Should Keller's Thin Sections Be Independently Tested?

There is an interesting dispute going on at Benny Peiser’s CCNet among some eminent scientists about the date of the Cretaceous/Tertiary (K/T) boundary. It’s not an issue that I’d ever thought about very much – although my 5-year old granddaughter has gotten interested in dinosaurs and asked me why they are “extincted”. (She’s the one that thought that Bon Appetit would be a good name for a restaurant for dogs.) The dispute caused one party, Jan Smit, to propose that the other party, Greta Keller, send thin sections of her critical sections to an independent analyst. Keller thought that this was a ridiculous proposal, saying:

Instead, Smit has called for “testing” Keller’s forams to demonstrate that they are dolomite crystals… If Smit wants to carry out such a foolish “test”, I say go ahead and use your own slides.

Without knowing anything of the debate, Keller loses me here. If an exploration geologist responded in an equivalent way to a request for an independent analysis of a drill core, you should call your broker instantly and short the stock. Now in fairness to Keller, there seems to be a lot of ingrained academic expectations of immunity from ongoing due diligence and it might be a little early to short the stock. However, to a draw a more general moral, surely Smit’s request can be construed as a form of ongoing peer review. If Smit doesn’t get anywhere with Keller, I think that the next step should be writing the journals involved and asking them to intervene.

Here is some of the recent correspondence at CCNet (and there are some other interesting comments there as well.)


Dear Benny,

Keller, Adatte and Harting mount their horses again in their quest for the holy grail: the denial that the Chicxulub impact is the K/T boundary impact, notwithstanding truckloads of positive evidence, to which Schulte et al. recently added another piece.

The Chicxulub impact coincides with the mass-extinctions and caused the worldwide iridium anomaly. In all our former debates (CCNet, 1 November 2003 and Gerta Keller c.s. have consistently refused to answer to this evidence, or dismiss it with a simple sweep of the hand.

The focal point of Keller, Adatte and Harting is their critique on Schulte’s placement of the K/T boundary in the Brazos river sections. As Keller states: “…does not justify the arbitrary assignment of the K-T boundary to the base of the event deposit in contradiction to the globally accepted defining criteria”.

But wait a minute! It is Keller, Adatte and Harting that make the crucial error of placing the K/T boundary at the first appearance of the Paleocene foraminifers, neglecting “globally accepted criteria”

Schulte’s assignment is not arbitrary, his K/T boundary placement is fully in agreement with the official K/T GSSP in Tunisia. Keller et al. claim the K/T boundary should be placed at the first occurrence of Paleocene foraminifers. Yet this is NOT the case at the K/T stratotype (GSSP) in Tunisia, where the first occurrence of Paleocene foraminifers is 5-20 cm (depending on who is cited), probably thousands of years above the K/T boundary. Therefore Keller’s statement is highly misleading. This first appearance criterion is an approximation, there are other criteria. As for biostratigraphy, the mass-extinction of foraminifers and nannofossils approximates the K/T boundary much more accurate. The official K/T boundary is DEFINED at the base of a 2-3mm thin iridium-rich layer. The K/T boundary does coincide, like elsewhere, EXACTLY with the mass-extinction of the vast majority of Maastrichtian foraminifers.

So far so good, the mass-disappearance of the Cretaceous planktic foraminifers at Brazos River coincides with the base of the event beds in Brazos River, as Schulte wrote. To state, as Keller does; “…. the remainder consists of dark laminated shale similar to the late Maastrichtian sediments below the storm event deposits” is misleading. Even untrained freshman students armed with a simple hand lens can easily distinguish the foram-loaded sediments below the event beds from the darker, foram poor sediments above the event bed.

But wait again! Keller c.s. deny that the iridium rich ejecta layer in the Pacific, Western Interior and the Tethys is the Chicxulub distal ejecta layer. According to them, that one should occur some 300ky, or between 4 and 12 meters below the KT boundary in the Tethys sections. Well, nobody has found any sign of such an ejecta layer outside the Gulf of Mexico and direct surroundings below the K/T boundary, although the Chicxulub impact is large enough to have produced such a global ejecta layer.

So why then should the Tethys-Pacific-Western Interior ejecta layer NOT be the Chicxulub ejecta layer? Keller states: “In fact, the Ir anomaly is never found in association with Chicxulub ejecta and no Chicxulub ejecta glass has ever been found and documented in K-T clay layers worldwide.”

Once again that is misleading, and hinges on a multitude of curious interpretations and twists.

Chicxulub glass has indeed almost nowhere been identified, because after 65 million years such glass, instable as it is, rarely survives that long. Pure glass has been found only at Beloc, Haiti, and extremely rare in the core of smectite spherules in Mimbral, Lajilla (Mexico). Glass devitrifies into a variety of secondary minerals over time, such as smectite, glaucony or goyazite (, but, on the other hand, its host structure is often preserved, like silicified wood from the petrified forest in Arizona preserves the original wood structure. Yet, because no wood was found, has anybody denied that the petrified forest ever was a forest with trees? The typical fingerprint of the Chicxulub glassy spherules and droplets is, therefore, not the presence of glass, but the presence of exquisitely preserved splash-forms and the remains of internal bubbles. Such forms have been found in the Pacific (GPC-3), Tethys (Agost) in the iridium-rich ejecta layer at the K/T boundary, and in the Western Interior (Dogie Creek). In the Western Interior these splash-forms occur in the sub-layer amalgamated to the iridium rich layer. The significance of the dual layer at the K/T boundary in the Western Interior is consistently downplayed or dismissed by Keller cs., yet is the prime evidence for tying the Chicxulub impact to the iridium-rich ejecta. “Schulte et al’s conflating the two events” is therefore fully justified. and supports the K/T boundary position at the base of the event bed, because the bubbly splash-forms occur abundantly at the base of the bed.

Additional evidence that the global iridium-rich layer is caused by the Chicxulub impact include shocked zircons from the iridium-rich part of the dual layer, whose age and isotopes are consistent with an origin from the Chicxulub crater (such shocked zircons are recently also found in Tethys sections)

Adatte’s comments on the sedimentology of Schulte are peppered with curious stratigraphic and sedimentologic concepts. Adatte considers the event beds a ” normal sedimentary succession”. This is a curious conclusion, regarding the presence of iridium and impact spherules in the beds. And the exquisite preservation of such “normal storm” beds in a regressive setting is highly unusual to say the least.

A statement of: “channel complex ……deposited in an incised valley …” is misleading. The event beds can be traced over more than a kilometer in the Brazos river bed. Its geometry is a blanket of sand rather than a confined channel.

“5 discrete storm beds …” can also read as: 5 discrete tsunami-surge beds, and “erosional surfaces” can also be caused by individual tsunami surges. These are not valid criteria to distinguish one from the other.

One can differ in opinion about the interpretation on the event beds being several storm beds, or the deposits from several tsunami waves. The sedimentary structures will be comparable, although not identical. However, the prime sedimentary structures justifying a storm interpretation are hummocky cross stratification HCS (oscillation driven, multidirectional). Such structures are widely quoted from the Brazos event beds (Yancey, Hansen), but are they real? Our observations of the event beds at low river levels, when the sandstones are exposed over a wide area do not agree with HCS, and display dominantly climbing ripple beds, the product of uni-directional currents from a high suspension load. (see ). Such uni-directional climbing ripples are easily explained by tsunami wave surges that each may last for an hour or more.

Whether burrow-rich upper surfaces of sand layers indicate a colonization over a long period of time, or can be caused by quick re-burrowing by excavated (storm, tsunami) organisms trying to re-occupy their normal habitat remains an undecided question. Crab burrows (ophiomorpha) elsewhere in comparable deposits in NE Mexico (Lajilla, Penon ) burrow down from the event bed surface to the upper surface of sandstone layers up to 70 cm deeper down.

And since when is a sequence boundary observable as such in an outcrop? A sequence boundary is a stratigraphic interpretation, not a fact. What is frequently observed at a postulated sequence boundary, is a burrowed hardground surface with evidence for a hiatus, and a transgressive sequence on top. Such sedimentary sequence has not been found in Brazos river at the base of the event bed.

So Keller, Adatte and Harting make a lot of noise, but present very little substance, and are plain wrong at a number of issues.

Yours sincerely,

Jan Smit

Prof. Dr. J. Smit
dept. of Sedimentology
Vrije Universiteit
de Boelelaan 1085
1081HV Amsterdam


Dear Benny

Jan Smit (CCNet Jan. 23) brings a spirited defense of Schulte et al.(CCNet Jan. 12) and his own work by rehashing old arguments he has made time and time again – arguments which have already been demonstrated to contradict empirical evidence in various publications. Here I will only address his claim that is relevant to the Schulte et al. paper, namely that the KT boundary at Brazos is correctly placed at the base of the event deposit based on the presence of Chicxulub spherules. His many other claims, including sedimentation, sea level systems tracts and KT spherules being the same as Chicxulub spherules, make little sense and have no scientific basis.

On CCNet Jan. 24 he continues pulling the wool over non-experts eyes by attempting to denigrade paleontologists (as Luis Alvarez has done in the past) by labeling them as little more than drawing “their conclusions behind the microscope, fully neglecting sedimentological and field aspects.” In this derogatory statement he includes anyone who disagrees with him – the sedimentologists, mineralogists and geochemists along with the biostratigraphers. He puts himself up as expert in all, but has done no research and published no studies on biostratigraphy for more than 20 years, nor has he published any in-depth studies on sedimentology, mineralogy or geochemistry comparable to the ones he derides, Instead, he accuses with innuendo and slander the detailed and serious research done by multidisciplinary teams of experts as not knowing what they are doing whenever the results do not agree with his own (that includes his comments on sedimentology, mineralogy, sea level, and geochemistry in CCNet Jan. 23).

In todays CCNet he raises a red herring by asking “Why can biostratigraphers not agree on the K-T boundary?” In fact, biostratigraphers do agree on the placement of the K-T boundary. More than any other boundary, the K-T boundary is very easily identified and there is hardly ever any disagreement among biostratigraphers on its placement. It is only when ideology and circular reasoning interfere that there are disagreements. Smit uses circular reasoning by claiming that Chicxulub is the K-T impact, therefore wherever Chicxulub ejecta is found must be the K-T boundary. In the process he dismisses the geochemical, lithological and stratigraphic criteria – while at the same paying them lip service as in his CCNet comment.

Let us first analyze Smit’s arguments in CCNet Jan. 24. He states that “What biostratigraphers forget is that boundary stratotypes are NOT defined on biostratigraphic criteria, but are a “golden spike” in a sedimentary succession, chosen on lithological criteria.” Evidently Smit assumes that lithological changes are time correlative, but anyone who has taken Geology 101 knows that lithological changes are time transgressive. A lithological succession changes with the local environment – it can therefore not become a globally correlateable boundary. Only when a lithological change is induced by a global event, such as the KT clay layer, the KT carbon-13 shift, the late Cenomanian carbon-13 shift in association with global anoxia and organic-rich sediments, can lithology be used as one of the criteria in conjunction with the geochemical markers.

However, even the KT clay layer and basal red layer with its Ir anomaly is dependent on local environments and its absence doesn’t necessarily mark the absence of the KT boundary. By Smit’s argument the KT boundary must always be placed at the base of the KT clay layer. All biostratigraphers agree and follow this practice, especially if this clay layer also records an Ir anomaly and the global shift in carbon-13 values.

But let us examine whether Smit and Schulte et al. follow this practice as they proclaim. For example, at Brazos there is no distinct KT clay layer. But an Ir anomaly is present above the event deposit coincident with the first Tertiary species and a mm thin rust colored sandstone layer (Rocchia et al.,l996). In addition, the global carbon-13 shift coincides with the Ir anomaly and biostratigraphic placement of the KT boundary. Yet, Smit and Schulte et al dismiss these accepted markers and place the KT boundary at the base of the event deposits on the sole basis that they contain Chicxulub spherules.

They do the same in Mexico’s sections where a clay layer and thin red layer with Ir anomaly is present above the event deposits, such as at El Mimbral, La Sierrita, La Parida and others. This KT clay and Ir anomaly also coincides with the mass extinction in planktic foraminifera and the first appearance of Tertiary species within a few cm above it. But as at Brazos, Smit and Schulte et al place the KT boundary at the base of the event deposit on the basis that the presence of reworked Chicxulub spherules there marks this as the KT boundary. They ignore Smit’s much touted identifying criteria (CCNet Jan. 24) and the fact that the original Chixulub spherule ejecta is interbedded in undisturbed marine marls 4-5 m below the event deposit (BBC Horizon 2004: What didn’t kill the dinosaurs). It is clear that while paying lip service to the defining criteria, they in fact only use their belief that the presence of Chicxulub spherules at the base of the event deposit places the KT boundary there because they assume that Chicxulub is the KT impact. The age of this impact, however, appears to predate the KT by about 300 ky (Keller et al., 2002, 2003, 2004a,b).

Smit and Schulte et al. thus conveniently dismiss lithological and geochemical data when it doesn’t fit their belief of where the boundary should be placed to be consistent with the hypothesis that the Chicxulub impact is the KT killer.

What about Smit’s slander of biostratigraphers? Who use only subjective criteria, who can never agree on first and last appearance datums? Smit trots out the brief keywords on the KT as selected by Jim Ogg for the latest ICS web page. These keywords include the Ir anomaly, the mass extinction in the major microfossil groups and dinosaurs (the latter is especially controversial and not a defining criteria since few dino bones if any can be found at the KT boundary). Jim Ogg obviously didn’t include all marker criteria for the KT boundary in his keywords. Notably absence is the carbon-13 shift, and the first appearances of Tertiary species, which are indeed key bio- and chemo-markers of the KT boundary. But neither are Chicxulub impact spherules mentioned as identifying citerion as Smit would have us belief.

Smit takes Oggs keywords as support to dismiss any and all biostratigraphers who employ the first appearances of Tertiary species as one of the markers of the KT boundary at Brazos – that indeed includes every study done by microfossil biostratigraphers to date. Even Schulte et al.s dinoflagellate expert on his team, Michael Prauss (see CCNet, Jan. 23), who excused himself from this paper. This agreement among biostratigraphers for the placement of the KT boundary in the Brazos section alone shows that evidently biostratigraphy works very well across different microfossil groups. The close coincidence of the first Tertiary species with the carbon-shift and the Ir anomaly – all above the event deposits – makes it rather clear that the KT boundary is not at the base of the event deposits.

Smits erroneously claims that:” They all five (Prauss, Tantawy, Keller, Adatte, Harting) mistakenly stress that the ONLY criteria based on first occurrences of Paleocene biota should be used..” This is at best misrepresentation by Smit since all biostratigraphers have made it clear that faunal turnovers from Cretaceous to Paleocene assemblages and the carbon-13 shift mark the placement of the boundary,

Biomarkers, first and last appearances of species and acmes, appropriately tested globally form the basis of biostratigraphy and age dating in conjunction with paleomagnetic and radiometric dates. At the K-T boundary, the mass extinction of tropical and subtropical planktic foraminifera marks a unique event. Equally unique is the first appearance of Tertiary species in foraminifera, as well as nannofossils and dinoflagellates. Smit stated that this occurs within the first 5 cm above the extinction horizon and Ir anomaly at El Kef, which is indeed very close. In fact, the first Tertiary forams appear already in the first 1 cm of the KT clay in the smaller than 63 micron size fraction (Keller et al., 1995; 2002) and coincident with the carbon-13 shift. Yet he exclaims: “Where she recognizes that first appearance of Danian species DO NOT coincide with the K/T boundary!?” Evidently he has not kept up with biostratigraphic KT studies over the past decades.

This is also evident in his comment: “As member of the subcommission on the K/T GSSP, I pledged for NOT ACCEPTING the first appearance of e.g. E. eugubina, of C. primus, knowing that these markers first occur 4-25 centimetres above the iridium anomaly…” (Disclosure: I was also a member of that subcommission). No biostratigrapher proposed on that commission that the first appearance of P. eugubina mark the KT boundary. The first appearance of this species has always been (at least for two decades) the index fossil for the P0/Pla boundary in the early Danian! Smit’s claim makes it clear that he is not familiar with biostratigraphy – while at the same time dismissing it. If Smit had carefully read Tantawy’s comment on CCNet, he would also have seen that C. primus is not a KT defining criterion. But a host of first and last appearances of species and faunal turnovers are including the nannofossil “Biantholithus sparsus the calcareous dinoflagellate Thoracosphaera operculata, the sudden change in nannofossil assemblage, the last appearance of Ankhangelskiella cympiformis and Micula decussata and/or the first appearance of Cyclagelosphaera alta.” Is Smit really so out of it with the current status of biostratigraphy? Or just arrogant in dismissing all that doesn’t fit his hypothesis?

Based on the above arguments Smit states “Biostratigraphically, the K/T boundary is therefore better defined and approximated by the major extinction horizon, and this level is in Brazos river below the event beds.” There is just one small problem with this criterion at Brazos – there is no conventional mass extinction. Planktic foraminifera, define this mass extinction in marine fossils with all tropical and subtropical species extinct (2/3 of the total assemblage) by KT time and the survivorship of environmentally tolerant taxa. At Brazos, biotic stress is very high through the late Maastrichtian because the area was a very shallow sea (less than 80m estimated by Schulte et al) excluding most specialized species. In this environment species diversity is therefore very low, 20-30 species in any given sample below the event bed, as compared with 55-60 species in normal marine environments (Keller, l989). Some specialized species (e.g., globotruncanids and rugoglobigerinids) are sporadically present and these disappear at the unconformity at the base of the event deposit. The event deposit and sediments above it were deposited in even shallower seas leading to even greater biotic stress, as evident by the sharp increase of benthic species in the benthic/planktic ratio. In this high-stress environment the disappearance of the rare and sporadic specialized species at the unconformity below the event deposit cannot be assumed to represent the KT mass extinction as Smit claims.

In contrast, the carbon-13 shift and Ir anomaly both of which coincide with the first appearance of Danian planktic foramifera, nannofossils and dinoflagellates and faunal turnovers from Cretaceous to Paleocene assemblages well above the event deposit reliably mark the KT boundary. Prauss, Tantawy and I have independently studied the microfossils in the core KT1/KT2 of Schulte et al. and found that the boundary is coincident with the carbon-13 shift at 1.6 m above the event deposit. Thus, three independent studies based on different microfossil groups all agree on the placement of the KT boundary. They agree because the KT faunal and floral turnovers and the carbon-13 shift are unique. The same was observed in other sequences at Brazos (e.g., KT3 core, and CM outcrops, Keller, l989; Barrera and Keller, l990). Smit and Schulte et al. conveniently ignore this evidence.

Now consider that reworked Chicxulub ejecta is always well below the KT and at the base of the event deposit. Based on the empirical evidence this is a clear case of two separate events – the KT and Chicxulub impacts.

Gerta Keller, Department of Geosciences, Princeton University


Barrera, E., and Keller, G., l990. Stable isotope evidence for gradual environmental changes and species survivorship across the Cretaceous-Tertiary boundary. Paleoceanography 5(6), 867-890.
Keller, l989. Extended Cretaceous/Tertiary boundary extinctions and delayed population change in planktic foraminifera from Brazos River, Texas. Paleoceanography 4, 287-332.
Keller, G., Li, L. and MacLeod, N., 1995. The Cretaceous/Tertiary boundary stratotype section at El Kef, Tunisia: How catastrophic was the mass extinction? Paleogeogr., Paleoclimatol.,
Paleoecol., 119: 221-254.
Keller, G., Adatte, T., Stinnesbeck, W., Luciani, V., Karoui, N., Zaghbib-Turki, D., 2002. Paleoecology of the Cretaceous-Tertiary mass extinction in planktic foraminifera. Paleogeogr., Paleoclimatol., Paleoecol., 178, 257-298.
Keller G., Stinnesbeck W., Adatte T. and Stueben D. 2003. Multiple impacts across
the Cretaceous-Tertiary boundary. Earth-Science Reviews 1283: 1-37.
Keller, G., Adatte, T., Stinnesbeck, W., Rebolledo-Vieyra M., Urrutia Fuccugauchi, J., Kramar, G., and Stueben, D., 2004a. Chicxulub predates the K/T boundary mass extinction. Proceedings of the National Academy of Sciences 101, 3753-37-58.
Keller, G., Adatte, T., Stinnesbeck, W., Stueben, D., Berner, Z., Kramar, U., Harting, M., 2004b. More evidence that the Chicxulub impact predates the K/T mass extinction. Meteoritics & Planetary Science 39 (7), 1127-1144,
Keller, G., Adatte, T., Berner, Z., Harting, M., and Stueben, D., 2005. Chicxulub impact predates K-T boundary in Texas and caused no mass extinction. 7th International Symposium on the Cretaceous 5-9 Sept. 2005, University of Neuchatel, Switzerland, Scientific Program and Abstracts, P. 118-119.
Remane, J., Keller, G., Hardenbol, J., Ben Haj Ali, M., l999. Report on the international workshop on Cretaceous-Paleogene transitions. Episodes 22(l), 47-48.
Rocchia, R., Robin, E., Froget, L. and Gayraud, J., l996. Stratigraphic distribution of Extraterrestrial markers at the Cretaceous-Tertiary boundary in the Gulf of Mexico Area: Implications for the Temporal Complexity of the event. In: The Cretaceous-Tertiary Event and other Catastrophes in Earth History, Geol. Society of America, Special Paper 307, P. 279-286.
Schulte, P, Speijer, R., Mai, H., Kontny, A. (2006) The Cretaceous-Paleogene (K-P) boundary at Brazos, Texas: Sequence stratigraphy, depositional events and the Chicxulub impact. Sedimentary Geology, 184, 77-109

Gerta Keller
Department of Geosciences
Princeton University
Princeton NJ 08544, USA


The Chicxulub Debate:


Dear Benny,

Gerta Keller spills a bunch of strong words in CCNet 26 Jan, but in the end, after you take away the smoke screen, almost completely agrees with Schulte and Smit about the placement of the K/T boundary!

Keller admits that tropical and specialized species disappear at the base of the event bed at Brazos river. This a major criterion to place the K/T boundary below the event beds at Brazos River.

But the funny thing is, take away the event beds, (that whatever mechanism you propose, have been deposited in a very short time), and the K/T boundary favoured by Keller and the K/T position favoured by Schulte/Smit are almost one and the same.

There are many arguments to make the iridium layer and Chicxulub ejecta one and the same but Keller never adressed them.

Neither has she ever shown any inclination to have her so-called Maastrichtian foraminifers (1) in the Yaxcopoil-1 core in the Chixulub crater tested. I have amply demonstrated (2) these to be inorganic dolomite crystals, yet Keller uses these to show that Chicxulub is older than the K/T boundary.

It is a pity she has not done so, because I think also full professors at Princeton are not beyond good scientific principles: Testing!

For detailed comments on Keller’s CCNet 26 Jan comments, see:

jan smit

Prof. Dr. J. Smit
dept. of Sedimentology
Vrije Universiteit
de Boelelaan 1085
1081HV Amsterdam


Dear Benny

I’m starting this morning with a big smile on my face after reading Jan Smit’s latest comment. He has gone way beyond science into the realm of fantasy – claiming that we almost completely agree with the placement of the K/T boundary – and further claiming that ” I have amply demonstrated these (Maastrichtian planktic foraminifera in Yaxcopoil-1) to be inorganic dolomite crystals”.

And once again he has entered the realm of personal slander and accusations “I think also full professors at Princeton are not beyond
good scientific principles: Testing!” In the absence of science, hiding behind slander is the next best thing.

What riles Jan Smit so much about the Chicxulub Yaxcopoil-1 crater core? It is simple. Keller et al. (2004a,b) demonstrated that the impact breccia and the K-T boundary are separated by a 50 cm interval of sediments that contain (1) five thin glauconite clay layers, each of which represents deposition over tens of thousands of years, thus incompatible with tsunami or backwash deposition, (2) burrows and bioturbation, which could not be present if this were a tsunami or backwash deposit, (3) late Maastrichtian carbon isotope signals and the negative shift that marks the K-T, both indicating normal sediment deposition, rather than reworking, (4) chron 29r that marks the last 500 ky below the K-T and first 200 ky above it, (5) the presence of late Maastrichtian zone CF1 foraminifera, which span the last 300 ky of the Maastrichtian, (6) the mass extinction of Cretaceous planktic foraminifera, (7) the first appearance of Danian species immediately above it, (8) absence of Ir anomaly, probably because of erosion. All this evidence points to placement of the K-T boundary 50 cm above the Chicxulub impact breccia. And most importantly, it demonstrates that the Chicxulub impact predated the K-T boundary mass extinction.

All this evidence against the hypothesis that Chicxulub is the K-T killer riles Jan Smit, who maintains that the 50 cm interval represents a tsunami deposit, backwash and crater infill after the impact. Of all this evidence, which stands alone even without the forams, he is particularly upset about the forams, which he maintains do not exist and are just dolomite crystals. He has maintained this claim despite the fact that the two foram experts he has consulted have confirmed the presence of late Maastrichtian planktic foraminifera. Smit supplied core Yaxcopoil-1 samples to Antonio Arz, who documented late Maastrichtian forams (Arz et al., 2004). He then supplied his own set of samples (Disclosure: they are the same sample set as mine) to the Cretaceous foram expert Michele Caron, who confirmed the presence of abundant Late Maastrichtian foraminifera, including the zone CF1 index species Plummerita hantkeninoides, which spans the last 300 ky of the Cretaceous. But he rejected her results, sayingt “I don’t believe it”, just as he rejected Arz results.

Instead, Smit has called for “testing” Keller’s forams to demonstrate that they are dolomite crystals. Since Michele Caron has identified them in his own samples, which are from the same set as mine, he could perform this test easily. But he hasn’t. And there is a reason why. It doesn’t make sense. To the uninitiated this test may sound like a valid proposal. But it isn’t. The foraminifera in these dolomitized sediments are recrystallized, this means that the original calcite shells are replaced by dolomite. In fact, diagenetic replacement of original test shells is common in most sediments, but this doesn’t change the fact that the forams are there, only that the original shell calcite is gone. Now Smit proposes to test the foram shells and show that they are dolomite, rather than the original shell calcite. Of course they are replaced by dolomite – but they are still forams. But he wants you to believe that this is a legitimate test to determine whether forams identified as such by several foram experts are really forams. If Smit wants to carry out such a foolish “test”, I say go ahead and use your own slides. Michele Caron will be happy to show you the forams that you tell her don’t exist.

Gerta Keller
Department of Geosciences
Princeton University
Princeton NJ 08544, USA


The Chicxulub Debate:


Dear Benny,

Since when is a call for objective, scientific testing “slandering in the realm of phantasy?” (G. Keller, CCNet, 31 Jan). Keller tries to hide behind a thin smoke screen in her refusal to let HER Chicxulub samples be scientifically tested.

It is not difficult to see through the all too transparent disguise when she exclaims at the end: “Go ahead and do a foolish test on your own slides”. When I would present an “unbiased” test of my own slides, then I would be foolish indeed.

By coincidence, I have just visited Michele Caron a few days ago at the Swissed meeting in Fribourg, and we have discussed at length with a microscope at hand whether the few foram-reminiscent forms in these slides are really the remains of foraminifers of not. Michele believes there are a few very badly preserved foraminiferal forms in there (deteriorated way beyond recognition at generic or species level) and I still have my doubts. However, Michele Caron has ample experience in identifying foraminifers in thin sections from the metamorphic “schistes lustrés” and she may have found a way to identify badly preserved specimens. But even if these forms turn out to be true foraminifers – and Arz et al 2004 have shown that there are a few reworked specimens in that core interval – then those occur in a cross-bedded sandstone and are, therefore, by definition reworked, and cannot be used for biostratigraphy.

About a year earlier I had sent to M. Caron, a renowned foraminiferal specialist, the thin sections I had made from that critical interval from impact to post-impact rocks for further analysis. I have done this to have my observations tested and to see if someone else could find foraminifers where I could not. Pure, simple, scientific testing by an experienced colleague, who had taken no stance in the debate.

Michele Caron spent weeks behind the microscope to find just a few examples that might be a badly preserved foraminiferal shell, and in one case she may be right. (See the images of these here:

Thus, when Keller quotes Michele Caron as having “… confirmed the presence of abundant Late Maastrichtian foraminifera”, she simply does not tell the truth. Nor has Caron found any “CF1 index species Plummerita hantkeninoides.”

Antonio Arz (1), who has also been “quoted” by Keller, has answered in a similar fashion in CCNet of 1 Feb. (Disclosure: Not I, but Antonio Arz has received the same sample set as Keller,
but he has dissolved all his sample material. My sample set was close, but offset from Keller’s, and cannot be used for one-on-one testing.)

I have provided M. Caron with a second set of polished thin sections, in which I earlier(2) demonstrated the zoned dolomite crystals, and I am curious to know what Caron will find in there.

Michele Caron had also earlier politely asked Gerta Keller if she could send her set thin sections to Fribourg, Switserland, so she could compare those with the set I had supplied. Thus far, Keller has failed to answer this request.

Keller, now for the first time, admits that all the calcite tests of her foraminifers in that critical interval were replaced by dolomite (3). Dolomitization, as every paleontologist knows, destroys very effectively the shape of fossils, especially microfossils, so nothing is left to recognize, and certainly not to the species level.

In the interval just above the claylayer in the Chicxulub core (794.11-793.85m) true Paleocene foraminifera are found. But also dolomite crystals can be found that are identical to the crystals in the interval 794.60-794.11m. Those crystals often pierce and destroy the foraminiferal test-walls, but never replace the original calcite wall (see

I therefore appeal again to Gerta Keller to send her set of thin sections to Michele Caron, so that all sample sets can be scientifically tested.


1) Arz J.A., Alegret, L., and Arenillas, I., 2004. Foraminiferal biostratigraphy and paleoenvironmental reconstruction at Yaxcopoil-1 drill hole (Chicxulub crater, Yucatan Peninsula). Meteoritics and Planetary Science 39(7), 1099-1111.

2) Smit, J., Gaast, S. J. v. d. & Lustenhouwer, W. T. 2004. Is the transition impact to post-impact rock complete? Some remarks based on XRF scanning, electron-microprobe and thin section analyses of the Yaxcopoil-1 Core, Chicxulub Crater. Meteoritics and Planetary Science 39(7), 1113-1126.

3) Keller, G. et al. More evidence that the Chicxulub impact predates the K/T mass extinction. Meteoritics and Planetary Science 39(7), 1127-1144 (2004).

Prof. Dr. J. Smit
dept. of Sedimentology
Vrije Universiteit
de Boelelaan 1085
1081HV Amsterdam


  1. Dave Dardinger
    Posted Feb 2, 2006 at 10:08 AM | Permalink

    I followed the original debate about the KT boundry and was convinced Alverez & son were correct. At some point they became the conventional wisdom, IMO. Likewise, the identification of the Chicxulub impact with the KT event was well proven IMO. So anyone denying it needs to be especially careful to lay out a set of well verified facts which support the new theory. While Keller presents a set of supposed facts, she doesn’t seem to be getting anywhere in proving them. Her language also does her no good. And logically one has to consider the physical processes involved in the aftermath of an impact before considering the biological consequences.

    Anyway from the last messages you have there it’s fairly clear that a couple of
    determinations could pretty well settle this argument for fair-minded people.

    1. Is Smit correct when he says,

    Dolomitization, as every paleontologist knows, destroys very effectively the shape of fossils, especially microfossils, so nothing is left to recognize, and certainly not to the species level?

    If so this pretty well knocks out Keller’s main argument.

    2. Just what is the statis of the statement by Keller that

    Keller et al. (2004a,b) demonstrated that the impact breccia and the K-T boundary are separated by a 50 cm interval of sediments that contain (1) five thin glauconite clay layers, each of which represents deposition over tens of thousands of years, thus incompatible with tsunami or backwash deposition?

    Perhaps it’s somewhere buried in this very long blog, but just what is she using as the signs of the KT boundry above this sediment layer? If it’s an iridium anomaly I’d like to see exactly how the iridium distribution looks compared to other places. I wouldn’t place much stock in the carbon 13 shift as there might be ways it could be distorted locally by one thing or another.

  2. The Knowing One
    Posted Feb 2, 2006 at 10:27 AM | Permalink

    The dinosaur-extinction debate has previously degenerated into something seriously lacking in scientific integrity. For some background, see “Political Corruption of K-T Science”, by Dewey McLean (Virginia Tech), at

    with some elaboration at


    and especially

    This last also highlights how dishonorably Keller has been treated by Science magazine in the past. It is really surprising that a magazine as prestigious as Science would act so dishonorably.

  3. Louis Hissink
    Posted Feb 2, 2006 at 8:07 PM | Permalink

    I’m watching the K-T extinction debate Steve raised with keen interest – many critics have described geology as essentially a religion – which considering the sect like argument documented above – tends to confirm.

    That said, it’s academia where all the controversy occurs since in exploration all one does is drill a hole and whap, theory proven or disproven. Mind you even exploration types sometimes tend to have an almost religious attitude to pet projects were they “know” there is a monster orebody hidden “out there”.

    The plasma guys have another take on the issue here

    Read at your intellectual peril.

    I’ll probably summarise the K-T debacle in the a forthcoming AIG NEWS since the fat lady has just started singing.

  4. Dave Dardinger
    Posted Feb 2, 2006 at 10:03 PM | Permalink


    Well, the default position is that the reason the ‘skeptic’ position there isn’t touted in Science is because it’s not scientifically sound. I know that might seem strange coming from one of the strong AGW skeptics here, but the fact is I spent a lot of time investigating the asteroid vs volcano, etc. theories quite a number of years ago and arrived at the conclusion that the non-asteroid positions just didn’t hold water.

    I came to global warming thinking I could do much the same, read a couple of books and some articles and be able to reach a conclusion. But I quickly found that the important facts and data were hidden behind lead shielded doors and only the in crowd could actually approach the holy of holies; the climate models and actually talk to them. Yes there are a few ‘open’ models and some data is available, but not enough to actually decide if the earth is really warming as fast as claimed and whether it’s likely to climb much farther. This doesn’t disprove the theory, of course, but it did turn me into a skeptic. Being told not to worry about the details unless you’re willing to spend vast amounts of time (and possibly money) is not the way to win friends and influence people.

  5. Louis Hissink
    Posted Feb 2, 2006 at 10:42 PM | Permalink


    Another way of looking at it is if the two main camps are so adamantly opposed to the other’s concept then maybe both are wrong and a third approach has to be considered.

    Happened in global tectonics – Plate Tectonics vs Expanding earthers, each screaming at each other and it got pretty nasty, much like we see now in the K-T thing and of course climate.

    Some of us never accepted either but were on the sidelines but then started looking at the data and the end result is “surge” tectonics, (not perfect by any means) that could explain the observations of both the PT and EA supporters. Most of the sideline crowd are in the exploration and mining game where dogma isn’t a useful approach to take, especially when exploration managers are breathing down your back asking “how much does it assay?”. Wonderful how such an approach can focus the mind on the essentials.

    Scepticism in this case is natural. All theories are provisional until contradicted by new data in any case. If a theory works, we use it. If it doesn’t we don’t.

  6. Posted Feb 2, 2006 at 11:56 PM | Permalink

    re #4
    I think that the distinction between scientists and activists has been blurred in the field of climate science.

  7. ET SidViscous
    Posted Feb 3, 2006 at 12:35 AM | Permalink


    I think it’s pretty clear, to be honest. They are inseparable, one in the same.

    Steve M, what’s your feeling on the work of Gould in relation to deep oil etc.?

    I always thought the work intriguing, and Gould interesting, but don’t have the background to review his work critically.

  8. Louis Hissink
    Posted Feb 3, 2006 at 12:59 AM | Permalink


    You mean Tommy Gold?

  9. Ray Soper
    Posted Feb 3, 2006 at 2:19 AM | Permalink

    Sid, I think that you mean Gold, not Gould. If you are interested in those ideas, you may be like to have a look at Vladimir Larin “Hydridic Earth” and C W Hunt, E A Skobelin and L G Collins “Expanding Geospheres”. These guys present a hypothesis that the earth was once a gas giant as Jupiter is now. The massive hydrogen atmosphere put the metals in the core in the situation of being saturated with hydrogen under great pressure. They argue that under these conditions, metals undergo a phase change and compress to perhaps one quarter or less of their volume when in their normal crystal lattice. Accordingly, they argue, the core of the earth has an SG of around 31 from this effect. Now that the hydrogen atmosphere has been abladed away, the confining pressure is released, and some of the core metals spring back to their “normal” lattice state and typical SG. This, they argue provides a mechanism for earth expansion. In the process H2 is released. As it ascends towards the surface of the earth, it passes through the SiC layer, and reacts with the Si and the C to form CH4 (methane) and SiH4 (silane). The CH4 is relatively stable and ascends to near the earth’s surface where it accumulates in traps. There are bacteria near the surface that convert (some) of the methane to long chain molecules, hence petroleum accumulations. The SiH4 is very reactive, and it reacts with O2 and H2O to form silicates, thus granites. The metals that are in the rocks that are crystallised to silicates can no longer be held in the lattice and are expelled, ascending up structural features around the granites until finally precipitated as sulphides.

    The theory also explains coalification, quartzite accumulations (Athabasca sands). Overall, an elegant and simple theory that has had almost no airtime. I have read the books and I am impressed. I also acknowledge that I probably haven’t done justice to the ideas. You can see the books at Off topic, I know, but I think pertinent. Many geologists will tell you that there is consensus, the science is settled. But, I tell you what, maybe it isn’t!!

  10. Peter Hearnden
    Posted Feb 3, 2006 at 3:08 AM | Permalink

    Fellas, you’re getting distracted by, in recent cases, the purely speculative margins of geology …Oh, and I suppose it’s a coincidence coal is full of fossils of coal forest trees?

    Anyway, get you eyes on the ball or loose it!

  11. John A
    Posted Feb 3, 2006 at 7:25 AM | Permalink

    I’m afraid Peter Hearnden has had a rare moment of lucidity.

    It’s not enough to claim that a new hypothesis explains some phenomena, it must also explain why the current theory has worked so well.

    Thus any hypothesis that purports to be scientific must explain why coal appears in seams dating from known periods of extensive plant growth (and much higher carbon dioxide levels than today) and has fossils of the plants that presumeably made the carbon deposits.

  12. Steve McIntyre
    Posted Feb 3, 2006 at 7:49 AM | Permalink

    I haven’t read Gold and I’m way too busy with other topics right now.

  13. ET SidViscous
    Posted Feb 3, 2006 at 10:22 AM | Permalink

    Yes I did mean Gold, and don’t know why I confused the two, other than Gold did work in so many areas, I thought Evolution might be one.

    Thanks Ray

    But to Peter and John. Gold’s theory, and I assume Larin’s, don’t exclude surface plant mass in Coal in fact it is required. Speaking of that, what do you guys think of the rare human skulls found in coal seams in Pennsylvania and Tuscanny.

    Fair enough Steve, just curious from a geologists perspective.

  14. John A
    Posted Feb 3, 2006 at 11:44 AM | Permalink

    Speaking of that, what do you guys think of the rare human skulls found in coal seams in Pennsylvania and Tuscanny

    You mean the infamous Ed Conrad? Completely cracked and then some.

    Apparently he’s being sued for defamation as well…impressive.

    Here’s the refuation of Conrad’s supposed “science” :

  15. ET SidViscous
    Posted Feb 3, 2006 at 12:10 PM | Permalink

    No I definitely would not say Ed Conrad who seems to be a creatonist, but on review I would withdraw the skull comment not finding any reference that can’t be traced back to some level of creatonist.

    Still does not rule out the abiogenic theory of hydrocarbons.

  16. Steve McIntyre
    Posted Feb 3, 2006 at 12:25 PM | Permalink

    Let’s not pursue this topic on this site any more if you don’t mind (and if you do mind, I’ll moderate anyway.)

  17. ET SidViscous
    Posted Feb 3, 2006 at 12:46 PM | Permalink

    No problem here. Like I said I was just curious.

  18. Louis Hissink
    Posted Feb 3, 2006 at 6:57 PM | Permalink

    What concerns me is that evidence for this theory is so scant – dolomitized forams that need acute observational skills to identify them.

    This is much like science at the thresh-hold of observation – where things are imagined to occur and where complex statistical analyses are required to extract the evidence.

    And the reason there is so much acrimony is because the physical evidence doesn’t jump out at you, so it reduces to a matter of belief, and hence stops being science.

    I mean, no one argues over the presence of the sun or moon do they.

  19. Louis Hissink
    Posted Feb 3, 2006 at 7:44 PM | Permalink


    (concluding para disappeared in the RSS feed reply !)

    So independent testing in this case would be mandatory since the evidence (chemical altered forams)is not self evident. Why Keller did not have the data independently checked before publication is another matter – and here we get into the ego area.

  20. Jaime Arbona
    Posted Feb 3, 2006 at 7:57 PM | Permalink

    Re # 18

    “I mean, no one argues over the presence of the sun or moon do they.”

    Louis, you should look here.

  21. Louis Hissink
    Posted Feb 4, 2006 at 12:59 AM | Permalink

    Re # 20


    Obviously I wrote a “courageous” statement – I should these things out beforehand…..

    In any case this Keller/Smit spat is getting interesting as I am drafting a short article on it and looking at the various other papers, which I’ll put up elsewhere once it gets published in dead tree form in March.

  22. Hans Erren
    Posted Feb 4, 2006 at 4:42 AM | Permalink

    Of course should the thin sections be indepently tested!
    Let me give you an example: In 1988 I was working in the Mineralogical Museum in Delft, which also holds holotypes of indonesian triassic foraminafera. On request from a researcher in the museum of national history museum in London, and with permission of the curator, I removed a chip from the holotype and brought it to London.
    That’s how science should work.

  23. Louis Hissink
    Posted Feb 4, 2006 at 4:52 AM | Permalink

    Re # 22


    Keller suggested Smit should have his slides tested, not hers, per se, or do you understand it differently?

  24. Hans Erren
    Posted Feb 5, 2006 at 6:09 AM | Permalink

    AFAIK smit suggested to test Kellers slides, keller responded that smit could use his own slides.

  25. Louis Hissink
    Posted Feb 6, 2006 at 6:47 PM | Permalink

    Re # 24

    Well then to be even handed, both sets should be independently checked. Interesting to see what the reaction would be. Losing thin section often happens I discovered, another archiving issue. Usually happens with absent minded academic types who don’t understand the importance of the data.

    This K-P issue remains too emotional for the moment.

  26. Graham Sowa
    Posted Apr 13, 2006 at 10:20 AM | Permalink

    I am traveling to Haiti for three months this summer and will be going to Beloc to get some samples of the K/T boundry Tektites. Does anyone know of the exact location of the outcrop or exposure? If anyone hear would like some samples I would be happy to bring them back with me.

    Please email

    Thank you,

%d bloggers like this: