Product has been added to the basket

And Yet It Smokes

Smitsmall.jpgJan Smit and colleagues say that the evidence claiming to show that the Chicxulub crater is not K/T in age just does not exist - period!

Jan Smit writes: Here is the other side of the Chicxulub story by Jan Smit from the Vrije Universiteit Amsterdam, who worked over 10 years in the area with colleagues (Tom Roep, Philippe Claeys) and two masters students

The main issues are:

  • The age of the Chicxulub impact relative to the K/T boundary
  • Clastic deposit interpretation, tsunami or not?
  • Burrowing in the clastic layers
  • Chicxulub (Yaxcopoil-1) core interpretation.

Age of Chicxulub relative to K/T

The impact spherule deposits in eastern Mexico and the US Gulf coast are all reworked and related to the Chicxulub impact as admitted by Gerta Keller. The occurrence of the lowest deposit would therefore approximate the age of the impact. Keller claims this is 300Ka below the boundary, on the assumption that the spherule layers are separated by normal, pelagic sediments. But the latter is highly questionable. The so-called normal layers between the spherule 'layers' (none of those are continuous for more than 10m in almost all of the >40 outcrops in Mexico) are all slumped, or part of a mass flow containing numerous soft Mendez clasts, that deform and weld to one another after (re)deposition, creating the illusion or 'normal' Cretaceous hemipelagic sedimentation. Look at the correlation by one of the students of Keller and Stinnesbeck's group, (Fig1) and see that none of the so-called 'layers' correlate in any systematic way, even at such short distances. They are all slumped and discontinuous, and occur on different levels. The spherule layers, therefore, can be easily reconciled with one original layer that has been re-deposited several times in rapid succession.

Stratigraphic columns of the Mesa san Juan-Sierrita area of Peter SchulteFigure 1, Stratigraphic columns of the Mesa san Juan-Sierrita area of Peter Schulte. The spherule layers even at these short distances do not correlate, because they are locally slumped, of discontinuously mixed with Mendez soft clay clasts. Click on the image to get a bigger one.

Clastic beds - tsunami or not tsunami?

The clastic unit overlying the spherule layers, is composed of several sandstone sublayers, each graded, without interruption by normal hemipelagic sedimentation. These layers bear some similarities to turbidites, so they could be interpreted as a stacked series of amalgamated turbidite layers, if not for their variable current directions measured on the ripple beds and flute casts (Smit et al., 1996). The current directions are often 180º different, although the ripple beds deposited in a downslope, south-easterly direction, are thickest. A simple explanation by turbidity currents that usually run in a downslope direction only, is therefore unlikely. However, a series of turbidites, each triggered by different tsunami waves (each about one hour apart) and modified by the surges of the tsunami waves running on the shelf from the deep Gulf of Mexico, is a plausible explanation. The duration of deposition of the entire package could be as short as two days.

The so-called 'sandy limestone layer', a term introduced by Keller in several papers, has no sedimentological significance. It is in fact not much different from the other sandstone layers except that it is washed clean from finer mud, usually an indication of wave action (grainstone). The sand grains are later cemented with sparry calcite, which makes the sandstone a weathering resistant feature, but not a normal hemipelagic limestone.

Burrowing in the clastic layers

The burrows occurring on some of the bedding planes do not necessarily indicate prolonged time thousands of years, as preferred by Keller.

The burrows occurring in the clastic units fall into two categories: 1) Those observed by the Keller, Stinnesbeck and Ekdale group in the basal layers of the often 10m thick unit clastic unit, and 2) rather abundant burrows of different types (Fig2) in the top 15cm or so of the unit. We spent considerable time at each of the locations to find burrows at those lower levels, but they remained elusive for us.

We found, however, several pseudo-burrow types that could have been mistaken for such. 1) Recent root traces surrounded by caliche or mud and wasp nests, both rather muddy structures that mimic burrows, in particular when partially eroded. 2) synsedimentary fault structures, that are lined by rusty haloes, fault traces that, when eroded, resemble burrow traces, and 3) flame structures, infolded at the base of sandstone layers such as those that are common at the loaded base of turbidites. The infolded flames carry spherules, and when cut horizontally, may resemble burrows.

Fig.2 Figure 2. Three sandstone layer surfaces at El Penon that display surficial burrow structures. The upper surface with the hammer is at the top of the sequence, and shows Thalassinoides, Planolites burrows. The lower two surfaces show abundant Chondrites. The upper surface is probably an omission surface that has been exposed at the seafloor for some time.

We believe for these reasons that true burrows at those lower levels are non-existent.

Burrows at the upper levels are nicely tiered and each level carries a different association of burrow types (Fig2). Such is typical for a single colonisation from above, when the conditions after the tsunami-waves and turbidites have returned to normal. If the burrowing were a normal feature of these sandstone layers, the tiered levels would migrate upwards through the sandstone layers, in line with the accumulation of additional sandstone layers overprinting the higher tiers. One would expect therefore that each of the different levels would show the same association/fabric, which is not the case. Some rare types of Thalassinoides (crab) burrows penetrate at least 70cm downwards into the top of the sandstone, before spreading out over a wide area within a silt layer. Vertical burrows 70cm long were observed in the Penon and Ramones sections, that achieve this penetration from the top. We believe that there is evidence only for one colonisation phase of burrowers, after the impact-tsunami events, when there is sufficient time to colonize the sandstone surface and underlying layers.

To summarise: None of the 'burrow evidence' calls for prolonged deposition over thousands of years.

Chicxulub (Yaxcopoil-1) core interpretation (Chicxulub Scientific Drilling Project CSDP)

Keller claims to have found Maastrichtian foraminifers in a 'normal, laminated, micritic limestone' in a 50 cm interval overlying suevitic ejecta.

As one of the Principal Investigators (PIs) of the CSDP project I was put in charge by the PI team of the sample division of the 75cm interval from 793.85-794.60m (Fig3) spanning the transition from suevitic impact rocks, to lower Paleocene post-impact mudstones. 105 samples were requested from that interval, and therefore each researcher received a limited amount of sample (see Fig. 4).
Fig.3 Figure 3. Core segment of the interval 793.85 a to 794.60m, before the samples were cut. The left part is undisputed Paleocene, separated from the cross-bedded and parallel-bedded sandstone on the right by a 2cm dark clay and a burrowed hardground.
Figure 4. Overview of the samples taken from the interval 793.85 a to 794.60m, the samples marked Keller/Arz are split in two, samples 306 to 324 were analysed by J. Smit. Click on the image for a bigger one.
Figure 5. Images of four thin sections from the interval (312, 314, 315, 316 clockwise from top, see Fig 4) that Keller claims to contain -in a micrite-, foraminifers of Cf1, Upper Cretaceous age. The pictures show the grainy nature of the rocks, consistent with a sandstone. (Scale=0.5mm). Click on the image for a bigger one.

The sample levels Keller requested were also requested by the micropaleontology group of Jose Antonio Arz from Zaragoza, Spain, and the Vrije Universiteit Amsterdam. Arz received a split of each sample sent to Keller, adjacent samples were sent to Amsterdam. I have studied several thin sections per sample from that interval (see Fig. 5) but did not find foraminifers - except for some benthic types enclosed in limestone clasts, and clearly reworked from older rocks. Arz used a new technique to extract foraminifers, but found only very scarce, either non-determinable foraminifers, or foraminifers of much older, Albian age. The latter have to be reworked from older rocks.

The interval in question is composed of parallel bedded, fine-grained sandstone and somewhat coarser grained, cross-bedded dolomitic sandstones, definitely not the micritic hemipelagic mudstones Keller claims them to be. That whole interval, 794.60-794.11m, up till the overlying clay layer and Paleocene mudstones, is sandstone, and therefore by definition reworked. Keller has shown at the EGS/AGU meeting in Nice and the impact meeting in Nordlingen, Germany slides that showed cross sections of dolomite crystals, and those often grow together to mimic a cross section of a foraminifer (see Fig. 6).

Fig.6 Figure 6 Backscatter electron micrograph of dolomite (dl) overgrowth of dolomite sand grains. Bar = 50 microns

This dolomite overgrowth has the same thickness as a foraminifer shell-wall. Figure 5 shows pictures of thin sections of the interval in question, showing the sandstone character without any foraminiferal shells. I expect that Gerta Keller shows some of her finds (claimed to be of Maastrichtian foraminiferal shells) so the readers can compare both results in this forum.

The best evidence in favour of a single impact, I repeat, is in the K/T record from the US western interior. In numerous outcrops from Alberta in Canada, through Dogie Creek in Wyoming to the Raton Basin in New Mexico an iridium-enriched clay layer occurs in coal swamp deposits at the palynological K/T boundary. This clay layer has a dual nature (Izett, 1990), and consist of two layers: a lower layer that contains spherules (best seen in Dogie creek (Fig. 7) morphologicaly indistinguishable from the Chicxulub spherules from the Gulf.


Fig.7 Figure 7 Thin section micrograph from the Dogie Creek, Wyoming site. The upper layer is enriched in iridium and shocked minerals, the lower layer is full of spherules. Fingernail gives scale.

The upper layer is strongly enriched in iridium and shocked minerals, such as quartz, feldspar and zircons. The shocked zircons are shown (Krogh, 1993) to have the isotopic properties (Sm/Nd) of the pan-African basement of the Chicxulub crater. In all the mentioned localities the two layers are in contact with each other, without an intervening layer. Not even a single layer of one fall season of leaves or plant material occurs between the two layers. If the upper, iridium-rich, layer is from another impact than the Chicxulub impact, they have to be simultaneous, and have to occur on the same pan-African basement - in itself highly unlikely, but not impossible. A 300Ka separation between the two layers in all the localities, as Keller posits for the separation between the Chicxulub impact and the iridium producing impact, is therefore excluded - barring a miracle.


Conclusion There may have been multiple impacts near the K/T boundary. Some craters have been (poorly) dated around 65Ma ago. The extrusion of the Deccan traps straddles the K/T boundary. But as yet, I believe that the majority of K/T researchers, like me, do not see a shred of evidence in the sedimentary record that supports multiple impacts. Neither is there any evidence that supports the influence on the oceanic ecosystems of the Deccan traps eruptions in sediments deposited below the iridium anomaly.

The evidence overwhelmingly points to a single large impact at the K/T boundary, and the odds are that this one is Chicxulub.


J. Smit et al., in The Cretaceous-Tertiary Event and Other Catastrophes in Earth History G. Ryder, D. Fastovski, S. Gartner, Eds. (Geol. Soc. of Amer., Boulder, 1996), vol. Sp. Pap. 307, pp. 151-182.

G. A. Izett, Geological Society of America Special Paper 249, 1-100 (1990).

T. E. Krogh, S. L. Kamo, B. F. Bohor, Earth and Planetary Science Letters 119, 425-429 (1993).