An emplacement mechanism for the mega-block zone within the Chicxulub crater, (Yucatán, Mexico) based on chemostratigraphy


Corresponding author. E-mail:


Abstract– To better constrain the emplacement mechanism of the so-called “mega-block zone,” a structurally complex unit of target rocks within the Chicxulub impact structure, the stratigraphic coherence of this zone is tested using its strontium isotopic composition. Forty-eight samples across the 616 m sequence of deformed Cretaceous rocks in the lower part of the Yaxcopoil-1 core, drilled by ICDP in 2002, were analyzed for their 87Sr/86Sr isotope ratio. The oceanic anoxic event 2 (OAE2 event), located near the base of the core forms the only stratigraphic anchor point. From this point upward to approximately 1050 m depth, the 87Sr/86Sr trend shows small oscillations, between approximately 0.7074 and 0.7073, characteristic of Cenomanian to Santonian values. This is followed by an increase to approximately 0.7075, similar to the one reported in the seawater strontium curve during the Campanian. Scattered Sr isotope ratios are attributed to local diagenetic effects, such as those expected from the possible presence of hot, impact-induced dikes and hydrothermal fluid flow, originating from the thick central melt sheet. The absence of Upper Maastrichtian Sr isotope values may result from the removal of upper target lithologies during the impact cratering process. Based on these results, the displaced Cretaceous sequence in Yax-1 appears to have preserved its stratigraphic coherence. During the modification stage, it probably moved as a whole into the annular basin during collapse of the crater wall, thereby breaking up into discrete units along previously weakened detachment zones. This model is consistent with the emplacement mechanism postulated by Kenkmann et al. (2004).


Regarding the general understanding of the formation process of meteorite impact structures, most problematic are the so-called complex structures, which on Earth develop in structures more than a few kilometers across (Melosh and Ivanov 1999). Complex craters form by the gravitational collapse of an initial bowl-shaped transient cavity: the crater floor rebounds to form a central uplift, and target rocks slump from the crater rim into the annular basin along concentric faults, forming a terraced zone or mega-block zone (Melosh 1989; French 1998). The mechanism of and interaction between these collapse regimes is poorly constrained. The approximately 180–200 km wide Chicxulub impact structure (Yucatán, Mexico) is the largest complex crater known to occur during the Phanerozoic, and the corresponding impact is thought to have caused the Cretaceous-Tertiary boundary mass extinction, 65 Ma ago (e.g., recent review by Schulte et al. 2010). The structure lies buried under approximately 1 km of Cenozoic sediments, half-onshore and half-offshore on the northern Yucatán Peninsula (Hildebrand et al. 1991) (Fig. 1). As the crater was rapidly buried in a tectonically quiet area, its inner structure has been remarkably well preserved, making Chicxulub an excellent candidate for studying the formation––and particularly the modification processes––of complex crater structures. The crater’s structure is primarily known from integrating data from drill cores and seismic experiments. These data include the information obtained from wells drilled by PEMEX in the 1950s–1960s for oil exploration purposes (Y1, Y2, Y4, Y5, Y6, C1, S1), the UNAM shallow scientific wells (U5 to U7) (Urrutia-Fucugauchi et al. 1996), and a marine seismic reflection study carried out in 1996 by the British Institutions Reflection Profiling Syndicate (BIRPS) in the offshore part of the basin (Morgan et al. 1997) (Fig. 1). The Chicxulub crater structure was interpreted to consist of a broad and shallow annular trough, a terrace zone, and a central peak ring. In the seismic reflection profiles, Morgan et al. (2000) found the peak ring to coincide with a low-velocity region that dips toward the center of the crater. These authors interpreted this feature to represent the boundary between two collapse regimes: the inwardly collapsing crater rim and the outwardly collapsing central uplift. Although Collins et al. (2002) developed a model that showed the interaction of the two collapse regimes, the final result of which is in fairly good agreement with the main structural observations made at Chicxulub, additional data are needed to constrain the behavior of the target material during impact.

Figure 1.

 Location of the Chicxulub crater on the Yucatán Peninsula with location of the BIRPS seismic reflection line Chicx-A-Chicx-A1 (seismic reflection line from Bell et al. 2004) and the onshore wells drilled in and outside the crater structure.

In 2002, the International Scientific Continental Drilling Project (ICDP) drilled the crater structure to a depth of 1511 m and recovered a complete sequence of both impactite lithologies and preimpact target rocks (Urrutia-Fucugauchi et al. 2004). The Yaxcopoil-1 well is located within the annular basin at approximately 60–65 km from the crater center (Figs. 1 and 2). The drill core encountered 795 m of postimpact Cretaceous sediments and a 100 m thick impactite sequence, before reaching a 615 m sequence of Cretaceous preimpact target rocks. This Cretaceous sequence consists of interlayered limestone/dolomite/anhydrite lithologies (Dressler 2002), constituting the carbonate platform that extended over the Yucatán Peninsula in the Late Cretaceous (López-Ramos 1975). The sequence is brittly deformed (Kenkmann et al. 2003). Kenkmann et al. (2004) observed changes in the dips of the bedding plane throughout the sequence, and identified three main structural units and nine subunits. In the upper part of the sequence (894–934 m), layers were found to be in a subvertical position. Normal faults occur throughout the entire sequence and are consistent with a displacement toward the center of the structure (Kenkmann et al. 2004). The Cretaceous sequence is cut at several levels by suevitic (909 and 916 m), impact melt (1347 and 1348 m), and polymict lithic breccia (1314.7 and 1374 m), dikes (Kenkmann et al. 2004; Wittmann et al. 2004). At a depth of 1036–1037 m, a para-conglomerate or diamictite occurs. It consists of angular and subrounded fragments of anhydrite, limestone, relics of shells, and shocked quartz grains, embedded in a fine-grained dolomite matrix (Kenkmann et al. 2004). It was interpreted by these authors to show some resemblance with the Albion Formation diamictite bed in southern Quintana Roo, a carbonate breccia that forms part of the continuous ejecta blanket and that crops out some 360 km SSE of the center of the crater (Ocampo et al. 1996; Pope et al. 1999). Based on these observations, the Cretaceous sequence has been referred to as the “mega-block zone,” a stack of structural units affected by and displaced during the cratering event (Kenkmann et al. 2004).

Figure 2.

 Reflection seismic profile (Chicx-A-Chicx-A1) of the offshore part of the Chicxulub structure. Modified from Morgan and Warner (1999). Note the ambiguous position of the Yax-1 core and the asymmetry of the crater’s structure.

The structural context of Yax-1 is further deduced by projecting the well onto the offshore seismic data. Its locality corresponds to the depression zone forming the annular trough between the inner peak ring and the crater rim (Fig. 2). The Yax-1 well seems to penetrate either subcontinuous dipping reflectors (Stöffler et al. 2004), marking the terrace zone of a complex crater, or, alternatively, it encounters the chaotic reflectors near the peak ring. The problem with this (structural) approach, as previously stated by Stöffler et al. (2004), is that seismic profiles were obtained from the offshore part of the crater, whereas the Yax-1 core was drilled onshore. Therefore, the inferred structural position of the Yax-1 core presumes that the Chicxulub crater is circular and its features are symmetrical, which is not the case, as can be seen from the seismic profiles (Gulick et al. 2008; see also Fig. 2). Moreover, at this stage, mathematical modeling of the cratering process cannot precisely document the evolution of the target material forming the mega-block zone, and the resolution of the seismic data is insufficient to image such small-scale structural features. This has left considerable uncertainties concerning the origin of the mega-block zone and its behavior during the cratering process.

Formation of the Mega-Block Zone

Based on the observations made on the Yax-1 core and the inferred structural position of this drilling site, two main hypotheses have been used to account for the origin and emplacement mechanism of the mega-block zone (Fig. 3) (Kenkmann et al. 2004; Stöffler et al. 2004). Kenkmann et al. (2004) argued that the mega-block zone is parautochthonous and originated from an external zone near the rim. It moved inward and slightly downward along normal faults during the modification of the crater. Contrarily, the hypothesis put forward by Stöffler et al. (2004) explains the mega-block(s) as being part of the components that were excavated or pushed away (“megabreccia”) during the evolution of the transient cavity. Consequently, the sequence would have moved horizontally outward into the annular trough, from the more central part of the crater, as the transient cavity collapsed at the end of the excavation phase. This scenario is consistent with the rather small thickness of the impactite sequence in Yax-1 (only 100 m), in comparison with the 800 m estimated by modeling results and the observation of a shock-metamorphic overprint in the para-conglomerate (Kenkmann et al. 2004; Stöffler et al. 2004).

Figure 3.

 Two-scenario model suggested for the mega-block sequence of the Yax-1 core, based on seismic data and numerous observations made in the core. The mega-block sequence could be (par)autochthonous (arrow 1), it was initially located outside the transient cavity and moved into the annular basin along normal faults during the modification stage, slumping from the crater rim and thereby remaining stratigraphically coherent. The second scenario explains the mega-block zone as part of the excavated ejecta, with big blocks being pushed out horizontally (arrow 2) during the evolution of the transient cavity. This chaotic emplacement mechanism implies that the mega-block sequence is allochthonous. Modified from Morgan et al. (2000).

One approach to discriminate between these hypotheses is to look at the stratigraphic coherence of the sequence. If the mega-block sequence is parautochthonous (Kenkmann et al. 2004), an intact stratigraphy would be expected, whereas the chaotic emplacement mechanism proposed by Stöffler et al. (2004) leads to the presumption that the stratigraphy was strongly disordered or even overturned. Stinnesbeck et al. (2004) attempted to correlate the Cretaceous sequence constituting the mega-block zone with the normal preimpact Yucatán Platform stratigraphy outside the crater, as described by Ward et al. (1995). However, the essentially microfossil- and macrofossil-free lithologies hinder such biostratigraphic correlations. Only one clear marker horizon (the ocean anoxic event 2 [OAE2] at the Cenomanian/Turonian boundary) was identified at the base of the Yax-1 core (Stinnesbeck et al. 2004), providing just one stratigraphic anchor point. This paper reports the results of an assessment of the stratigraphic coherence of the mega-block zone using strontium isotope stratigraphy.

Strontium Isotope Stratigraphy

Strontium isotope stratigraphy has proved a workable method in dating strontium-bearing marine sediments. It relies on the assumption that the 87Sr/86Sr ratio of ocean water has evolved throughout Phanerozoic times, due to a variable input from different sources contributing to the Sr isotopic composition of the oceans. As Sr2+ substitutes for Ca2+ during chemical precipitation of marine carbonates and evaporites, the 87Sr/86Sr ratio of the precipitate reflects the 87Sr/86Sr ratio of the seawater at the time of deposition, assuming no isotopic fractionation has occurred (Faure and Powell 1972). Over the past decades, a marine reference curve has been constructed displaying the variation of the Sr isotopic composition of the oceans throughout the Phanerozoic (Burke et al. 1982; Smalley et al. 1994). McArthur et al. (2001) have recently optimized the curve using the LOWESS fit. For the Cretaceous sequence constituting the mega-block zone, the technique is of particular interest, due to (1) the improved resolution of the Sr isotope seawater curve in Mesozoic to Cenozoic times, and (2) the almost uninterrupted increase in the seawater 87Sr/86Sr ratio during this interval (Fig. 4). The working hypothesis is that, in the case of intact stratigraphy of the mega-block zone, the experimental Sr isotope ratio trend should follow the marine reference curve in the time window between the lowermost anchor point (OAE2) and the uppermost Cretaceous (time of impact).

Figure 4.

 Part of the marine reference curve with variation of 87Sr/86Sr throughout the Cretaceous (modified from McArthur et al. 2001). The gray overlay section highlights the part of the trend expected for an intact stratigraphy. The white full line highlights part of the curve found in the experimental Sr isotopic trend along the mega-block zone (Fig. 5).

Although the seawater 87Sr/86Sr ratio is incorporated in the CaCO3 crystal lattice without isotopic fractionation, diagenetic processes subsequent to deposition may modify these ratios (Veizer 1989). Assessing the effect of diagenetic alteration on the preservation of the primary Sr isotopic signal in marine carbonates is therefore critical when using 87Sr/86Sr ratios for comparison with the strontium isotope seawater reference curve. In this work, a petrographic study, along with measurement of stable isotope ratios (δ13C and δ18O), has been carried out to evaluate the degree of diagenetic alteration of the mega-block samples, and its potential effect on the 87Sr/86Sr ratio.

Analytical Methods

Forty-eight samples were taken along the 616 m mega-block sequence. Samples were polished and cleaned with ultrapure 18.2 MΩ cm water (obtained from an Element Milli-Q system; Millipore, USA); clean samples were crushed and further ground to powder using a simple agate mortar and pestle. Approximately 50–200 mg of carbonate powder was dissolved in 4.5 mL of 1 M HCl in precleaned teflon beakers (Savillex). From anhydrite samples, about 15–25 mg was dissolved in ultrapure 18.2 MΩ.cm water. The dissolved (carbonate) fraction was pipetted off and centrifuged, to prevent any possible contribution from Sr leached from clay fractions in further analytical steps involving strong acid attack. The sample solution was evaporated to near-dryness, followed by dissolution of the precipitate in 2 mL of 7 M HNO3 on a hotplate at 70 °C. Extraction chromatography was applied to isolate the Sr fraction from its concomitant matrix, using the commercially available strontium-specific extraction chromatographic resin Sr spec, and following the separation protocol of De Muynck et al. (2009). After loading of the sample digest, dissolved in 7 M HNO3, onto the resin, the resin was rinsed with 5 mL of 7 M HNO3 solution to remove matrix elements, while the resin retained the Sr. The purified Sr fraction was subsequently stripped off the resin in a quantitative manner through rinsing with 9 mL of 0.05 M HNO3. The purified Sr fractions were measured using a Thermo Scientific Neptune multicollector ICP-MS instrument at the Department of Analytical Chemistry of Ghent University.

All samples were run in a sample-standard bracketing sequence with a 200 μg L−1 Sr isotopic standard solution of NIST SRM 987 SrCO3. The Sr content in the samples and the standard was matched within ±10% to avoid any effect from the analyte content on the extent of instrumental mass discrimination. Blank Sr signals were negligible compared with the Sr intensities for standards and samples (<0.1%), and therefore blank correction was unnecessary. Rinsing the sample introduction system for a few minutes with 0.14 M HNO3 between consecutive runs minimized the memory effects. The intensities obtained for 82Kr+, 83Kr+, and 85Rb+ were used to correct for the Kr interferences on the m/z ratios of 84 and 86, and for the Rb interference on m/z 87. Russell’s law was used for mass discrimination correction, assuming an invariant 86Sr/88Sr ratio of 0.1194. The corrected 87Sr/86Sr ratios are displayed in Table 1. Repeated analysis of NIST SRM 987 SrCO3 yielded an average 87Sr/86Sr ratio of 0.710259 with 2s = 0.000030 (n = 16). This is in full agreement with the accepted 87Sr/86Sr ratio of 0.710248 for this reference material (Thirlwall 1991). Based on 30 dynamic cycles of data collection, for each reference sample, the internal precision (standard deviation) was between ±0.000017 and ±0.000051.

Table 1.   Samples analyzed for 87Sr/86Sr, δ13C, and δ18O isotope ratios with associated uncertainties.
Sample no.LithologyDepth (m)87Sr/86SrSERSDδ13C (‰) calδ13C (‰) dolδ18O (‰) calδ18O (‰) dol
  1. Notes: Cal = calcite; Dol = dolomite; n.a. = not analyzed.

  2. Lithological subdivision from Dressler (2002).

Yax_1497Limestone897.030.7077870.0000070.00252.33 −7.64 
Yax_1502Limestone907.490.7075100.0000080.00281.94 −3.43 
Yax_1503Limestone909.210.7077480.0000060.00212.55 −7.72 
Yax_1505Limestone914.070.7075370.0000140.00471.87 −3.16 
Yax_1510Calcarenite924.30.7074940.0000080.00281.98 −2.68 
Yax_1515Calcarenite935.210.7074830.0000060.00212.02 −2.65 
Yax_1522Calcarenite950.230.7074660.0000100.00342.57 −2.40 
Yax_1533Calcarenite974.270.7075450.0000150.00532.30 −4.30 
Yax_1545Calcarenite1002.440.7074820.0000060.00192.05 −3.20 
Yax_1547Calcarenite1004.910.7077240.0000150.00521.64 −7.45 
Yax_1564Calcarenite-anhydrite1045.880.7073730.0000140.00503.03 −2.95 
Yax_1689Calcarenite1397.890.7075080.0000110.00384.17 −5.13 
Yax_1690Calcarenite1400.780.7073900.0000180.00624.35 −4.51 
Yax_1691Calcarenite1407.870.7074050.0000070.00253.91 −4.29 
Yax_1695Calcarenite1417.450.7074110.0000160.00552.87 −4.87 
Yax_1709Calcarenite1456.790.7074250.0000110.00371.21 −5.23 
Yax_1721Calcarenite1489.710.7074260.0000110.00384.89 −3.86 

The stable isotopic compositions of carbon and oxygen were measured in the same batch of powdered whole rock samples (Table 1). Analysis was restricted to samples consisting purely of carbonate, while mixed carbonate-anhydrite or purely anhydrite samples were excluded from analysis. Approximately 10 mg of powder was treated with 100% orthophosphoric acid, to liberate CO2 from the sample. Evolved CO2 was quantitatively isolated through cryogenic trapping in a vacuum line. The δ13C and δ18O were measured using a Finnigan Delta E Dual Inlet mass-spectrometer at the Department of Geology of Vrije Universiteit Brussels. A sequential double extraction procedure was applied to the samples containing both calcite and dolomite phases. First, the sample was reacted with 100% orthophosphoric acid at 25 °C for 3 h, to liberate CO2 from the calcite phase. After quantitative isolation of this CO2, the reaction was continued for 48 h to liberate the CO2 from the dolomite phase. All rock samples were analyzed in duplicates. Only results on CO2 batches large enough to produce at least a 100 mV signal are reported here. Based on duplicate analyses, the analytical uncertainties (2 SD) for δ13C are ±0.10‰ for the calcite phase and ±0.12‰ for the dolomite phase. For δ18O, analytical uncertainties are ±0.20‰ and ±0.21‰ for calcite and dolomite phases, respectively. All values reported in Table 1 are the mean values of duplicate analyses and relative to the V-PDB standard.

Petrographic analyses were carried out on polished thin sections with optical light transmitted microscopy and scanning electron microscopy (JEOL model JSM-6400 at the Department of Metallurgy, Electrochemistry and Material Science of Vrije Universiteit Brussels). Compositional characteristics were revealed by Energy Dispersive X-ray elemental analysis using a Li-Si detector (Pioneer) connected to the same SEM instrument.


Sr Isotope Ratios

The mass bias-corrected 87Sr/86Sr ratios are reported in Table 1. Strontium isotope ratios over the Cretaceous sequence vary between approximately 0.7073 and 0.7078 (Table 1). In the lowermost part of the Cretaceous sequence, the 87Sr/86Sr ratios yield values of approximately 0.7074. Upward from this part of the drill core, the Sr isotopic composition decreases to approximately 0.7073, after which it rises again to values of up to approximately 0.7078.

Stable Isotopes

The C and O isotopic values are highly scattered along the sequence. Carbon isotope ratio values are positive and range from 1.17‰ to 4.89‰. Oxygen isotope ratio values are negative and range from −7.72‰ to −0.62‰. The upper part of the mega-block sequence is characterized by δ18O values ranging from −2‰ to −4‰, whereas the lowermost part of the sequence exhibits more negative δ18O values of approximately −5‰. Disparities between the oxygen isotopic signal of calcite and dolomite phases within mixed dolomite-calcite samples are only minor. Anomalous δ18O values occur in samples Yax_1497, Yax_1503, and Yax_1547, shifting toward more negative δ18O values down to −7.72‰. These shifts do not correlate with excursions in the δ13C signal.


Caution is required when interpreting 87Sr/86Sr isotope ratios for chronostratigraphic reconstructions. Not only may diagenetic alteration modify the original Sr isotopic signature, but problems may also arise during the analytical treatment of the sample. This could involve leaching of radiogenic Sr during acid attack on the noncarbonate residue, especially from K-rich clay minerals (Veizer 1989). As a result, Sr isotopic ratios may exhibit large disparities from the seawater reference curve. The integrity of the Sr isotope ratios is discussed in the section below by integrating the Sr isotope ratio data with stable δ13C and δ18O isotopic analysis and petrographic examination.

Comparison With the Marine Reference Curve

Stinnesbeck et al. (2004) identified only one clear correlatable biostratigraphic horizon in the depth range from 1455 to 1495 m (the Oceanic Anoxic Event 2-OAE2), inferring a Cenomanian-Turonian age for the lowermost part of the Cretaceous mega-block zone of Yax-1. Samples analyzed from this depth interval exhibit 87Sr/86Sr isotope ratios ranging from 0.70744 to 0.70742 (Fig. 5). When compared with the marine reference curve for the Cretaceous, these are values typical for the Cenomanian/Turonian boundary. Consequently, the match between the age as inferred from the marine reference curve and the measured 87Sr/86Sr in the samples provides one anchor point (or age) in the stratigraphy of the Cretaceous sequence within Yax-1 in full agreement with the biostratigraphy. From this point on, the observed trend shows small oscillations, ranging from 0.7074 to 0.7073, which are characteristic of Cenomanian to Santonian values. In the upper part of the sequence, Sr isotopic ratios reach a maximum value of 0.7078. When comparing the experimental 87Sr/86Sr versus depth profile with the marine strontium reference curve (McArthur et al. 2001) (Fig. 4), the trend reflects the 87Sr/86Sr variation from Cenomanian/Turonian to the Late Cretaceous.

Figure 5.

 Compilation figure for the mega-block zone. A) δ13C-reference curve for the Cretaceous (modified from Jarvis et al. 2006) based on the English Chalk section. B) δ13C values for the mega-block sequence. The Oceanic Anoxic Event 2 (OAE2), identified near the base of the Yaxcopoil-1 core, is confirmed by the presence of a positive excursion in the δ13C signal that characterizes the OAE2 globally. The second positive excursion may reflect the “Late Turonian Events.” The whole sequence in the 1300–1500 m depth interval contains organic-rich limestones. C) 87Sr/86Sr variation with depth for the Cretaceous mega-block interval in Yaxcopoil-1. D) δ18O ratios for the Cretaceous mega-block interval. Outliers in the Sr curve correlate with negative excursions in the oxygen isotopes.

On first approximation, it would seem that the match between the marine reference curve and the general experimental trend implies that the mega-block sequence is stratigraphically intact. It must be pointed out that the variation in the Sr isotope ratios obtained is presented as a function of depth, whereas the marine reference curve is expressed as a variation of Sr isotope ratios with time. If the assumption is correct that the observed trend in 87Sr/87Sr with depth matches the marine reference curve from Cenomanian/Turonian to the Late Cretaceous, this implies that the samples in the 250 m depth interval (1050–1300 m) span a time period of only 3 Myr (91 to 88 Ma). In the Yax-1 core, massive layers of anhydrite and dolomite dominate this depth interval. An interesting feature of most evaporites is that the time period for deposition and accumulation is usually very short (Schreiber et al. 2007). Modern deposition rates in shallow water are up to 1–2 m per 1000 yr for CaSO4 (Schreiber and Hsu 1980). During the Mid-Cretaceous, such evaporites were deposited in a sabkha environment on the shallow Yucatán carbonate platform (Stinnesbeck et al. 2004), and indeed must have had a very high sedimentation rate.

Some 87Sr/86Sr ratios in the upper part of the sequence differ from the steep, monotonously rising trend expected for the late Cretaceous. Samples Yax_1497, Yax_1503, and Yax_1547 differ from the trend as they represent comparatively higher Sr isotope values (Fig. 4). These anomalies correlate with three pronounced negative excursions in the oxygen isotopic trend (Fig. 5), providing strong hints that the increased 87Sr/86Sr ratios result from diagenetic overprinting. The diagenetic effect is confirmed by microscopic examination of the samples: all three samples exhibit a completely recrystallized carbonate matrix, consisting of anhedral, vermiculate calcite and dolomite, and rhombs of replacement dolomite (Fig. 6).

Figure 6.

 Microphotographs of representative samples along the mega-block interval. A) Yax_1497 (from depth 897.03 m). Complete recrystallization of the sample. Matrix composed of anhedral, irregular, wavy calcite en dolomite. B) Yax_1497 (XPL). Porous matrix and fissures facilitated diagenetic and hydrothermal fluid flow. C) Yax_1547 (from depth 1004.91 m) fractured anhydrite and replacement dolomite rhombs. Partial rhombs (1) with corroded rims (2, arrow) and core replacement (3). D) BSE-mode image of Yax_1547. Dolomite rhombs (Dol) are replaced by calcite (Cc) E, F, G) Yax_1502 (from depth 907.49 m). Biomicritic limestone. Fossils are still recognizable, but poorly identifiable. Recrystallization occurred in equilibrium with the host rock. Dol = dolomite; Cc = calcite; Anh = anhydrite.

Sample Yax_1497 is from 897 m depth. The contact between the Cretaceous sequence and the overlying impactite sequence is slightly above 894 m. The rather porous matrix is crosscut by microfractures. Such high permeability is a key factor in providing channels and traps for hydrothermal fluids (Newhouse 1969; Zürcher and Kring 2004). In Yax-1, hydrothermal alteration is evident from extensive potassium metasomatism in the impactite units (e.g., Hecht et al. 2003) that display high porosities, and within the suevite dikes, which crosscut the Cretaceous sequence. The hydrothermal system at Chixculub is thought to originate from the combined effects of preexisting saline brines and heat radiating from the thick, central melt sheet (Zürcher and Kring 2004).

Sample Yax_1503 stems from a depth of 909 m. Precisely at the depths of 909 m and 916 m, strongly altered suevitic dikes cut the mega-block zone (Wittmann et al. 2004). According to Wittmann et al. (2004), those dikes are surrounded by a contact aureole, implying that their emplacement/injection as hot material probably affected the surrounding lithologies. Moreover, around these dikes, steeply inclined fractures filled with phyllosilicate minerals are frequent within the host (carbonate) rock (Wittmann et al. 2004). Those secondary minerals probably formed as a result of the diagenetic effect of the hot siliceous, glassy components from the suevite dikes. As these phyllosilicates contain an excess of radiogenic 87Sr, the pore waters from the host rock probably interacted with and subsequently became enriched in 87Sr.

Sample Yax_1547 corresponds to a depth of 1005 m. The excellent agreement of the duplicate result indicates that the deviating 87Sr/86Sr result obtained for this sample is not an analytical artifact. The strong correlation with the anomalously depleted δ18O value implies a diagenetic overprint. This is supported by microscopic examination indicating complete recrystallization (Fig. 6). Dolomite rhombs are corroded at the rims, and partially to almost completely replaced by calcite. Moreover, backscattered electron (BSE) imaging and X-ray elemental analysis using Energy Dispersive X-ray analysis (EDX) revealed the presence of rare micrometer-sized halite crystals. Lüders and Rickers (2004) attributed the presence of halite crystals in the suevite and upper Cretaceous units of the Yax-1 core to migration of diagenetic saline brines. At this stratigraphic level, fluid migration was facilitated by a brecciated interval occurring precisely at a depth of 1005 m as observed by Stinnesbeck et al. (2004).

A relatively small shift toward higher 87Sr/86Sr values also occurs at a depth of 1398 m (sample Yax_1689). This sample is characterized by relatively negative oxygen isotopic values. Unusual amounts of non-carbonate residue (15 wt%) could have contaminated the sample during acid-digestion dissolution. Indeed, SEM-EDX microanalysis reveals that the sample contains an abundant fraction of K-feldspar (Fig. 7). An interesting observation is that samples Yax_1689 and Yax_1721 have an overall similar appearance. They both originate from the stratigraphic interval (1400–1500 m) enriched in organic matter. These lithologies consist of bituminous limestone, with abundant framboidal pyrite and K-feldspar in the matrix (Fig. 7). However, in contrast to Yax_1689, the 87Sr/86Sr ratio of Yax_1721 is consistent with the expected Sr isotope ratio for this stratigraphic interval. This supports the idea that leaching of radiogenic Sr from noncarbonate phases may not necessarily have contributed to the origin of the elevated Sr isotope ratios in Yax_1689. Moreover, a striking feature is the presence of unusual, black clastic polymict dike breccias in the 1397.5–1398.5 m interval (Wittmann et al. 2004). Although they do not contain impact-derived matter (contrary to other polymict dike breccias in the Yax-1 core), these breccias are slightly enriched in potassium content (Wittmann et al. 2004). The presence of these black, K-enriched clastic dikes close to sample Yax_1689 may account for the increased 87Sr/86Sr ratio observed in Yax_1689: during fluid-rock interaction, K-rich fluids may impart radiogenic 87Sr into the host rock and subsequently increase its 87Sr/86Sr ratio.

Figure 7.

 Backscattered electron-images obtained by EDX. A) Yax_1547 (from depth 1004.91 m). Small halite crystals in recrystallized calcite and dolomite matrix. B) Yax_1689 (from depth 1397.89 m) abundant framboidal pyrite and K-feldspar and carbonate disseminated in matrix of organic matter. C) Yax_1721 (from depth 1489.71 m) has overall same appearance as Yax_1689. Mineral abbreviations Py = pyrite; Cc = calcite; Kfs = K-feldspar.

The observation of alteration in samples Yax_1497, Yax_1503, and Yax_1547 strongly contrasts with samples Yax_1502 and Yax_1578, which represent biomicritic limestones. Although the latter are recrystallized, remnants of microfossils are still recognizable, although poorly identifiable (Figs. 6E and 6G). In addition, oxygen isotopic values of Yax_1502 and Yax_1578 are relatively low and consistent with the δ18O values reported for Mid- to Late Cretaceous marine limestones (Veizer et al. 1999). As their 87Sr/86Sr ratios are consistent with the coeval marine Sr isotopic signal expected at the inferred stratigraphic level, this supports the idea that diagenesis took place in a rock-dominated system, and that the primary 87Sr/86Sr values of these samples have most likely been retained.

The δ13C values across the mega-block zone range from 1.17‰ to 4.89‰. This corresponds to values reported for the Mid- to Late Cretaceous marine carbonates (Arthur et al. 1985). As carbon isotopic signals are commonly less prone to diagenetic modification compared with δ18O (Anderson and Arthur 1983), the carbon isotopic values were most likely buffered by the precursor limestone.

When unaltered, carbon isotope ratios may provide supportive information for stratigraphic interpretations, through comparison with the carbon marine reference curve (Arthur et al. 1985; Veizer et al. 1999; Jarvis et al. 2006). However, approximately 50% of our mega-block samples comprise anhydrite; an incomplete record of δ13C thus limits the use of carbon isotope stratigraphy as a proper correlation tool. In addition, unlike the marine Sr isotopic signal, spatial variability of oceanic δ13CCO2 (Kroopnick 1985) and biological factors in shell formation (McConnaughey and Whelan 1997) result in a significant spread of δ13C values in marine carbonates throughout time (Veizer et al. 1999). Consequently, the large sampling interval in the mega-block sequence (ranging from 5 to 40 m) obscures the recognition (and correlation) of rather short-term fluctuations that characterize the evolution of the marine δ13C signal. In this study, the marine carbon reference curve is of some interest due to the presence of Cretaceous “Oceanic Anoxic Events” (e.g., Schlanger and Jenkyns 1976; Arthur and Schlanger 1979) in the inferred stratigraphic framework for the mega-block zone. These OAEs are short-time (1Ma or less) periods of exceptional organic matter burial and preservation (Schlanger and Jenkyns 1976; Locklair et al. 2011). They are thought to result from enhanced burial of organic matter during ocean anoxia, storing isotopically light organic carbon in the sedimentary record and enriching the residual ocean-atmosphere reservoir in isotopically heavy carbon (Jones and Jenkyns 2001). In the global carbon isotope record of marine carbonates, they are expressed as prominent positive δ13C excursions. Such OAEs may serve as correlation horizons.

The OAE2, at the Cenomanian/Turonian boundary, was recognized by Stinnesbeck et al. (2004) at the base of the Yacxopoil-1 core (1490 m). It is characterized by the presence of organic-rich, hemipelagic limestones, the presence of planktic foraminifera that define the biozone R. cushmani, and a high TOC (total organic carbon) of more than 7% (Stinnesbeck et al. 2004). The positive δ13C excursion marks the OAE2. It is followed by an abrupt decrease and subsequent positive shift in the 1477–1338 m interval. The second positive excursion may correspond to the “Late Turonian Events” defining the long-term positive excursion in the Upper Turonian (Jarvis et al. 2006). Although not visible in Fig. 5, the amplitude of the Late Turonian Events varies throughout different sections. In the δ13C profiles from Gubbio (Italy), the amplitudes for the Turonian Events and the OAE2 are very similar. A more detailed overview is found in Jarvis et al. (2006).

Explanations for the scattered 87Sr/86Sr ratios within the uppermost part of the mega-block zone could be found in the diagenetic modification of the signal. Another possibility is that the sudden jump to higher values indicates that the blocks in the uppermost part of the mega-block zone in Yax-1 originate from different stratigraphic levels. In this scenario, the uppermost part of the mega-block zone forms part of the allochthonous ejecta or mega-breccia. This would also explain the occurrence of a shock metamorphic overprint within the paraconglomerate (Kenkmann et al. 2004), which would represent the basal fragment of an allochthonous mega-block sequence starting at approximately 1038 m. However, this would imply that structural boundaries are to be identified and that these should correlate with lithostratigraphic changes. Kenkmann et al. (2004) advocated the existence of structural boundaries (consistent with changes in lithostratigraphy) within the upper part of the mega-block zone at depths of 894, 916, 934, 1014, and 1038 m. In addition, the proposed structural boundaries frequently coincide with horizontal core breaks (Kenkmann et al. 2004). Nevertheless, these structural boundaries would not necessarily explain the excursions to higher Sr isotope values at 897, 909, and 1005 m. These positive excursions basically take place within a defined, coherent structural unit, unless later sealing of decoupling horizons occurred due to cementation or other factors hiding more of these structural boundaries (Kenkmann et al. 2004). Until further examination of the core can be carried out, it seems, for now, more plausible to attribute the discussed elevation of the 87Sr/86Sr ratios to local diagenetic events.

If the assumption of the mega-block sequence being stratigraphically coherent is correct, and the higher Sr isotope ratios of samples Yax_1497, Yax_1503, and Yax_1547 are attributed to diagenetic overprinting, this implies that the upper part of the mega-block zone yields maximum 87Sr/86Sr values of approximately 0.7075. This is not consistent with the expected Late Maastrichtian age of the uppermost part of the target, where Sr isotopic values of approximately 0.7078 are expected (Fig. 4). The uppermost Cretaceous is present at many sites around the crater (Pope et al. 1999; Smit 1999; Grajales-Nishimura et al. 2000; Fouke et al. 2002). Considering the position on the carbonate platform, there are no clear indications of a depositional hiatus at this location, which, in addition, would span the entire time interval from Mid-Campanian to Latest Maastrichtian (time of impact). The structural position and deformation pattern of the Yaxcopoil-1 core (Kenkmann et al. 2004) may yield a possible explanation for the missing Cretaceous: due to a relatively near-transient cavity rim position, upper target lithologies may have been separated from the lower target lithologies. At this location, interference of the shock wave with the free surface could have triggered weak spallation, leading to subhorizontal detachments within the mega-block zone (Kenkmann et al. 2004). Such detachments may be explained by the difference in strength of the mega-block lithologies. The alternating competent carbonate and weaker, more incompetent anhydrite succession within the target lithologies would highly favor this type of movement. This is in agreement with studies in the Ries impact structure (Kenkmann and Ivanov 2006) that have demonstrated that the uppermost target layers beneath the ejecta blanket are mechanically decoupled along incompetent beds. The combination of both the dragging of the fast moving ejecta curtain (Kenkmann and Ivanov 2006), which erodes the underlying substrate, and weak spallation (Kenkmann et al. 2004), could have resulted in the removal of the upper target lithologies of late Cretaceous age.

Conclusion and Emplacement of the Yax-1 Target Rock Sequence

Three arguments support the idea that the primary 87Sr/86Sr isotope ratio signal along the mega-block zone has been preserved (1) the OAE2, identified at the base of Yaxcopoil-1, exhibits 87Sr/86Sr and δ13C values that match the strontium and carbon marine reference curves, (2) 87Sr/86Sr isotope ratios throughout the entire mega-block sequence are consistent with expected ratios for the inferred stratigraphic interval as indicated by the seawater reference curve, and (3) excursions in the Sr isotopic trend are most likely the result of diagenetic effects as they correlate with depleted oxygen isotopic values. Sr isotope stratigraphy thus supports the hypothesis that the stratigraphy of the Yax-1 carbonate/evaporate sequence is preserved, inferring a (par)autochthonous nature for the mega-block zone as postulated by Kenkmann et al. (2004). The following mechanism is suggested for the emplacement of the Cretaceous sequence constituting the mega-block zone (Fig. 8):

Figure 8.

 Emplacement model for the mega-block zone following Kenkmann et al. (2004). During the formation of the transient cavity, the mega-block zone is located close to, but outside, the cavity rim, where interference with the rarefaction wave causes spallation and decoupling along incompetent beds. Shallow excavation along with the dragging of the ejecta curtain causes the removal of uppermost target lithologies, and additional decoupling within the underlying beds. The whole sequence of Cretaceous decoupled mega-blocks moves inward the annular basin, and discretely breaks up into structural units along the decoupled horizons.

  • 1 Formation of the transient cavity and outward excavation flow. The mega-block zone is located near (but outside) the transient cavity rim (Kenkmann et al. 2004).
  • 2 Interference of the rarefaction wave with the free surface causes spallation (Kenkmann et al. 2004). In combination with the fast moving ejecta curtain, which has a strong horizontal component (Kenkmann and Ivanov 2006), the upper target lithologies are removed from the sequence and become incorporated into the erosive flow of the continuous ejecta blanket. Thick polymict breccias dominated by carbonate and anhydrite platform lithologies have been recovered in three wells of the UNAM drilling program (U5, U6, and U7) close to the crater rim, at about 120–150 km from the crater center (Urrutia-Fucugauchi et al. 1996). At Albion island (Belize) approximately 360 km SSE from the crater center, the continuous ejecta blanket consists of a diamictite bed with large limestone and dolomite boulders from Yucatán Platform lithologies (Ocampo et al. 1996). The studies of Pope et al. (1999) and Wigforss-Lange et al. (2007) demonstrated that these boulders were most likely allogenic in origin.
  • 3 Interference of the rarefaction wave with the free surface also causes weaker spallation (Kenkmann et al. 2004) and subsequent horizontal zones of weakness to be formed (with discrete detachment) within the lower parts of the Cretaceous sequence. Additional dragging by the high velocity ejecta curtain causes more discrete movements within the lower part of the sequence (Kenkmann and Ivanov 2006).
  • 4 During collapse of the crater wall, the entire sequence moves into the annular trough along normal faults. It discretely breaks up into separate blocks along pre-existing weakened detachment zones as discussed by Kenkmann et al. (2004).


Acknowledgments–– We thank Kris Latruwe for assistance with MC-ICP-MS measurements and Oscar Steenhaut for technical support with SEM-EDX. R. Skala and G. Shields are gratefully acknowledged for reviewing this manuscript. We thank W. U. Reimold for helpful suggestions and careful editing that substantially improved this manuscript. This project formed part of the first author’s master dissertation and was supported by FWO Research Foundation––Flanders (project G.0021.11 & G.A078.11N to Ph. C). We thank Research Foundation-Flanders (FWO) for funding a Ph.D. fellowship to J. Belza and a Postdoctoral Fellowship to S. Goderis. J. Belza thanks the VUB research council for support.

Editorial Handling–– Dr. Uwe Reimold