fooled by the skeptics who say that because one detail is unexplained, the whole theory is wrong. That's part of how we learn from science, is by properly explaining these details.
Is it fair to lump Keller with the people who claim that evolution is wrong and global climate has nothing to do with carbon dioxide (and even if it does, people aren't to blame)? Yes, she's been making similar arguments for years. (Decades? Yes, I think it's been decades now.) And the Chicxulub impact is a compelling story (especially when told in a talk by Walter Alvarez). Maybe Keller's conclusion is wrong, but (based on Suvrat's reading) she seems to have done some careful work and uncovered data that need explaining.
And that's an important difference from the skeptics about evolution or climate change. She's not bringing up something we already know. She's not saying "A-ha! The eye is too complex to be evolved!" or "Radioactive dating doesn't work because of fractional crystallization" or "If humans cause global warming, how come the Cretaceous was so hot, huh?" When people bring up those (and many other arguments), they're repeating arguments that have already been discussed and examined and have been incorporated into models used by people doing the research. They're arguing against over-simplified intro-class-textbook science at best, and out-of-date science at worst. Anyone who teaches intro classes or writes science journalism needs to know that those arguments are out there, but NSF can be forgiven if it doesn't fund research projects based on those arguments.
Keller is different, I think. She's collecting new data. Maybe she's wrong, and maybe her paper doesn't deserve to generate headlines like "New blow for dinosaur-killing asteroid theory". But lots of scientists are wrong - even scientists who are right most of the time. That doesn't make those scientists the equivalent of the Discovery Institute.
Life Science
Physical Science
Comments
I left a similar comment over at Suvrat's blog ... the key here is this statement (from press release):
"...4 and 9 metres of sediments were deposited at about 2-3 centimetres per thousand years after the impact."
From this they estimate the impact predates the K-T extinction boundary by ~300 kyr. I'm going to need to get the full paper, but did they get the 300 kyr from this estimate of sedimentation rate (i.e., the 2-3 cm/kyr)? I hope there is more to it ... I've got some issues w/ using sed rates in that fashion (especially pre-Quaternary) to determine durations of depositional episodes.
The press release also states:
"The study found that the sediments separating the two events were characteristic of normal sedimentation, with burrows formed by creatures colonising the ocean floor, erosion and transportation of sediments, and no evidence of structural disturbance."
What is "normal" sedimentation? Also, I want to see these trace fossils ... in some cases, burrows can form very quickly (hours to days) ... we know this from studies of the modern sea floor.
I suppose I should hold off on pre-judging *too* much until getting a look at the paper. :) Fun!
Regarding the point of your post, yes Keller is different ... as you point out, she is collecting data and doing science. People may think it is bad data or crap science, but at least she is *doing* some for others to evaluate. The vast majority of the internet psuedo-/anti-science punditry doesn't actually do any work (because they are lazy and/or dishonest), which puts them in an entirely different category.
Posted by: BrianR | April 28, 2009 9:34 AM
Argh - Brian, you're right - Keller's paper is in the May issues, and the contents are listed on the site, but the papers aren't online. I want to see someone with a sed background look at the evidence. (Those press releases that precede real publication are really, really annoying. It's hard to have a real discussion without access to the evidence being discussed.)
And unrelated, while I was running, I decided I should comment with a personal disclosure of my biases.
Reasons I would be pro-impact explanation:
- One of my PhD committee members worked on dating Chicxulub tectites while I was in grad school. Those things are so cool!
- My undergrad structural geology prof was a PhD student of Walter Alvarez's.
- I've seen Alvarez talk (hour-long seminar) about the K-T boundary, and had dinner with him afterwards.
- A friend from undergrad did his PhD with people who also worked on the Chicxulub tsunami deposits.
So I've got all kinds of connections to people on the impact side in my greater academic social network. I haven't worked on it myself, but friends-of-friends have.
I don't have any connections with Keller, on the other hand, and don't have a feel for how she's regarded in the community. But I have worked with people who were on the losing side of scientific debates, but whose observations and ideas were still worth listening to. The community is a lot more complicated than "legitimate scientists" vs "denialists." Over-simplifying it gives too much credit to creationists and their ilk.
Posted by: Kim Hannula | April 28, 2009 10:04 AM
Excellent post! I think that calling Keller a "skeptic" and lumping her with skeptics/denialists in the worst sense of those words is somewhat inflammatory. She's a geologist out doing geology, right or wrong, hopefully advancing our state of knowledge!
It will be interesting to read the paper - looks like it might be online at the JGS site in a few months. Here's the list of what's in the May issue: http://www.geolsoc.org.uk/gsl/op/edit/publications/journals/jgs/page432.html
It looks like they have the Feb. issue online right now as "current." Wonder if Chris has a subscription.
Posted by: Silver Fox | April 28, 2009 10:23 AM
Brian writes: From this they estimate the impact predates the K-T extinction boundary by ~300 kyr. I'm going to need to get the full paper, but did they get the 300 kyr from this estimate of sedimentation rate (i.e., the 2-3 cm/kyr)?
no they did not. the temporal constraints (I have left a comment on my post also) comes from a combination of magnetostratigraphy and foraminifera biostratigraphy.
Keller has found at different sites that the sediment between the impact breccia and the K-T layer lies within reverse polarity C29r (A few cm above the K-T layers the zone is 29n) and within Late Maastrichtian foraminifer biozone CFI which spans the last 300,000 years of the Cretaceous. the base of the CF1 zone is just below the event layer.
as for a real publication you can get a feel for Keller's work and data from this paper.
Posted by: suvrat | April 28, 2009 11:29 AM
Suvrat ... thanks! The magneto-/bio-strat constraints make me feel a lot better. I will look at the link you provide later. I assume that the shift in chrons occurs w/in the proposed 300 kyr interval?
Posted by: BrianR | April 28, 2009 11:47 AM
hmmm ... looking at the Keller et al. (2004), which I'll call K04, that Suvrat links to above ... since neither the chron nor the biozone actually shift between the impact breccia and the K-T boundary (fig. 2), doesn't the interpretation of 300 kyr duration rest almostly exclusively on the interpretation of depositional environment (and its implications)?
Posted by: BrianR | April 28, 2009 11:56 AM
I've read through K04 (the paper Suvrat links to above) quickly over lunch; the paper concludes that the Chicxulub impact predates the K-T boundary by ~300 kyr. If I have time tonight, I'll try and put together a blog post about it.
K04 dwell on the 50 centimeter thick fine-grained interval above the 100 meter thick impact breccia and below the biostrat/chronostrat-constrained K-T boundary. This interval pre-dates the K-T extinction boundary -- this aspect of the story appears to be pretty solid.
The rest of the paper argues that the 50 cm-thick fine-grained interval is temporally unrelated to the underlying 100 m thick impact breccia. They conclude that the 50 cm thick interval represents 'quiet' deposition (i.e., not high-energy backwash related to impact). I have no problem with that interpretation ... the grain size and structures certainly lead one to that conclusions. But, why can't it be low-energy deposition right after the high-energy backwash? At some point, the higher-energy sediment transport/deposition wanes and the finest material will deposit (this could take days to weeks [months?] I'd imagine).
However, K04 brings up the presence of glauconitic horizons, which is interesting and the most compelling part to me (so far) for their interpretation ... K04 draws upon it as an indicator of slow sedimentation rates. But, at the same time, they state that the glauconite is re-sedimented. I need to revisit that part in the paper in more detail.
So, K04 concludes that the Chicxulub impact predates the K-T boundary ... as I said above, that is solid ... but I still don't understand how they get to the 300,000 years part.
Posted by: BrianR | April 28, 2009 1:31 PM
Kim, I appreciate you listing your biases, uncomfortable as that might be. It points to a real phenomenon in science: no matter how much we talk about truth and beauty, there are people you don't want to cross. There are authorities, with human powers and human failings. And usually the authorities are right. And a cool theory has the power to cloud our minds with certainty.
Alvarez has every right to his pride, but Keller is the absolutely necessary whetstone that every prevailing theory needs to keep it sharp--or occasionally kill it. She takes science seriously and asks if the rocks really show what Alvarez's camp says they should.
I don't have a dog in the K-T hunt, and I've always thought that the volcanists had a good case. Amazingly, even the biggest impacts don't seem to be as deadly as we think, and Earth herself has plenty of nasty surprises without invoking the heavens.
Posted by: Andrew | April 28, 2009 6:11 PM
Brian-
You're right. The paper K04 doesn't directly explain the 300,000. But I had left another reference on the comments on my blog which does. That is a description of another section from Brazos Texas, where the base of the CF1 biozone (which spans the last 300,000 years of the Maastrichtian) is seen to be just below the event layer. So they argue that deposition of the event layer took place in the lower part of CF1. They don't mention that in K04 but since the event layer can be treated as an isochron, Keller seems to be extending that reasoning of 300,000 years to other locales, kind of a composite stratigraphy argument.
I agree though that if the chrons had shifted inside the interval the evidence would be stronger!
Posted by: suvrat | April 28, 2009 8:53 PM
Suvrat ... ahh, okay ... that makes a bit more sense. It's a little late for me, so I'll have to dive in a bit more tomorrow if I have the time. Interesting stuff.
Posted by: BrianR | April 28, 2009 10:07 PM
Has anyone taken the time to look at Keller's excellent biostratigraphic documentation; it is very good, indeed. Similarly, Bob Sullivan's interesting dinosaur and non-dinosaur vertebrate extinction analysis, now "ancient" but still accurate, documents the early extinction of many vertebrate taxa, well before any proposed impact event. Dave Archibald, following on Van valen and Sloan demonstrates the complexity of mammal evolution across the boundary. Of course Jim Fassett's very recent entry into dinosaurs in the Tertiary, seen in the Ojo Alamo, follows on Keith Rigby's earlier and similar proposition, not as well done as Fassett, is a matter of contention, but interesting.
The coincidence of withidrawal of the epiric seaway, the likely wave of extinction that follows that withdrawal as habitats were reduced, points to an equally compelling scenario. Deccan traps would seem to have had a powerful and longer term impact on worldwide environments--the production of continent sized masses of volcanics must have an impact (no pun intended). Complex answers for complex events are much more interesting than superficial deus ex mechina modalities, and it is likely Keller is on an interesting track, with similarities to that Chuck Officer traced some years ago.
Posted by: Donald Wolberg | April 29, 2009 10:58 PM
Deccan Traps geochronology review:
http://www.springerlink.com/content/p224366g67433156/
Posted by: Lab Lemming | April 30, 2009 6:13 AM
As a dinosaur researcher myself, I've always been a "little from column A, little from column B" sort of guy when it comes to the KT extinction event. No single agent was the "smoking gun" that did it in for ~60% of the species at the time. In most places around the world we don't have the stratigraphic resolution necessary to properly constrain what happened when, and its potential causality with the extinctions. This wiggle room is what helps keep the debate alive, which I think is exciting and absolutely necessary. Keep the data coming!
As for a related, but I would say far more contentious, topic in dinosaur paleontology see this recent article from Palaeontologica Electronica on evidence for Paleocene dinosaurs: http://palaeo-electronica.org/2009_1/149/index.html
Why the media hasn't picked up on this is beyond me. Anyways, it should provide for some pretty lively debate at future meetings!
Posted by: Chris Noto | April 30, 2009 10:41 AM
Someone correct me if I'm wrong ... but the Keller work isn't a impact vs. non-impact cause, but whether or not the Chicxulub impact in particular is responsible for the end-K extinction.
From the Keller et al. (1997) GSA Bulletin paper:
"The stratigraphic separation of the K/T boundary and siliciclastic deposits and the evidence of long-term deposition between them, suggests the presence of two events: (1) a globally recognized K/T boundary (impact) event marked by Ir anomaly and the mass extinction, and (2) a Caribbean event (impact or volcanic and probably linked to the Chicxulub structure) that predates the K/T boundary and is marked by glass and siliciclastic or breccia deposits." -- in addition, Keller et al. (2003) Earth-Science Reviews is titled 'Multiple Impacts Across the K-T Boundary'.
Some of the reports and things I've read in mainstream press seem to miss this completely.
I'm still not wholly convinced of the 300,000 yr interpretation ... that number is cited in almost every Keller pub from the past 7-8 years ... I realize that is the CF1 biozone for the uppermost Maastrichtian, but how exactly is it pinned to that? In some other papers they assume a sedimentation rate and extrapolate, but it's all still w/in the CF1 zone; or, they assume the strat interval starts at the base of the CF1 zone (which is 300 kyr before K-T).
By the way, I'm not saying they are wrong ... I just tend to get fixated if I don't undersand the rationale :)
Posted by: BrianR | April 30, 2009 2:06 PM
Brian writes-In some other papers they assume a sedimentation rate and extrapolate, but it's all still w/in the CF1 zone; or, they assume the strat interval starts at the base of the CF1 zone (which is 300 kyr before K-T).
They are not just assuming that the strat interval starts at the base of CF1. In several Mexico and Texas sections the base of CF1 lies just below what has been interpreted as the primary spherule layer (original impact fallout deposit). The geochemistry of the glass in this layer has been linked to the geochemistry of the Chicxulub melt rock. so there is a case to be made that this spherule layer records the Chicxulub impact.
This stratigraphy position of the base of CF1 with the lowermost spherule layer is interpreted by Keller to mean that the impact took place in the lower part of CF1 ~ 300,000 years before K-T boundary.
The reference above has links to a detailed explanation by Keller on the sedimentology and stratigraphy of late Maastrichtian and a rebutal by Jan Smit.
Posted by: suvrat | April 30, 2009 10:57 PM
The 300K time for the CF1 can be reference linked back as far as some 1998 Li & Keller ocean drilling papers, but I had to do real work before getting a chance to look into those.
Posted by: Lab Lemming | May 1, 2009 4:55 AM
Suvrat - the link to the online comment/reply is great ... that's a nice way to catch up on the debated points rather quickly. Thanks.
Posted by: BrianR | May 1, 2009 8:54 AM
My hypothesis: Impact killed most dinos, and antipode (Deccan traps) gradually finished off the rest. And hey, we had 2 earthquakes today in Eureka, Nor. Calif. holottashakin~!
Posted by: DDeden | May 4, 2009 12:05 AM
I need a break from medical articles and blogs, so this is going to be fun, and I'm writing from the perspective of a complete amateur!
First, I was getting my graduate degrees (in biological and medical sciences) when I read Alverez's theory in Science in 1980. I also remember there was quite a kerfuffle about his theory at that time, with many paleontologists and geologists questioning the theory. I have no clue if Keller is right or wrong (and she sure seems to be alone on her theory), but, in science, as long as you present a scientific case, it will eventually be merged into the main theories. It takes work.
Second, I think that the K-T extinction should be divorced from the K-T boundary. One is a biological event and one is geological, of course. Given that many of the dinosaur clades were in serious decline well before the extinction event, it's possible that both events contributed to the final demise of the dinosaurs (except for Aves).
I still think there are a lot of problems with both theories that bother me as an amateur. For example, the size of the bolide is large enough to have incinerated the earth. Yet, I'm here, evolved from small mammals that passed through the event. And birds survived. And so did turtles, crocodilians, and seemingly random animals and plants. I'm also concerned about the fossil patterns. I'd think that the K-T boundary layer would be rich with fossil remains of a huge variety of plants and animals. Yet it isn't, which may mean that everything was incinerated...and we're back to the conundrum.
Anyways, I think that Keller has some valid theories, and she is most definitely not a pseudoscientist. She's not a creationist (at least I haven't read anything that would indicate she is). She's not a nutjob proposing something that doesn't even have a basis in science. My bet is that her theories are going to be incorporated into an overarching K-T theory.
Posted by: Michael Simpson | May 12, 2009 10:42 AM
Michael says: "I think that the K-T extinction should be divorced from the K-T boundary. One is a biological event and one is geological, of course."
While seemingly distinct, remember that the boundaries in the geologic time scale are where they are *because* of changes in the fossil record.
Posted by: BrianR | May 13, 2009 9:58 PM
Slight revision: Cosmic impact causes antipodal mass flood basalts, mass extinction. See TED talk (at 14 minutes):
here, the H2S and CO2 are just reactions to the impact and subsequent flood traps.
Posted by: DDeden | May 18, 2009 6:18 PM
Sorry, I'm a little late to this 'party'. The article is finally up and available here. Unfortunately, there is a mistake with the formatting so look for "Gerta et al."
BrianR is correct in that these data say nothing about an impact at the end-K. It just says that Chicxulub is too early. Alvarez et al. (1980) did a decent job demonstrating that there exists a stratigraphic signal that is consistent with a bolide at the end of the Cretaceous (Chicxulub wasn't considered the 'culprit' until Hildebrand et al 1991).
Where Alvarez et al. (and most other extinction studies) go beyond their data is inferring a tentative correlation as a causal mechanism. Correlation doesn't equal causation (not to say it definitely wasn't an impact, we just can't know with our current understanding of stratigraphy). This is a pervasive problem with extinction studies. We lack stratigraphic resolution to say much about extinction events. We can't determine a rate of extinction (Signor and Lipps, 1982), measure changes in diversity (Hurlbert and Archibald, 1995), or distinguish between extinction and extirpation (Smith, 1995).
DDeden, the Deccan traps predate even the revised date of Chicxulub. So it is unlikely that they were caused by that particular impact, but it is a good thought nonetheless. Plus, if I remember correctly, I think Chicxulub's antipode is in the SE. Pacific or S. Indian ocean (I could be wrong on that though).
I agree with Chris that this wiggle room keeps this debate alive but respectfully disagree that it keeps the debate exciting. It would be much more interesting if researchers attempted to surmount the problems listed above (and others yet to come). As it stands, any evidence in support of one theory can be viewed in support of contrary theories. This is a hallmark of an intractable line of questioning. Until these issues are addressed, the discussion will go round in circles.
Michael, I half agree with what you are saying about the boundary, but BrianR is right. The whole geologic time scale was initially sketched out based on biostratigraphy. However, you are right that it should be divorced from the extinction because this is contrary to convention (Bourgeois, 1990). The boundary should be placed at the location of the first appearance of Pg taxa (though in most places this is approximately the same horizon). However, I really wish people would be more specific about which KPg boundary they are studying, as they are not interchangeable, and they are not necessarily chronostratigraphically significant with the exception of the geochronologic KPg boundary (65.95 Maa [sorry I can't remember the error bars off the top of my head], Kuiper et al. 2008).
Cited:
Alvarez, L.W., Alvarez, W., Asaro, F., and Michel, H.V., 1980, Extraterrestrial Cause for the Cretaceous-Tertiary Extinction: Science, v. 208, no. 4448, p. 1095-1108.
Bourgeois, J., 1990, Boundaries; A stratigraphic and sedimentologic perspective, in Sharpton, V.L., and Ward, P.D., eds., Global catastrophes in Earth history; An interdisciplinary conferene on impacts, volcanism, and mass mortality; Geological Society of America Special Paper 247. p. 411-416.
Hildebrand, Alan R.; Penfield, Glen T.; Kring, David A.; Pilkington, Mark; Zanoguera, Antonio Camargo; Jacobsen, Stein B.; Boynton, William V. (September 1991). Chicxulub Crater; a possible Cretaceous/Tertiary boundary impact crater on the Yucatan Peninsula, Mexico: Geology, v.19, no. 9, p.867–871.
Hurlbert, S. H., and Archibald, J. D., 1995, No statistical support for sudden (or gradual) extinction of dinosaurs: Geology, v. 23, no. 10, p. 881-884.
Kuiper, K. F., Deino, A., Hilgen, F. J., Krijsman, W., Renne, P. R., and Wijbrans, J. R., 2008, Synchronizing Rock Clocks of Earth History: Science, v. 320, p. 500-504.
Signor, P. W., and Lipps, J. H., 1982, Sampling bias, gradual extinction patterns, and
catastrophes in the fossil record, Geological Implications of Impacts of Large
Asteroids and Comets on the Earth: Special Paper - Geological Society of
America, p. 291-296.
Smith, R. M. H., 1995 Changing Fluvial Environments across the Permian-Triassic boundary in the Karoo Basin, South Africa and possible causes of tetrapod extinctions. Palaeogeography, Palaeoclimatology, and Palaeoecology. v. 117 no 1-2 p. 81-104.
Posted by: Bryan | May 20, 2009 9:53 AM