cathodoluminescence petrography and isotope geochemistry

Cathodoluminescence petrography and isotope geochemistryof KT impact ejecta deposited 360 km from the Chicxulub crater,at Albion Island, Belize

BRUCE W. FOUKE*, AUBREY L. ZERKLE*, WALTER ALVAREZ , KEVIN O. POPEà ,ADRIANA C. OCAMPO§, RICHARD J. WACHTMAN*, JOSE MANUEL GRAJALESNISHIMURA– , PHILLIPE CLAEYS** and ALFRED G. FISCHER *Department of Geology, University of Illinois, 245 Natural History Building, 1301 W. Green Street,Urbana, IL 61801, USA (E-mail: [email protected]) Earth and Planetary Science, University of California Berkeley, 307 McCore Hall, Berkeley, CA 94720, USAàDepartment of Geology and Geophysics, University of California, Berkeley, CA 94720, USA§Jet Propulsion Laboratory, California Institute of Technology, MS183-601, 4800 Oak Grove, Pasadena,CA 91109, USA, and National Aeronautics and Space Administration, Headquarters, Code SD,Washington, DC 20546, USA–Instituto Mexicano del Petroleo, Eje Lazaro Cardenas #152, DFCP 07730, Mexico**Department of Geology, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium Department of Earth Sciences, University of Southern California, 3651 University Avenue, Los Angeles,CA 90089, USA

ABSTRACT

The depositional and diagenetic history of Cretaceous–Tertiary (KT) impact

ejecta deposited 360 km from the Chicxulub crater, at Albion Island, Belize,

has been investigated using integrated cathodoluminescence and isotopic

analyses. A quarry exposes 26 m of Upper Cretaceous Barton Creek Formation

dolomitized marine limestone overlain by 16 m of dolomitized Albion

Formation impact ejecta. The Albion Formation consists of a lower fine-

grained »1-m-thick spheroid bed and an upper 15-m-thick coarse

conglomeratic diamictite bed. A 14-event paragenetic sequence has been

documented and used as a temporal framework to interpret chemostratigraphic

trends in bulk rock d18O, d13C and 87Sr/86Sr. The uppermost surface of the

Barton Creek Formation was subaerially exposed before the KT impact, as

indicated by a brecciated palaeosol that caps upsection decreases in d13C and

d18O. Small 1-cm-diameter spheroids in the spheroid bed exhibit vermicular

crystalline textures but lack the concentric zonations common to accretionary

lapilli. These spheroids are hypothesized originally to have been impact glass

or reactive Ca and Mg oxide dusts that adhered to water vapour particles

condensing from the cooling impact vapour cloud. The spheroids were

dolomitized soon after deposition. The earliest dolomitization in the matrix

sediments of the Albion Formation was also post-depositional, replacing clays

formed by devitrification of impact glass. Dolomite and clay 87Sr/86Sr exhibit a

distinct symmetrical distribution in the spheroid bed ranging from 0Æ707745 to

0Æ707872. Although unproven, this may represent primary changes in the

chemical composition of the impact glass. The limestone clasts in the

diamictite bed were dolomitized before the KT impact and exhibit upsection

decreases in bulk rock 87Sr/86Sr. This suggests that the clasts were excavated

from strata equivalent in age or older than the Barton Creek Formation at

locations closer to, or in, the Chicxulub crater.

Keywords Carbonates, cathodoluminescence, Chicxulub, Cretaceous–Ter-tiary (KT) boundary event, ejecta, isotopes.

Sedimentology (2002) 49, 117–138

Ó 2002 International Association of Sedimentologists 117

INTRODUCTION

Evidence from drill cores and geophysical datahave demonstrated that the »200-km-diameterringed Chicxulub structure in the subsurface ofthe northern Yucatan in Mexico is the Creta-ceous–Tertiary (KT) large-body impact crater(Hildebrand et al., 1991, 1995; Pope et al., 1991,1994; Alvarez et al., 1992; Sharpton et al., 1992,1993, 1996; Swisher et al., 1992; Ward et al.,1995; Morgan & Warner, 1997; among others).Recognition of the impact site has been followedby an intensive search for circum-Caribbean andGulf of Mexico deposits of sedimentary ejecta thatrecord direct evidence of the transport dynamicsand chemistry of the impact vapour cloud andejecta curtain (Smit et al., 1992, 1996; Ocampoet al., 1996; Fourcade et al., 1998, 1998; Kelleret al., 1997; Bralower et al., 1998; Pope et al.,1999).

To date, the most proximal KT ejecta sedimentshave been recovered in the subsurface from deepexploratory wells drilled by Petroleo Mexicanos(PEMEX) and shallow wells drilled by the Uni-versidad Nacional Autonoma de Mexico (UNAM)in and near the deeply buried Chicxulub crater(Ward et al., 1995; Urrutia-Fucugauchi et al.,1996). Although of great scientific value, thecores from these wells have provided limitedinformation on the geographic distribution, sed-imentological composition and transport historyof KT ejecta. The most proximal and stratigraph-ically complete exposure of KT ejecta sedimentscropping out at the modern-day land surface is aquarry on Albion Island, Belize (Ocampo et al.,1996; Pope et al., 1999). The Albion Islandsection exposes the lowermost part of the Chic-xulub ejecta blanket at an intermediate distancebetween the crater rim and the siliciclastictsunami deposit sections (Smit et al., 1996).

The present study of the Albion Island sectionis the first to apply cathodoluminescence (CL)petrography and isotope geochemistry to recon-struct the depositional and post-depositionalalteration (diagenetic) history of sediments ejec-ted from the KT impact crater. Stratigraphic andpetrographic analyses of the ejecta have beensynthesized into a detailed paragenetic sequence,which provides a high-resolution relative timeframe of sediment deposition and diagenesis forthese deposits. Bulk-rock chemical analyses of87Sr/86Sr, d13O and d13C have been determinedand evaluated with respect to predicted water–rock interaction covariation trends. The temporalframework provided by the paragenetic sequence

integrated with the whole-rock isotopic trendshas been used to: (1) constrain the timing ofdolomitization of the ejecta with respect to the KTimpact; (2) determine that the Albion Island sitewas subaerially exposed before the impact; (3)interpret the formation of a unique class of ejectaproducts called dolomite spheroids; and (4)evaluate the source of coarse allochthonous clastsin the ejecta to test the applicability of currentmodels of impact sediment dispersal.

GEOLOGICAL SETTING

Albion Island (18°7Æ5¢ N, 88°42¢ W), which lies360 km SSE from the centre of the Chicxulubcrater (Fig. 1), is a site at which KT impact ejectawere deposited on the interior of the MesozoicYucatan carbonate platform. The base of theAlbion Island quarry is composed of 26 m ofBarton Creek Formation dolomitized limestone,interpreted to be of a Late Cretaceous andpossibly Maastrichtian depositional age (Ocampoet al., 1996; Vega et al., 1997; Pope et al., 1999;Fig. 2). Overlying the Barton Creek Formation is apervasively dolomitized ejecta deposit called theAlbion Formation, which is divided into: (1) alower 1-m-thick fine-grained clay and dolomitespheroid bed; and (2) an upper 15-m-thick coarsediamictite bed with abundant dolomitized lime-stone blocks as large as 6 m in diameter in amatrix of fine-grained dolomite (Ocampo et al.,1996; Pope et al., 1999; Fig. 2). The upper surfaceof the exposed diamictite bed is at the modernerosional land surface and, therefore, the originalthickness of the bed is unknown.

Ocampo et al. (1996) and Pope et al. (1999)presented detailed descriptions of the AlbionFormation lithologies, suggesting that at leastpart of the sedimentary debris is derived fromthe KT Chicxulub impact crater. Clay spheroidscomposed of palagonite have been interpreted tobe devitrified impact glass similar to the glasstecktites deposited in Haiti (Sigurdsson et al.,1991a,b), whereas the dolomite spheroids havebeen interpreted as altered impact-derived accre-tionary lapilli (Ocampo et al., 1996; Pope et al.,1999). Other evidence that the Albion Islandsection is an impact ejecta deposit includes itsproximity to the Chicxulub crater, consistent ifimprecise biostratigraphic age controls, coarseand poorly sorted sediments, one shocked quartzgrain, large accretionary blocks and polishedcobbles with striations and injected rock frag-ments (Ocampo et al., 1996; Pope et al., 1999).

118 B. W. Fouke et al.

Ó 2002 International Association of Sedimentologists, Sedimentology, 49, 117–138

METHODS

Centimetre-scale sampling of the lithologies crop-ping out in the Albion Island quarry (Table 1) wasclosely co-ordinated with sedimentologicaldescriptions presented by Ocampo et al. (1996)and Pope et al. (1999). Twenty spheroids collec-ted from the Albion Formation spheroid bed wereimpregnated, thin sectioned and analysed. Pol-ished thin sections and rock chips were examinedon a CITL 1200 cold cathodoluminoscope oper-ating at 11 kV and 550 lA in the Department ofGeology at the University of Illinois. An Optron-ics DEI-750 three-chip HCCD thermoelectronical-ly cooled camera was used to capture andmanipulate the low-light CL images directly(Fouke & Rakovan 2001). CL results are summar-ized in Table 2.

Sample powders for isotopic analysis wereprepared by crushing »1-cm3 rock chips in anHCl-cleaned porcelain mortar and pestle. Eachrock chip was scrubbed, ultrasonically cleaned in1% HCl, rinsed in deionized water and dried in adust-free low-temperature oven before crushing.Sample powders were separated into three splits,each of which was treated in the following

manner: (1) calcite–dolomite–clay sample split –unacidified without further treatment to retain allbulk-rock mineral components including calcite;(2) dolomite–clay sample split – leached in 4%glacial acetic acid for »1 min to remove calcite(according to techniques described by Staudtet al., 1993); and (3) clay sample split – leachedin 4% glacial acetic acid for »1 min followed by a24-h leach in 5% HCl to concentrate the clays andremove the dolomite mineral fraction. The sam-ple powder splits analysed geochemically in thisstudy are presented with their specific 87Sr/86Sr,d13C and d18O analyses in Table 3.

Carbon- and oxygen-isotope analyses weremeasured on the CO2 released during digestionof 20–50 lg of the sample powder in 100%phosphoric acid at 50 °C on a Finnigan-Mat massspectrometer in the Sawyer EnvironmentalLaboratory at the University of Maine. Data arereported as d13C and d18O values for CO2 gasrelative to VPDB using the standard delta notation(Swart et al., 1991). Analytical precision was0Æ1 ml–1 for oxygen and 0Æ2 ml–1 for carbon.Strontium separation was performed by standardchemical methods. This included dissolving 1 mgof sample powder in 3 M HCl at 110 °C for 10 h,centrifuging to remove organics, drying at 110 °C,acidification in 1Æ5 M HCl, transport throughcation exchange columns loaded with AminexQ15S and loading onto outgassed zone-refined Refilaments. 87Sr/86Sr was measured on a FinniganMAT 261 fixed multicollector thermal ionizationmass spectrometer in the Department of EarthSciences at Carleton University. Normalizationwas made to a 88Sr/86Sr ratio of 8Æ37521(86Sr/88Sr ¼ 0Æ1194). Sr isotope ratios from 10replicate analyses of NIST SRM 987 yielded amean value of 0Æ710251 (1 sigma ±0Æ000016).Average in-run precision was ±0Æ000015 for thestandard and ±0Æ000014 for each sample. Water–rock interaction modelling of the isotopes wascompleted by integrating equations described byBanner & Hanson, 1990) and Langmuir et al.(1978) into VISUAL BASIC macro programs embeddedwithin Microsoft 98 EXCEL (Fouke et al., 1996a,b).

SEDIMENTOLOGY AND CATHODO-LUMINESENCE PETROGRAPHY

Multiple complex episodes of carbonate sedimen-tation and diagenesis have created the depositsnow exposed in the Albion Island quarry. Thesedimentology and stratigraphy of the AlbionIsland deposits were thoroughly described by

Fig. 1. A map showing the location of Albion Island,Belize, with respect to the KT Chicxulub impact crateron the Yucatan peninsula of Mexico.

Petrography and isotope geochemistry of KT impact ejecta 119


Ocampo et al. (1996) and Pope et al. (1999).Therefore, only basic contextual summaries ofthe sedimentology and stratigraphy will be pre-sented here. CL petrography is an essential toolfor characterizing and fingerprinting individualevents of carbonate sedimentation and diagenesissuch as those comprising the Albion Formationejecta (Meyers, 1974; Meyers & Lohmann, 1985;Cander et al., 1988; Tucker & Wright, 1990; Foukeet al., 1996a,b; Montanez, 1997). The pore space-to outcrop-scale stratigraphic distribution ofcarbonate sediments and dolomite and calcitecrystals, mapped using their distinct colour andconcentric zonations under CL, has been used toestablish a high-resolution paragenetic sequence.This paragenetic sequence creates a relativetemporal framework from which the depositionaland diagenetic history of the KT impact ejecta hasbeen reconstructed. The paragenetic sequencehas also provided a framework with which tohelp constrain interpretation of the bulk-rockgeochemical analyses completed in this study.

The sedimentological and stratigraphic frame-work and CL petrography of each event in theparagenetic sequence are described and synthes-ized in the following descriptions of the BartonCreek and Albion Formations.

Barton Creek Formation

The basal portion of the Albion Island quarry iscomposed of pervasively dolomitized marinelimestones of the Barton Creek Formation(Fig. 2A and B, Table 1). Lithologies vary fromtidal flat carbonate mudstones containing gastro-pods, brachyuran crabs and anhydrite moulds tolagoonal cross-bedded fossiliferous packstones(Ocampo et al., 1996; Vega et al., 1997; Popeet al., 1999). Therefore, the Barton Creek Forma-tion at Albion Island represents Late Cretaceoussedimentation in restricted, shallow-water ma-rine environments well inboard from the marginof the extensive Yucatan platform. The top 20–50 cm of the Barton Creek Formation is an

Fig. 2. Field photographs of the Albion Island quarry. (A) and (B) Western wall of the quarry showing the completestratigraphic sequence of the Barton Creek Formation (BCF) and Albion Formation spheroid bed (SB) and diamictitebed (DB) with large dolomitized limestone clasts. (C) Close-up of the spheroid bed in (B) showing spheroids and alarge clay clast. (D) Close-up of the diamictite bed in (B), indicating a dolomitized limestone clast with a striatedsurface.



irregular red-tinted matrix-supported breccia(Figs 3D and 4), which has been interpreted torepresent an ancient soil (calcrete palaeosol or

caliche) formed during preimpact subaerial expo-sure and erosion of the Yucatan platform(Ocampo et al., 1996; Pope et al., 1999).

Table 1. Composition of the BartonCreek and Albion Formations in theAlbion Island quarry section(detailed descriptions are presentedin Pope et al., 1999).

Stratigraphic unit Thickness Sediment components

Diamictite bed 15 (+?) m Crystalline dolomite matrixDevitrified glass clay spheroidsAllochthonous limestoneblocks and clastsLimestone clasts with grooved

facetsFine-grained dolomite

spheroidsRadial fibrous calcite spheroids

Spheroid bed 1Æ2–2Æ0 m Fine-grained sucrosic dolomitematrix

Devitrified glass clay spheroidsFine-grained dolomite

spheroidsRadial fibrous calcite spheroidsShocked quartz

Barton CreekFormation

26 m Pervasively dolomitizedmarine limestones with thinred breccia at the top

Table 2. Description of each component of the paragenetic sequence in Fig. 4 comprising the Barton Creek andAlbion Formations.

Paragenetic Crystal Style of Crystal size

Cathodoluminescence

Stratigraphic

event morphology precipitation (mm) Colour Zonations occurrence

Dolomite 1 Large euhedralrhombs

Cement 150–300 ORB CZ, SZ BCF

Dolomite 2 Outer concentricrims

Cement 15–100 ORB CZ BCF

Dolomite 3 Rounded crystalsCCCR

Replacement 15–100 M CZ BCF

Dolomite 4 Small subhedralrhombs

Replacement 20–75 ORB CZ tomottled

BCF brecciatedtop

Dolomite 5 Small subhedralrhombs

Cement? 20–75 ORB CZ BCF brecciatedtop

Calcite 1 Large euhedralcolumnar

Cement 100–300 ORB CZ BCF

Dolomite 6 Euhedral Uncertain 15–40 ORB CZ tomottled

SB spheroids

Dolomite 7 Euhedral Replacement 50–75 ORB CZ tomottled

SB, DB

Dolomite 8 Rounded Uncertain 50–75 NL Not detected DBDolomite 9 Blocky Uncertain 40–100 ORB, M CZ DBDolomite 10 Concentric band Uncertain 10–15 ORB, M Mottled DBCalcite 2 Concentric band Uncertain 10–20 Y CZ DBCalcite 3 Concentric band Uncertain 5–15 M CZ DBCalcite 4 Blocky Uncertain 50–300 ORB, M CZ DB clasts

ORB, orange-red-brown; M, mauve; Y, yellow; NL, non-luminescent; CCCR, cloudy centre clear rim; CZ, concentriccrystal zonation; SZ, sectoral crystal zonation, BCF, Barton Creek Formation; SB, spheroid bed; DB, diamictite bed.



Table

3.

Geoch

em

ical

an

aly

ses

of

the

lith

olo

gie

scom

pri

sin

gth

eB

art

on

Cre

ek

Form

ati

on

(BC

F)

an

dth

esp

hero

id(S

B)

an

dd

iam

icti

te(D

B)

bed

sof

the

Alb

ion

Form

ati

on

.

Calc

ite–d

olo

mit

e–cla

yD

olo

mit

e–cla

yC

lay

Sam

ple

no.

(ele

vati

on

,m

)D

esc

rip

tion

d18O

(&V

PD

B)

d13C

(&V

PD

B)

d18O

(&V

PD

B)

d13C

(&V

PD

B)

87S

r/86S

r87S

r/86S

r

BC

F21Æ5

0D

olo

ston

e0Æ7

01Æ8

30Æ3

51Æ8

50Æ7

07722

ND

BC

F22Æ0

0D

olo

ston

e0Æ4

21Æ4

7N

DN

DN

DN

DB

CF

22Æ5

0D

olo

ston

e0Æ6

71Æ4

7N

DN

DN

DN

DB

CF

23Æ0

0D

olo

ston

e0Æ1

31Æ3

3N

DN

DN

DN

DB

CF

23Æ5

0D

olo

ston

e)

0Æ1

60Æ2

2N

DN

DN

DN

DB

CF

24Æ0

0D

olo

ston

e)

0Æ9

1)

0Æ2

7N

DN

DN

DN

DB

CF

24Æ5

0D

olo

ston

e)

1Æ8

9)

1Æ4

6)

1Æ6

9)

0Æ5

00Æ7

07776

ND

BC

F24Æ9

0D

olo

ston

e)

1Æ8

5)

1Æ4

1N

DN

DN

DN

DB

CF

25Æ5

0D

olo

ston

e)

1Æ6

9)

1Æ3

4N

DN

DN

DN

DB

CF

25Æ8

5D

olo

ston

e)

0Æ2

40Æ5

9N

DN

DN

DN

DB

CF

26Æ0

0T

hin

red

bre

ccia

)0Æ5

21Æ6

9)

0Æ4

01Æ7

10Æ7

07737

ND

SB

26Æ0

5C

last

san

dm

atr

ix)

0Æ2

2)

0Æ1

7)

0Æ1

50Æ5

10Æ7

07790

0Æ7

07850

SB

26Æ1

0C

last

san

dm

atr

ixN

DN

D)

0Æ0

9)

0Æ5

20Æ7

07808

0Æ7

07872

SB

26Æ1

5C

last

san

dm

atr

ix)

0Æ8

5)

1Æ1

8N

DN

DN

DN

DS

B26Æ1

8C

last

san

dm

atr

ix)

1Æ3

3)

2Æ2

3N

DN

DN

DN

DS

B26Æ2

0C

last

san

dm

atr

ix)

1Æ2

0)

2Æ0

2N

DN

DN

DN

DS

B26Æ4

5C

last

san

dm

atr

ix)

1Æ1

5)

2Æ4

7)

1Æ0

4)

1Æ7

80Æ7

07809

0Æ7

07837

SB

26Æ8

5C

last

san

dm

atr

ix)

0Æ9

1)

1Æ3

2)

1Æ3

1)

1Æ2

80Æ7

07783

0Æ7

07745

DB

27Æ0

0C

last

san

dm

atr

ix)

0Æ9

3)

0Æ2

4)

1Æ4

6)

0Æ5

10Æ7

07692

ND

DB

27Æ5

0C

last

san

dm

atr

ix)

0Æ4

9)

0Æ9

2N

DN

DN

DN

DD

B28Æ0

0C

last

san

dm

atr

ix)

1Æ5

1)

0Æ5

7N

DN

DN

DN

DD

B32Æ0

0C

last

san

dm

atr

ix)

0Æ8

0)

2Æ2

1)

0Æ7

9)

1Æ6

80Æ7

07838

ND

DB

32Æ2

5C

last

san

dm

atr

ix)

0Æ2

90Æ3

6)

0Æ4

60Æ5

00Æ7

07778

ND

DB

32Æ5

0C

last

)0Æ4

01Æ9

5N

DN

DN

DN

DD

B32Æ7

5C

last

)0Æ3

42Æ2

7N

DN

DN

DN

DD

B32Æ0

0C

last

)0Æ1

51Æ2

1)

0Æ5

30Æ8

00Æ7

07749

ND

DB

32Æ2

5C

last

)0Æ6

70Æ4

8N

DN

DN

DN

DD

B33Æ5

0C

last

)0Æ4

9)

0Æ0

9N

DN

DN

DN

DD

B33Æ7

5C

last

)0Æ4

30Æ8

9N

DN

DN

DN

DD

B34Æ0

0C

last

)1Æ0

70Æ3

1)

1Æ5

50Æ0

90Æ7

07747

ND

ND

,n

ot

dete

rmin

ed

.



Five generations of dolomite cementation(Dolomites 1–5; Figs 4, 5 and 6A and B; Table 2)and one event of calcite cementation (Calcite 1;Figs 4, 5 and 6C and D; Table 2) occur in theBarton Creek Formation limestone. The absenceof Dolomites 1–5 and Calcite 1 in overlyinglithologies (Fig. 4) suggests that these cementsprecipitated before deposition of the AlbionFormation (Fig. 5). Although it is possible forthe lateral movement of diagenetic groundwaterto create these types of stratiform cement distri-butions (Fouke et al., 1996a), no evidence wasobserved to support this interpretation for theBarton Creek Formation. The relative age of thecements (Fig. 5) is provided by the observationthat Dolomite 3 replaces Dolomites 1 and 2,disrupting their sharp euhedral concentric CLzonations (Figs 5 and 6A and B). Calcite 1 coats

and therefore post-dates Dolomites 4 and 5 withinthe uppermost 1 m of the Barton Creek Formation(Figs 5 and 6C and D). The angular, poorly sortedclasts in the red breccia at the top of the BartonCreek Formation range from 1 mm to 2 cm indiameter (Fig. 2D) and are composed of Dolomite4, whereas the dolomitized matrix of the brecciais composed of fine sucrosic Dolomite 5 crystals(Fig. 6E and F; Table 2). These pervasive replace-ment dolomitization textures indicate that Dolo-mites 4 and 5 were precipitated after the forma-tion of the thin red breccia at the top of the BartonCreek Formation (Fig. 5).

Albion Formation spheroid bed

The spheroid bed at the base of the AlbionFormation is a 0Æ9- to 1Æ5-m-thick layer of fine-

Fig. 3. (A) Sample number SB 26Æ20 from the spheroid bed exhibiting dolomite and clay spheroids. (B) Dolomitespheroids removed from the hand specimen shown in (A). (C) A cross-section through the centre of one of thedolomite spheroids in (B), illustrating that the spheroid lacks a nucleus and has no distinct concentric layering. (D)Sample number BCF 26Æ00 of the thin breccia capping the Barton Creek Formation.



Fig. 4. Stratigraphic column of the Barton Creek Formation and Albion Formation deposits, indicating the strati-graphic distribution of field sampling, geochemical analyses and sediment and diagenetic components.



grained dolomite and clay-rich sediments, themost abundant of which are spheroids and smallirregular clasts composed of dolomite and/or clay(Figs 2A–C and 3A–C; Table 1). X-ray diffractom-etry indicates that the clay is palagonite and hasbeen interpreted as devitrified impact glass(Ocampo et al., 1996; Pope et al., 1999). Severaltypes of spheroids have been observed within thespheroid bed that account for at least 30% of thelithology (Pope et al., 1999). Many of thesespheroids are concentrically zoned, reach diam-eters of up to 2 cm and are composed of dolomite,mixtures of dolomite and clay or pure clay(Ocampo et al., 1996; Pope et al., 1999). The 20spheroids analysed from the spheroid bed in thepresent study were »1 cm in diameter and spher-ical to ellipsoidal in shape with flattening parallelto the plane of bedding (Fig. 3A and B). Thesespheroids did not contain a nucleus, exhibited noconcentric layering (Fig. 3C) and were composedof small (£40 lm) euhedral crystals, here calledDolomite 6 (Fig. 7; Table 2).

Spheroids containing Dolomite 6 were notobserved in the Barton Creek Formation (Fig. 4).Therefore, Dolomite 6 has been interpreted topost-date Dolomite 5 (Fig. 5). Each spheroid ex-hibited a unique vermicular texture distributedalong parallel lineations that extend across eachspheroid (Fig. 7A–F). This texture cross-cuts theDolomite 6 crystal boundaries (Fig. 7E and F),suggesting that it was a primary depositionalfabric that was preserved during post-depositionaldolomitization. The fine-grained matrix support-

ing the spheroid bed is composed of finelydisseminated interstitial clays and 50–75 lmrhombohedral crystals of Dolomite 7 (Fig. 7Aand B; Table 2). Dolomite 7 crystals were com-monly observed floating within and cross-cuttingthe matrix clays (Fig. 7A and B). Therefore, theprecipitation of Dolomite 7 has been interpreted topost-date the diagenetic formation of the clay(Fig. 5). Dolomite 7 is slightly larger than Dolomite6 and exhibits brighter and more widely spacedconcentric CL zonations (Fig. 7A and B; Table 2).

Albion Formation diamictite bed

The next unit of the Albion Formation is thediamictite bed, a 15-m-thick layer of coarsematrix-supported conglomeratic breccia thatexhibits no obvious bedding or sorting (Fig. 2Aand B; Table 1; Ocampo et al., 1996; Pope et al.,1999). It is composed of a sucrosic dolomitizedmatrix that supports dolomitized limestone clastsfrom 1 mm to several metres in diameter (Fig. 2D)and spheroids composed of dolomite, calcite,and/or clay. The diamictite bed clasts commonlycontain marine fossils and represent a variety ofshallow-water, open to restricted marine carbon-ate facies (Ocampo et al., 1996; Pope et al., 1999).Many clasts exhibit variably oriented lineargrooves, some of which have polished surfaces(Ocampo et al., 1996; Pope et al., 1999; Fig. 2D. Inaddition, some blocks contain Barremian to Albianrudists, whereas others contain Albian to LowerCampanian foraminifera (Ocampo et al., 1996).

Fig. 5. Paragenetic sequence observed within lithologies of the Barton Creek Formation and the Albion Formation.Descriptions of each diagenetic component are presented in Table 2.



As in the spheroid bed, the fine-grained matrixsupporting the diamictite bed is composed ofDolomite 7 and interstitial clays (Fig. 8; Table 2).The spheroids in the diamictite bed are signifi-cantly less abundant than those in the spheroid

bed, lack the vermicular crystalline textures andare composed of Dolomite 7. Although no directevidence was observed to determine the relativetiming of precipitation of Dolomite 6 vs. 7, thelack of Dolomite 6 in the diamictite bed implies

Fig. 6. Paired plane light (left) and cathodoluminescence (right) photomicrographs of lithologies comprising theBarton Creek Formation. The relative timing and a description of each component are presented in the text (Table 2and Fig. 5). (A) and (B) Dolomite 1 (D1), Dolomite 2 (D2) and Dolomite 3 (D3) in sample number BCF 24Æ50. (C) and(D) Calcite 1 (C1) in sample number BCF 25Æ00. (E) and (F) Dolomite 4 (D4) and Dolomite 5 (D5) comprising thebreccia at the top of the formation in sample number BCF 26Æ00.



that Dolomite 6 is older (Fig. 5). Several otherdistinct types of dolomite and calcite crystals(Dolomites 8, 9 and 10; Calcites 2, 3 and 4) occur

within the large coarse clasts distributedthroughout the diamictite bed (Figs 4, 5 and 9;Table 2).

Fig. 7. Paired plane light (left) and cathodoluminescence (right) photomicrographs of lithologies comprising thespheroid bed. All photomicrographs are of sample number SB 26Æ20. The relative timing and a description of eachcomponent are presented in the text (Table 2 and Fig. 5). (A) and (B) Dolomite 6 (D6) comprising a spheroid (upperleft corner) and Dolomite 7 (D7) comprising the matrix and replacing clays that emit no CL. Note the vermiculartexture in the spheroid. (C) and (D) Thinner and less well-developed vermicular texture in a spheroid. (E) and (F)Higher magnification view of the vermicular structure in (A) and (B).



ISOTOPE GEOCHEMISTRY

Bulk-rock strontium, carbon and oxygen isotopicanalyses from the Barton Creek and AlbionFormations are presented in Table 3 and Fig. 10.

Documentation of whether the chemical analysiswas performed on a calcite–dolomite–clay, dolo-mite–clay or clay sample split is provided inTable 3. The composition of each sample splitcategory depended on the stratigraphic position

Fig. 8. Paired plane light (left) and cathodoluminescence (right) photomicrographs of lithologies comprising thematrix of the diamictite bed. The relative timing and a description of each component are presented in the text(Table 2 and Fig. 5). (A) and (B) Dolomite 7 (bright in CL) and Dolomite 8 (non-luminescent) in sample number DB27Æ50. (C) and (D) Dolomite 7 (D7) in sample number DB 28Æ00. Note dolomite replacing clays in the matrix. (E) and(F) Dolomite 7 (D7) and Dolomite 8 (D8) in sample number DB 28.00.



from which the samples were collected. For theBarton Creek Formation, the calcite–dolomite–clay sample splits contained Dolomites 1–5 andCalcite 1 (Fig. 4). For the spheroid bed, the

calcite–dolomite–clay sample splits containedvarying proportions of Dolomites 6 and 7 andclay (Fig. 4) and, for the diamictite bed, thecalcite–dolomite–clay sample splits contained

Fig. 9. Paired plane light (left) and cathodoluminescence (right) photomicrographs of lithologies comprising theclasts in the diamictite bed. The relative timing and a description of each component are presented in the text(Table 2, and Fig. 5). (A) and (B) Dolomite 8 (D8) in sample number DB 33Æ50. The irregular circular grain in theupper centre of the photomicrograph is the test of a foraminifera. (C) and (D) Dolomite 8, Dolomite 9 (D9), Calcite 2(C2) and Calcite 3 (C3) in sample number DB 33Æ75. Note Dolomite 7 replacing clays in the matrix. (E) and (F)Dolomite 8 and Calcite 4 in sample number DB 34Æ00.



mixtures of Dolomites 7–10 and Calcites 2–4(Fig. 4). Covariant trends in d13C (–2Æ47 to +2Æ27&

VPDB) and d18O (–1Æ89 to +0Æ70& VPDB) through-out the Albion Island section exhibit averagedifferences of »1& between the calcite–dolomite–clay and dolomite–clay sample splits (Fig. 10).Bulk-rock 87Sr/86Sr values also exhibit clearstratigraphic trends and range from 0Æ707692 to0Æ707808 (Fig. 10). The maximum difference be-tween the 87Sr/86Sr of the dolomite–clay and claysample splits was 0Æ00008 (Fig. 10).

Three important chemostratigraphic trendshave been observed in these bulk-rock chemicalanalyses. The first is a large synchronous co-varying decrease in d13C and d18O of –3& in thecalcite–dolomite–clay sample splits with up-ward stratigraphic progression in the BartonCreek Formation (21Æ5–26Æ0 m elevation;Fig. 10). The second is the symmetrical strati-graphic distribution of 87Sr/86Sr within thespheroid bed. Dolomite–clay sample split87Sr/86Sr increases by +0Æ000072 from the topof the Barton Creek Formation to the middle of

the spheroid bed, reaching a maximum of0Æ707809 (between 26Æ2 and 26Æ4 m elevation;Fig. 10). The dolomite–clay sample split87Sr/86Sr then decreases by –0Æ000117 to thebase of the overlying diamictite bed (Fig. 10).The clay sample split 87Sr/86Sr mirrors this trendbut at higher values, reaching a maximum of0Æ707872 (Fig. 10). These results suggest that thedolomite–clay sample split primarily reflects thedolomite composition, whereas the compositionof the clay split suggests that it is nearly a pureclay concentrate. A third distinct trend is exhib-ited by the bulk-rock chemistry of clasts in thediamictite bed (Fig. 10). The carbon and oxygenisotopes show variable trends near the bottomand top of the bed, with maximum fluctuationsof 3Æ9& d13C and 1Æ3& d18O (Fig. 10). Thedolomite–clay sample split 87Sr/86Sr increasesby 0Æ00145 from an elevation of 27 m at the baseof the diamictite bed to an elevation of 32 m(Fig. 10). A systematic decrease of 0Æ00091 thenoccurs vertically over the next 2 m of stratigra-phy (Fig. 10).

Fig. 10. Isotope chemostratigraphy of the Barton Creek and Albion Formations.



INTERPRETATION AND DISCUSSION

Integrating the results of the CL petrography andisotope geochemistry presented above permitsseveral valuable constraints to be placed on thedepositional and diagenetic history of the KTimpact ejecta deposited at Albion Island. Theseinclude: (1) dolomitization history of the ejecta;(2) reconstruction of the preimpact environmen-tal setting on the Yucatan platform; (3) prelimin-ary interpretation of the formation of the dolomitespheroids in the spheroid bed; and (4) evaluationof the mechanisms of sediment transport from theimpact crater.

Dolomitization of the Albion Island section

The petrography and geochemistry assembled inthe present study indicate that the Albion For-mation has experienced multiple events of pre-and post-impact dolomitization. Therefore,reconstruction of a depositional and diagenetichistory for these impact ejecta has requireddiscrimination between primary and secondarydolomitization. This has been done by combiningevidence of the stratigraphic distribution of eachdolomite in the Albion Island section (Fig. 4)with information on the timing of dolomitizationprovided by the paragenetic sequence (Fig. 5).

Dolomites 1–5 are stratigraphically restricted tothe Barton Creek Formation (Fig. 4), suggestingthat they formed before deposition of the AlbionFormation ejecta. Dolomite 6 occurs exclusivelywithin spheroids of the spheroid bed (Fig. 3) andexhibits a unique vermicular texture (Fig. 7).However, no petrographic evidence was observedto determine whether Dolomite 6 representsprimary dolomite precipitation during spheroidformation or is a result of post-impact dolomiti-zation. As described above, the stratigraphicdistributions of Dolomites 6 and 7 suggest thatDolomite 6 predated Dolomite 7 (Fig. 5). Rhombsof Dolomite 7 were observed floating within andthus replacing clays in the matrix of the spheroidbed (Fig. 7A and B, 8C and D). The clays areinterpreted to have formed by post-depositionaldiagenetic alteration of impact melt glass by localgroundwater (Ocampo et al., 1996; Pope et al.,1999). Therefore, the replacement of these clayssuggests that the precipitation of Dolomite 7 is apost-impact event that took place after depositionand impact glass diagenesis.

A prerequisite to interpreting mechanisms ofejecta dispersal is to determine whether the clastsin the diamictite bed were derived from near the

surface of the Yucatan platform (Barton CreekFormation) or from deeper subsurface horizonswithin the platform. The presence of Dolomites8–10 exclusively in clasts and not in the matrix ofthe diamictite bed implies that these dolomiteswere not formed during post-impact diagenesis.Therefore, Dolomites 8–10 presumably representpreimpact dolomitization of Yucatan platformlimestones. In addition, Dolomites 1–5 in theBarton Creek Formation (Fig. 6) are significantlydifferent in size, shape and CL character fromDolomites 8–10 in the clasts of the diamictite bed(Fig. 9; Table 2). This indicates that the diamic-tite bed clasts were not locally derived from theBarton Creek Formation at the Albion Island site.

Pre-impact environmental setting

The isotopic data collected in this study confirmprevious interpretations that the thin red brecciaat the top of the Barton Creek Formation (Fig. 4) isa palaeosol formed during subaerial exposure(Ocampo et al., 1996; Pope et al., 1999). Thissuggests that the KT ejecta sediments at AlbionIsland were deposited in a terrestrial rather thanin a shallow-marine setting. The stratigraphicrestriction of Dolomites 1–5 and Calcite 1 to theBarton Creek Formation (Fig. 4) and the isotopechemostratigraphy (Fig. 10) indicate that the topof the Yucatan platform was subaerially exposedbefore the deposition of the overlying AlbionFormation ejecta (Fig. 5). To evaluate the compo-sition of the groundwater responsible for thismeteoric diagenesis, the bulk-rock isotope datawere compared with an estimated sea-waterdolomite (ESD). The ESD was constructed frompreviously published d18O, d13C and 87Sr/86Srcompositions of marine low-Mg calcite cementsand shells precipitated from Campanian andMaastrichtian sea water (Jones et al., 1994;Howarth & McArthur, 1997; Vonhof & Smit,1997; McArthur et al., 1998; Veizer et al., 1999).The other hypothetical end-member in theseevaluations was an estimated freshwater dolomite(EFD), reconstructed from modern-day freshwateranalyses in the Yucatan and southern Caribbeanregion (Stoessel et al., 1989; Fouke, 1994). TheBarton Creek Formation isotopic data fall alongan inverted J-shaped distribution with respect tothe ESD (Fig. 11A), suggesting that the oxygenisotopes may have been diagenetically resetbefore the carbon isotopes. These covariationtrends, in combination with the progressivelylighter d18O and d13C values upsection (Fig. 10),are consistent with limestone diagenesis in fresh-



water containing dissolved soil-gas CO2 (Fig. 10;Allen & Matthews, 1982; Meyers & Lohmann,1985; Lohmann, 1987; Banner & Hanson, 1990;Goldstein et al., 1991; Fouke et al., 1996a,b).

A subset of the Barton Creek Formation d13Cand d18O analyses could also be interpreted as alinear distribution (Fig. 11A). If correct, this mayimply that there was: (1) physical mixing ofmeteoric groundwater with sea water that diage-netically altered the top of the formation; or (2)physical mixing of the Dolomite 1–5 and Calcite 1crystals in the bulk-rock sample powders (i.e.Faure, 1986). Although the data set is small, atleast some physical mixing is implied by the

correlation of the Barton Creek Formation datawith a binary mixing hyperbola calculated fromthe 87Sr/86Sr vs. d18O data (Fig. 11A). If accurate,this mixing parabola may imply that the uppersurface of the Barton Creek Formation at AlbionIsland was Campanian in age at the time of theimpact rather than Maastrichtian.

Genesis of the dolomite spheroidsin the Albion Formation spheroid bed

Concentrically layered spheroids with distinctnuclei have been observed in the Albion Forma-tion (Ocampo et al., 1996; Pope et al., 1999). These

Fig. 11. Geochemical cross-plots of bulk-rock 87Sr/86Sr, d13C and d18O compositions collected from the Barton CreekFormation (A) and the Albion Formation spheroid bed (B) and diamictite bed (C). Numerical data are presented inTable 3. Iterative water–rock interaction trajectories (WR) for varying meteoric water compositions are shown forfreshwater reaction with limestone using mixing equations in Langmuir et al. (1978) and water–rock interactionequations in Banner & Hanson (1990). Binary mixing lines (Mixing) were calculated using equations in Faure (1986).The composition of an estimated sea-water dolomite (ESD) precipitated from Campanian and Maastrichtian sea waterwas constructed from data in Jones et al. (1994), Howarth & McArthur (1997), Vonhof & Smit (1997), McArthur et al.(1998) and Veizer et al. (1999). Calcite–dolomite fractionation factors in Land (1980) were applied to estimate theESD d18O composition. The symbol key is the same as that shown in Fig. 10.



have been interpreted as being primary accretion-ary lapilli formed in the impact ejecta curtain ascondensing water vapour adhered fine dust tospinning grains in the impact debris clouds,forming lapilli similar to those created duringvolcanic eruptions (e.g. Graup, 1981; Schumacher& Schmincke, 1991). However, the dolomite sphe-roids from the spheroid bed analysed in thepresent study are distinct from accretionary lapil-li. Their lack of distinct nuclei and concentriclayering in cross section (Fig. 3C) may have beenthe result of fabric-destructive diagenesis (Popeet al., 1999). However, the vermicular crystallinetexture in these spheroids, which has not beendocumented previously in these or other KTimpact spheroids (Montanari, 1990), has beenpreserved (Fig. 7). This implies that the dolomitespheroids either: (1) originally lacked a nucleusand concentric layering; or (2) experienced dia-genesis that simultaneously preserved the vermic-ular fabric and destroyed the nuclei and layering.

Although no evidence was obtained conclu-sively to support or disprove the mechanism bywhich the dolomite spheroids were formed, atleast three hypothetical scenarios are plausible.The first is that the dolomite spheroids wereoriginally impact glass that was devitrified andaltered to clay. As is observed in thin section(Fig. 7A and B), these clays have commonly beenreplaced by dolomite. Therefore, each clay spher-oid may be a pervasively dolomitized clay spher-oid. The second hypothesis stems from theassumption that the cloud of vaporized rock(carbonate and silicate) and water created by thebolide impact cooled rapidly as it expanded awayfrom the Chicxulub crater. The condensation ofwater vapour in this cooling cloud may havetriggered the nucleation of calcium and magnes-ium oxides, forming a dust of highly reactive‘quicklime’ that accreted and rained out asspheroids (Ocampo et al., 1996). Where thesespheroids fell into sea water, they may have beendestroyed by rapid hydration, explaining thegeneral lack of large spheroids in marine sectionsof KT ejecta (Montanari, 1990). In contrast tosubaqueous environments, deposition at the ter-restrial Albion Island site may have permitted atleast short-term preservation of the spheroidsuntil percolating surface water led to pervasivediagenesis. Gradual hydration, conversion tovarious metastable calcium and magnesium com-pounds and the eventual formation of dolomitemay then have rapidly altered the spheroids. Ifcorrect, the vermicular texture may be a productof gas release during hydration of the spheroids.

An analogous series of chemical reactions isobserved during the formation of Portland ce-ment. Powdered carbonate and silicate rocks aremixed in varying proportions and heated to1400 °C (Erlin, 1969). Small proportions of waterare then added to drive the precipitation of belite(Ca2SiO4) in the form of 20- to 40-lm-diametersubspherical cement particles, called ‘clinkers’,containing parallel crystal striations (Petersen,1983a,b). It may be possible that the vermiculartexture is somehow analogous to these parallelcrystals. Larger spherical nodules are then formedas the clinkers are bound together by water viaboth coalescence and accretion processes. Relat-ively large proportions of water are required inthis process to form clinkers without outer rimscomposed of silica cements (Taylor, 1997), whichis consistent with the observations of the dolo-mite spheroids.

A third related hypothesis is that the dolomitespheroids formed by mechanisms similar to thosethat create lunar and meteoritic chondrules.Chondrules are millimetre-sized spherical tosubspherical bodies composed of silicate crystal-lites (‘parallel crystal barring’) that exhibit evi-dence of melting as well as the incorporation ofparticulate sulphides and Fe oxides (Roedder &Weiblen, 1977; Graup, 1981; Grossman & Wasson,1983). The parallel crystal barring may somehowbe analogous to the vermicular crystallization inthe spheroids, in that both crystalline fabricsextend across the host spherical grain. Two typesof impact-produced chondrules have been iden-tified, which include fluid drop chondrulescomposed of shock-melted silicates and lithicchondrules derived from rock fragments (Fre-dricksson et al., 1973a,b; Graup, 1981; Boss,1996). Secondary pervasive dolomitization maythen have replaced either of these types ofchondrules to form the dolomite spheroids.

Depositional history and 87Sr/86Sr chemostra-tigraphy of the Albion Formation

Synthesis of the sedimentological and geochem-ical analyses completed in this study permits anevaluation of the source and transport mecha-nisms of the Albion Formation ejecta. The87Sr/86Sr chemostratigraphy of the spheroid bedexhibits a remarkably symmetrical distribution(Fig. 10). If it can be demonstrated that the Sr inthese sediments has not been significantly alteredduring post-depositional diagenesis, then thistrend may reflect variations in the original com-position of the impact glass. The d18O vs. d13C



compositions suggest that the spheroid bed hasexperienced significant diagenetic alterationbased on the coincidence of the data withdiagenetic reaction trajectories (Fig. 11B). Oneinterpretation is that the spheroid bed sedimentswere altered via water–rock equilibration withfreshwater containing soil-derived CO2 (Fig. 10B;Allen & Matthews, 1982; Meyers & Lohmann,1985; Lohmann, 1987; Banner & Hanson, 1990;Goldstein et al., 1991; Fouke et al., 1996a,b).Alternatively, the subset of the spheroid bedd13C vs. d18O data that falls along a line (Fig. 11B)may suggest the binary mixing of sea water withfreshwater or the mixing of a sea water-deriveddolomite with a dolomite precipitated from fresh-water (Faure, 1986). Conversely, neither diagen-esis nor mixing is suggested by the 87Sr/86Sr vs.d13C and 87Sr/86Sr vs. d18O cross-plots (Fig. 11B).

Therefore, although unproven, this impliesthat, while the absolute values of 87Sr/86Sr inthe spheroid bed sediments have been partiallyaltered, their relative trends may still reflect aprimary change in the chemistry of the impactglass. If previous interpretations are correct andthe clay is derived from melt glass devitrification(Pope et al., 1999), then the higher 87Sr/86Sr of theclay sample split (Fig. 10) indicates that theoriginal impact glass contained more radiogenicSr than the dolomite. An associated possibility isthat the symmetrical 87Sr/86Sr trend could havebeen caused by larger proportions of clay todolomite in the middle of the spheroid bed, butno such trend was observed in thin section.

Black and yellow impact glasses derived frommelting of the siliceous basement at the bottom ofthe Chicxulub crater have been deposited in Haiti(Sigurdsson et al., 1991a; Blum & Chamberlain,1992; Blum et al., 1993). The black glass, inter-preted to be relatively pure crustal melt, has a87Sr/86Sr value of 0Æ70901 (Sigurdsson et al.,1991a). The yellow glass has been interpreted assiliceous melt mixed with carbonate sedimentsand contains a 87Sr/86Sr ratio of 0Æ70799 (Sig-urdsson et al., 1991a). The 87Sr/86Sr values of thespheroid bed dolomite–clay (0Æ707783–0Æ707809)and clay (0Æ707745–0Æ707872) sample splits arelower than those of the yellow glass 87Sr/86Sr(Fig. 10; Table 3). However, they are equivalent tohigher than the 87Sr/86Sr of the Upper Jurassic toUpper Cretaceous marine limestone that com-prised the impacted Yucatan platform (0Æ70685–0Æ70785; Howarth & McArthur, 1997). The inter-mediate 87Sr/86Sr composition of the spheroidbed may therefore imply that the original devit-rified glass in the ejecta was some type of yellow

glass created by mixing silicate melt with impac-ted carbonates.

As a result, it is possible that the symmetrical87Sr/86Sr stratigraphic trend in the spheroid bedrepresents complex mixing among Sr derivedfrom the impacted crustal rocks (limestone, evap-orites, silicate basement), the bolide and sea waterat the time of impact. Drill cores indicate that theshallowest depth of siliceous basement at thetime of the KT impact was at a depth of »1Æ7 kmbeneath the Yucatan platform (Ward et al., 1995).Therefore, the presence of impact glass in thespheroid bed requires that the Chicxulub craterwas excavated to at least this depth by the time ofspheroid bed deposition to provide a source ofsiliceous melt. As crater excavation proceededfrom the surface to the base of the Yucatanplatform, the liberated 87Sr/86Sr would have beenexpected to decrease progressively (Howarth &McArthur, 1997). This is not consistent with thesymmetrical distribution of 87Sr/86Sr in the spher-oid bed (Fig. 10), suggesting that this trend doesnot represent progressive crater excavation.Therefore, the implication of the 87Sr/86Sr chem-ostratigraphy in the spheroid bed with respect tocratering and ejecta deposition is uncertain.

An ejecta transport process permitting the fine-grained glass-rich sediments of the spheroid bedto outrun all other ballistic ejecta and be depos-ited as a thin undisturbed layer beneath thediamictite bed has not yet been identified. How-ever, two models for ejecta dispersal applicable tothe diamictite bed have been reconstructed forother impact sites. The first is the ballisticsedimentation model interpreted from the Buntesbreccia ejecta blanket from the Ries crater inGermany (Horz et al., 1983). In this model, largeballistically launched clasts and blocks wouldhave fallen back to earth near the crater, formingsecondary impacts that triggered ground surgedebris flows. Scouring of the earth surface duringthis process would have incorporated clasts fromthe Barton Creek Formation bedrock (Oberbeck,1975). Therefore, if this model is applicable, thediamictite bed debris clasts may be composed of amixture of primary clasts from the crater site andregionally to locally derived clasts eroded fromthe Barton Creek Formation. The second ejectasedimentation model is the ring vortex mechan-ism for continuous ejecta emplacement (Schultz,1992; Barnouin-Jha & Schultz, 1996, 1998). In thisscenario, the trajectories of ejecta expelled fromthe impact crater initially form a curtain in theshape of an inverted hollow cone that expandsoutwards. The ejecta curtain acts as a barrier that



forces the atmosphere away from it, creating ringvortices. Atmospheric drag acts to decelerate theejecta within the vortex and reduces the velocityof the ejecta curtain, causing the inverted hollowcone to collapse into a turbulent flow of debristhat is deposited several crater diameters from thecrater rim. If applicable, the diamictite bedshould be dominantly composed of clasts derivedfrom the crater rather than from the Barton CreekFormation.

The evidence collected in the present study isinconclusive regarding whether ballistic sedi-mentation or ring vortex collapse more accuratelydescribes ejecta dispersal from the Chicxulubcrater. Ocampo et al. (1996) and Pope et al. (1999)interpreted the majority of diamictite bed clasts tobe primary ejecta and thus suggested that the ringvortex model is most applicable. Their interpret-ation was based on: (1) significant differences insedimentological composition observed betweenthe Barton Creek Formation limestone and theclasts in the diamictite bed; and (2) rudistbiostratigraphy, suggesting that some of the dia-mictite bed clasts may be Lower Cretaceous.However, significant variations in sedimentolog-ical and chemical composition of the limestonewould be expected across the broad expanse ofthe Yucatan platform. Therefore, the differencesin dolomite petrography and chemistry may be aproduct of lateral variations rather than an indi-cation of deep crater excavation. In addition, thebulk rock 87Sr/86Sr from the diamictite bed clastsexhibit consistent upsection decreases (Fig. 10).Therefore, although conjectural at this point, thischemostratigraphic trend in clast 87Sr/86Sr isconsistent with increasing depth of excavationwithin the Yucatan platform. This may imply thatat least some of the clasts were primary ejectaderived during cratering from subsurface hori-zons in the Yucatan platform.

CONCLUSIONS

The depositional and diagenetic history of KTimpact ejecta deposited 360 km from the Chicxu-lub crater at Albion Island, Belize, has beenevaluated with CL petrography and isotope geo-chemistry. The base of the section, exposed in anactive quarry on Albion Island, consists of LateCretaceous Barton Creek Formation marine lime-stone deposited on the Yucatan carbonate plat-form before the impact. The overlying 16 m ofsediment are KT impact ejecta deposits called theAlbion Formation. This unit is composed of an

»1-m-thick spheroid bed of fine-grained dolomiteand clay-rich sediments with small spheroidalpebbles, and a 15-m-thick diamictite bed com-posed of a coarse conglomeratic breccia withclasts up to 7Æ5 m in diameter. A parageneticsequence of 14 depositional and diagenetic eventshas been documented in the Albion Islandsection, which has been used as a contextualframework to interpret bulk-rock d18O, d13C and87Sr/86Sr analyses.

A thin red dolomitized breccia at the upper-most surface of the Barton Creek Formation capsupsection decreases in bulk-rock d13C and d18O,indicating that the Albion Island site was subae-rially exposed and pervasively dolomitizedbefore the KT impact. Abundant 1-cm-diameterdolomite spheroids in the spheroid bed exhibitunique vermicular crystalline textures and lackthe concentric zonations common to other accre-tionary lapilli in the Albion Formation. Thesedolomite spheroids are hypothesized originallyto have been impact glass, or Ca and Mg oxidedusts that adhered to condensing water particlesin the impact vapour cloud and underwent rapidhydration and dolomitization after deposition.The earliest precipitation of dolomite observed inthe matrix of the Albion Formation post-datedclays formed by devitrification of impact glass.This indicates that the dolomites comprising thematrix of the Albion Formation are also productsof post-depositional diagenesis. Bulk-rock87Sr/86Sr in the spheroid bed exhibit a distinctsymmetrical stratigraphic distribution. Compar-ison with modelled water–rock interaction trendssuggests that, although the spheroid bed hasexperienced extensive diagenetic alteration, var-iations in 87Sr/86Sr may reflect primary changesin the composition of the original impact glass.The petrography and bulk-rock 87Sr/86Sr of largeclasts in the diamictite bed indicate that theywere not locally derived from the Barton CreekFormation limestone. However, they may havebeen eroded from equivalent or older limestonehorizons in the Yucatan platform closer to, orwithin, the Chicxulub crater. The results indicatethat both the ballistic and the ring vortex modelsremain viable models for the interpretation ofejecta dispersal from the Chicxulub impactcrater.

ACKNOWLEDGEMENTS

NSF Geology and Paleontology Program AwardEAR-9909560 and ACS–Petroleum Research



Fund Award 34549-G2 to B. Fouke supported thisresearch, as well as funding from the NASAExobiology Program and The Planetary Society toK. Pope and A. Ocampo. Isotopic analyses byJ. Blenkinsop at Carleton University and C.Pedrone at the University of Maine are gratefullyacknowledged. Editorial reviews by J. McArthur,P. Wignall and I. Jarvis served to improve themanuscript significantly.

REFERENCES

Allen, J.R. and Matthews, R.K. (1982) Isotope signatures as-

sociated with early meteoric diagenesis. Sedimentology, 29,797–817.

Alvarez, W., Smit, J., Lowrie, W., Asaro, F., Margolis, S.V.,Kastner, M. and Hildebrand, A.R. (1992) Proximal impact

deposits at the Cretaceous–Tertiary boundary in the Gulf of

Mexico: a restudy of DSDP Leg 77 sites 536 and 540. Geol-ogy, 20, 697–700.

Banner, J.L. and Hanson, G.N. (1990) Calculation of simulta-

neous isotopic and trace element variations during water–

rock interaction with application to carbonate diagenesis.Geochim. Cosmochim. Acta, 54, 3123–3127.

Barnouin-Jha, O.S. and Schultz, P.H. (1996) Ejecta entrain-

ments by impact-generated ring vortices: theory and

experiments. J. Geophys. Res., 101, 21099–21115.Barnouin-Jha, O.S. and Schultz, P.H. (1998) Lobateness of

impact ejecta deposits from atmospheric interactions.

J. Geophys. Res., 103, 25739–25756.Blum, J.D. and Chamberlain, C.P. (1992) Oxygen isotope

constraints on the origin of impact glasses from the Creta-

ceous–Tertiary boundary. Science, 257, 1104–1107.

Blum, J.D., Chamberlain, C.P., Hingston, M.P., Koeberl, C.,Marin, L.E., Schuraytz, B.C. and Sharpton, V.L. (1993)

Isotopic comparison of the KT boundary impact glass with

melt rock from the Chicxulub and Mason impact structures.

Nature, 364, 325–327.Boss, A.P. (1996) A concise guide to chondrule formation

models. In: Chondrules and the Protoplanetary Disk (Eds

R.H. Hewins, R.H. Jones and E.R.D. Scott), pp. 257–264.University Press, Cambridge.

Bralower, T.J., Paull, C.K. and Leckie, R.M. (1998) The Cre-

taceous/Tertiary boundary cocktail: Chicxulub impact trig-

gers margin collapse and extensive sediment gravity flows.Geology, 26, 331–334.

Cander, H.S., Kaufman, J., Daniels, L.D. and Meyers, W.J.(1988) Regional dolomitization of shelf carbonates in the

Burlington–Keokuk Formation (Mississippian), Illinois andMissouri. In: Constraints from Cathodoluminescent ZonalStratigraphy. Sedimentology and Geochemistry of Dolo-stones (Eds V. Shukla and P.A. Baker), SEPM Spec. Publ.,43, 129–144.

Erlin, B. (1969) Portland cement and concrete. Rev. Eng. Geol.,2, 81–103.

Faure, G. (1986) Principles of Isotope Geology. John Wiley andSons, New York, 589 pp.

Fouke, B.W. (1994) Deposition, Diagenesis and Dolomitizationof Neogene Seroe Domi Formation Coral Reef Limestones onCuracao, Netherlands Antilles. Publications Foundation forScientific Research in the Caribbean Region, The Hague,

197 pp.

Fouke, B.W. and Rakovan, J. (2001) An integrated cathodo-

luminescence video-capture microsampling system. J. Sed.Res., 71, 509–513.

Fouke, B.W., Meyers, W.J., Hanson, G.N. and Beets, C.J.(1996a) Chronostratigraphy and dolomitization of the Seroe

Domi Formation, Curacao, Netherlands Antilles. Facies, 35,293–320.

Fouke, B.W., Everts, A.J.W., Zwart, E.W., Schlager, W.,Smalley, P.C. and Weissert, H. (1996b) Subaerial exposure

unconformities on the Vercors carbonate platform (SE

France) and their sequence stratigraphic significance. In:High Resolution Sequence Stratigraphy: Innovations andApplications (Eds J.A. Howell and J.F. Aitken), Geol. Soc.London Spec. Publ., 104, 295–320.

Fourcade, E., Rocchia, R., Gardin, S., Bellier, J.P., Debrabant,P., Masure, E., Robin, E. and Pop, W.T. (1998) Age of the

Guatemala breccias around the Cretaceous–Tertiary

boundary; relationships with the asteroid impact on the

Yucatan. Sci. Terre Planet., 327, 47–53.Fredricksson, K., Dube, A., Milton, D.J. and Balasundaram,

M.S. (1973a) Lonar Lake, India: an impact crater in basalt.

Science, 180, 862–864.Fredricksson, K., Noonen, A. and Nelen, J. (1973b) Meteoric,

lunar and Lonar impact chondrules. Moon, 7, 475.

Goldstein, R.H., Anderson, J.E. and Bowman, M.W. (1991)

Diagenetic responses to sea-level change: integration offield, stable isotope, paleosol, paleokarst, fluid inclusion,

and cement stratigraphy research to determine history and

magnitude of sea-level fluctuation. In: Sedimentary Model-ing (Eds E.K. Franseen, W.L. Watney and C.G.S.C. Kendall),Kansas Geol. Survey Bull., 223, 139–162.

Graup, G. (1981) Terrestrial chondrules, glass spherules and

accretionary lapilli from the suevite, Ries Crater, Germany.Earth Planet. Sci. Lett., 55, 407–418.

Grossman, J.N. and Wasson, J.T. (1983) The compositions of

chondrules in unequilibrated chondrites: an evaluation of

models for the formation of chondrules and their precursormaterials. In: Chondrules and Their Origins (Ed. E.A. King),

pp. 88–121. Lunar and Planetary Institute, Houston.

Hildebrand, A.R., Penfield, G.T., Kring, D.A., Pilkington, M.,Camargo, A., Jacobsen, S.B. and Boynton, W.V. (1991)Chicxulub crater: a possible Cretaceous/Tertiary boundary

impact crater on the Yucatan Peninsula, Mexico. Geology,

19, 867–871.

Hildebrand, A.R., Pilkington, M., Connors, M., Ortiz-Aleman,C. and Chavez, R.E. (1995) Size and structure of the Chic-

xulub Crater revealed by horizontal gravity gradients and

cenotes. Nature, 376, 415–417.Horz, F., Ostertag, R. and Rainey, D.A. (1983) Bunte Breccia of

the Ries: continuous deposits of large impact craters. Rev.Geophys. Space Phys., 21, 1667–1725.

Howarth, R.J. and McArthur, J.M. (1997) Statistics for stron-tium-isotope stratigraphy: a robust LOWESS fit to the

marine Sr-isotope curve for 0–206 Ma. J. Geol., 4, 441–456.

Jones, C.E., Jenkyns, H.C., Coe, A.L. and Hasselbo, S.P. (1994)

Strontium isotope variations in Jurassic and Cretaceousseawater. Geochim. Cosmochim. Acta, 58, 3061–3074.

Keller, G., Lopez-Oliva, J.G., Stinnesbeck, W. and Adatte, T.(1997) Age, stratigraphy, and deposition of near-K/T silici-clastic deposits in Mexico: relation to bolide impact? Geol.Soc. Am. Bull., 109, 410–428.

Land, L.S. (1980) The isotopic and trace element geochemistry

of dolomite: the state of the art. In: Concepts and Modelsof Dolomitization (Eds D.H. Zenger, J.B. Dunham and

R.L. Ethington), SEPM Spec. Publ., 28, 87–110.



Langmuir, C.H., Vocke, R.D. and Hanson, G.N. (1978) A

general mixing equation with applications to Icelandic

basalts. Earth Planet. Sci. Lett., 37, 380–392.

Lohmann, K.C. (1987) Geochemical patterns of meteoric dia-genetic systems and their application to studies of paleok-

arst. In: Paleokarst (Eds P.W. Choquette and N.P. James),

pp. 10–42. Springer, Berlin.

McArthur, J.M., Thirwall, M.F., Engkilde, M., Zinsmeister,W.J. and Howarth, R.J. (1998) Strontium isotope profiles

across the K/T boundary sequences in Denmark and

Antarctica. Earth Planet. Sci. Lett., 160, 179–192.Meyers, W.J. (1974) Carbonate cement stratigraphy of the Lake

Valley Formation. (Mississippian), Sacramento Mountains,

New Mexico. J. Sed. Petrol., 44, 837–861.

Meyers, W.J. and Lohmann, K.C. (1985) Isotope geochemistryof regionally extensive calcite cement zones and marine

components in Mississippian limestones. In: CarbonateCements (Eds N. Schneiderman and P.M. Harris), SEPMSpec. Publ., 36, 223–239.

Montanari, A. (1990) Authigenesis of impact spheroids in the

K/T boundary clay from Italy: new constraints for high-

resolution stratigraphy of terminal Cretaceous events.J. Sed. Petrol., 61, 315–339.

Montanez, I.P. (1997) Secondary porosity and late diagenetic

cements of the Upper Knox Group, Central Tennessee re-

gion: a temporal and spatial history of fluid flow conduitdevelopment within the Knox regional aquifer. In: Basin-Wide Diagenetic Patterns (Eds I.P. Montanez, J.M. Gregg and

K.L. Shelton), SEPM Spec. Publ., 57, 101–118.

Morgan, J. and Warner, M. (1997) Size and morphology of theChicxulub impact crater. Nature, 390, 472–476.

Oberbeck, V.R. (1975) The role of ballistic erosion and sedi-

mentation in lunar stratigraphy. Geophys. Space Phys., 13,337–362.

Ocampo, A., Pope, K.O. and Fischer, A.G. (1996) Ejecta

blanket deposits of the Chicxulub Crater from Albion Island,

Belize. In: The Cretaceous–Tertiary Event and OtherCatastrophes in Earth History (Eds G. Ryder, D. Fastovsky

and S. Gartner), Geol. Soc. Am. Special Paper, 307, 75–88.

Petersen, I.F. (1983a) Isothermal sintering of Portland cement

raw mixes – Part 1. World Cement, 14, 190–195.Petersen, I.F. (1983b) Isothermal sintering of Portland cement

raw mixes – Part 2. World Cement, 14, 220–228.

Pope, K.O., Ocampo, A.C. and Duller, C.E. (1991) Mexican site

for the K/T impact crater? Nature, 351, 105.Pope, K.O., Baines, K.H., Ocampo, A.C. and Ivanov, B.A.

(1994) Impact winter and the. Cretaceous/Tertiary extinc-

tions; results of a Chicxulub asteroid impact model. EarthPlanet. Sci. Lett., 128, 719–725.

Pope, K.O., Ocampo, A.C., Fischer, A.G., Alvarez, W., Fouke,B.W., Webster, C.L., Vega, F.J., Smit, J., Fritsche, A.E. and

Claeys, P. (1999) Chicxulub impact ejecta from AlbionIsland, Belize. Earth Planet. Sci. Lett., 170, 351–364.

Roedder, E. and Weiblen, P.W. (1977) Barred olivine ‘chond-

rules’ in lunar spinel troctolite 62295. Proc. Lunar Sci.Conf., 8, 2641–2654.

Schultz, P.H. (1992) Atmospheric effects on ejecta emplace-

ment. J. Geophys. Res., 97, 11623–11662.

Schumacher, R. and Schmincke, H.U. (1991) Internal struc-ture and occurrence of accretionary lapilli – A case study at

Laacher See Volcano. Bull. Volcanol., 53, 612–623.

Sharpton, V.L., Dalrymple, G.B., Marin, L.E., Ryder, G.,Schuraytz, B.C. and Urrutia-Fucugauchi, J. (1992) Newlinks between the Chicxulub impact structure and the

Cretaceous/Tertiary boundary. Nature, 359, 819–821.

Sharpton, V.L., Burke, K., Camargo-Zanoguera, A., Hall, S.A.,Lee, D.S., Marin, L.E., Suarez-Reynoso, G., Quezada-Muneton, J.M., Spudis, P.D. and Urrutia-Fucugauchi, J.(1993) Chicxulub multiring impact basin: Size and othercharacteristics derived from gravity analysis. Science, 261,1564–1567.

Sharpton, V.L., Marin, L.E., Carney, J.L., Lee, S., Ryder, G.,Schuraytz, B.C., Sikura, P. and Spudis, P.D. (1996) A modelof the Chicxulub impact basin based on evaluation of geo-

physical data, well logs, and drill core samples. In: TheCretaceous–Tertiary Event and Other Catastrophes in EarthHistory (Eds G. Ryder, D. Fastovsky and S. Gartner), Geol.Soc. Am. Special Paper, 307, 55–74.

Sigurdsson, H., D’Hondt, S., Arthur, M.A., Bralower, T.J.,Zachos, J.C., van Fossen, M. and Channell, J.E.T. (1991a)Glass from the Cretaceous/Tertiary boundary in Haiti.

Nature, 349, 482–487.

Sigurdsson, H., Bronte, P., Turpin, L., Chaussidon, M.,Metrich, N., Steinberg, M., Pradel, P. and Hondt, D.S.(1991b) Geochemical constraints on source region of Creta-

ceous/Tertiary impact glasses. Nature, 353, 839–842.

Smit, J., Montanari, A., Swinburne, N.H.M., Alvarez, W.,Hildebrand, A.R., Margolis, S.V., Claeys, P., Lowrie, W.and Asaro, F. (1992) Tektite-bearing, deep water clastic

unit at the Cretaceous–Tertiary boundary in northeastern

Mexico. Geology, 20, 99–103.Smit, J., Roep, T.B., Alvarez, W., Montanari, A., Claeys, P.,

Grajalez-Nishimura, J.M. and Bermudez, J. (1996) Coarse-

grained, clastic sandstone complex at the K/T boundary

around the Gulf of Mexico: deposition by tsunami wavesinduced by the Chicxulub impact?. In: The Cretaceous–Tertiary Event and Other Catastrophes in Earth History (Eds

G. Ryder, D. Fastovsky and S. Gartner), Geol. Soc. Am. Spec.Paper, 307, 151–182.

Staudt, W., Oswald, E.J. and Schoonen, M.A.A. (1993)

Determination of sodium, chloride and sulfate in dolomites:

a new technique to constrain the composition of dolomi-tizing fluids. Chem. Geol., 107, 97–109.

Stoessell, R.K., Ward, W.C., Ford, B.H. and Schuffert, J.D.(1989) Water chemistry and CaCO3 dissolution in the saline

part of an open-flow mixing zone, coastal Yucatan Penin-sula, Mexico. Geol. Soc. Am. Bull., 101, 159–169.

Swart, P.K., Burns, S.J. and Leder, J.J. (1991) Fractionation of

the stable isotopes of oxygen and carbon dioxide during the

reaction of calcite with phosphoric acid as a function oftemperature and technique. Chem. Geol., 86, 89–96.

Swisher, C.C.I., Grajales-Nishimura, J.M., Montanari, A.,Margolis, S.V., Claeys, P., Alvarez, W., Renne, P., Cedillo-Pardo, E., Maurrasse, F.J.-M.R., Curtis, G.H., Smit, J. and

McWilliams, M.O. (1992) Coeval 40Ar/39Ar ages of 65.0

million years ago from Chicxulub crater melt rock and

Cretaceous/Tertiary boundary tektites. Science, 257, 954–958.

Taylor, H.F.W. (1997) Cement Chemistry. Thomas Telford

Publishing, London, 597 pp.

Tucker, M.E. and Wright, V.P. (1990) Carbonate Sedimento-logy. Blackwell Science, Cambridge, 482 pp.

Urrutia-Fucugauchi, J., Marin, L. and Trejo, G.A. (1996)

UNAM scientific drilling program of the Chicxulub impactstructure: evidence for a 300 kilometer crater diameter.

Geophys. Res. Lett., 23, 1565–1568.

Vega, F.J., Feldmann, R.M., Ocampo, A.C. and Pope, K.E.(1997) A new species of Late Cretaceous crab (Brachyura:Carcineretidae) from Albion Island, Belize. J. Paleontol., 71,615–620.



Veizer, J., Ala, D., Azmy, K., Bruckshen, P., Buhl, D., Bruhn,F., Carden, G.A.F., Diener, A., Ebneth, S., Godderis, Y.,Jasper, T., Korte, C., Pawellek, F., Podlaha, O.G. and

Strauss, H. (1999) 87Sr/86Sr, d13C and d18O evolution ofPhanerozoic seawater. Chem. Geol., 161, 59–88.

Vonhof, H.B. and Smit, J. (1997) High-resolution late Maas-

trichtian early Danian oceanic 87Sr/86Sr record: implica-

tions for Cretaceous–Tertiary boundary events. Geology, 25,347–350.

Ward, W.C., Keller, G., Stinnesbeck, W. and Adatte, T. (1995)

Yucatan subsurface stratigraphy: implications and con-

straints for the Chicxulub impact. Geology, 23, 873–876.

Manuscript received 14 July 2000;revision accepted 19 September 2001.



cathodoluminescence petrography and isotope geochemistry

Documents