Table Of Content

KTEVENTS IN INDIA: IMPACT, RIFTING, VOLCANISM AND DINOSAUR EXTINCTION SANKARCHATTERJEE&DHIRAJ K. RUDRA Chatterjee,S.&Rudra,D.K. 1996 1220: KTeventsinIndia: impact,rifting,volcanismand dinosaurexticntion. Memoirsofthe QueenslandMuseum 39(3): 489-532. Brisbane. ISSN 0079-8835. Formorethanadecade,anumberofimpactsiteshavebeenlinkedtothemassextinctionat theKT(Cretaceous/Tertiary)boundary.Theprimecandidatetoday isthe—ChicxulubCrater —inYucauinPeninsula,Mexico.RecentlyanotherpotentialKTimpactscar theShivaCrater hasbeenidentified fromsubsurfacedataattheIndia-Seychelles riftmargin. Thecrucial evidence in support of this impact structure comes from the Bombay High field, a giant offshoreoil basin in India, andassociated alkaline intrusiveswithin the DeccanTraps. The KTboundary ageofthecraterisinferred fromitsDeccan lava floor, Palaeocene ageofthe overlying sediments, isotope dating (~65Ma) of presumed melt rocks, and the Carlsberg riftingevent(chron29R)withinthebasin.SeismicreflectiondataandIndia-Seychellesplate reconstruction at65Ma reveal aburiedoblongcrater, 600km long, 450km wideand 12km deep, carved through DeccanTrapsand intounderlying Precambrian granite. Itrepresents the largest impact structure of Phanerozoic age. The crater shows the morphology of a complex impact structure and basin, with adistinct central uplift in the form ofa series of peaks, an annular trough and a slumped rim. The oblong shape of the crater and the asymmetric distribution offluidejecta indicateoblique impact in a SW-NEtrajectory. We speculate thata40km diametermeteoritecrashed on the western continental shelfofIndia around 65Ma, excavating the Shiva Crater, shattering the lithosphere and inducing the India-Seychellesrifting.Thecraterappearstonarrow in the formofateardroptotheNEor downrangewheretheejectameltrockswereemplacedradiallyoutwardbytheimpactshock. TheshapeoftheShivaCraterandtheasymmetricejectadistributionmimicthoseofartificial cratersproducedbyobliqueimpactsinlaboratoryexperiments.Thesynchronyandnear-antipodal positionsoftheShivaandChicxulubCratersmay indicatetwoalternativemodesof theirorigin. Either,both cratersoriginatedfromsplittingofalargerdiametermeteorite,or, large impacton one side ofthe Earth produced a similar signature on the far sideby axial focusing of seismic waves. Since India was ground zero for both an impact and Deccan volcanism,theircausalrelationshipsandbioticeffectswereassessed.ItappearsthatDeccan volcanismbegan 1Mabeforethe KTeventandwasnottriggeredbytheimpact. Itsorigin is attributedtotheDeccan-Reunionhotspot.TheextensivearealdistributionofDeccanTraps isowingtointercanyonflowsalongthedrainageoftheNarmada,GodavariandtheCambay basins. During the early stage ofDeccan eruption, sauropods, theropods and ankylosaurs flourished in India, but they died out suddenly at the KTimpactboundary. Although both impact and Deccan volcanismare hypothesized as contributing to the deleterious environ- mentalconsequencesleadingtobioticcrisisatthe KTboundary,theimpactissuggestedas having played the major role as the killing mechanism. Impact, Cretaceous-Tertiary boundary, India, dinosaur, extinction, volcanism. S. Chatterjee, Museum ofTexas Tech University, Lubbock, Texas 79409-3191, USA; D.K. Rudra, Geology Unit, Indian Statistical Institute. 203 Barrackpore Trunk Road, Calcutta 700035, IndianMay 1996. Sinceitsemergenceandsubsequenteruptionof diedoutfairlyrapidly. Allmassextinctions,how- diversity, life has shown a tenacious and wildly ever, have been followed by at least a partial successful holdonthisplanet.Buttherichhistory evolutionary recovery in which the number of oflifehas been repeatedly punctuatedbyequally species on Earth has increased again, awesome displays of its destruction. It is es- There are five major episodes of mass extinc- timated that 99% of plant and animal life that tions during the past 600 million years: Late haveeverlivedonEarth arenowextinct(Wilson, Ordovician (440Ma), Late Devonian (365Ma), 1992). The history of life is replete with major Late Permian (245Ma), Late Triassic (210Ma) episodes of biotic catastrophes or mass extinc- and Late Cretaceous (65Ma). Ofthese mass ex- tions, when 50%ormoreofthe unrelated species tinctions, the one that has captured the greatest 490 MEMOIRS OFTHEQUEENSLANDMUSEUM TOTALEXTINCTION >#r«. (Nanavtan Dinosaurs! [LargeMarineReptiles) FIG. 1. VictimsandsurvivorsaftertheKTextinction.Theprimaryvictimswerenonaviandinosaurs,pterosaurs, largemarinereptilessuchasplesiosaursandmosasaurs,andvariousinvertebratessuchasammonitesandrudists. Lizards,snakes,turtles,crocodiles,birdsandmammalsenduredthiscatastropheandrebounded. Bothbirdsand mammals underwentexplosiveevolutionsafterthiscrisis. attention of earth scientists has been the KT Bytheendofthe Cretaceous, harshchanges of (Cretaceous/Tertiary) extinction when the environments were taking place as a result of dinosaurs and two-thirds of all marine animal plate movements, mountain buildings, volcanic speiceswerewipedout.Thesuddenextinctionof emissions and sea regressions. Exactly what dinosaurs has puzzled both scientists and public causedthebioticcrisisremainshighlycontrover- formorethan acentury. Having survived for 160 sial. Currently two competing models have been million years, dinosaurs seemed indestructible. proposed to explain this apocalyptic disaster at Not only the dinosaurs died out during that rela- the KT boundary: meteorite impact hypothesis tively brief period; all land animals weighing and volcanic hypothesis. The impact theory pos- more than 25kg disappeared fromthe planet. All tulatesthattheenvironmentswerelethallyaltered pterosaurs, plesiosaurs, mosasaurs as well as or destroyed at the end ofthe Cretaceous by the severalfamiliesofbirdsandmarsupialmammals, collision of a large meteorite leading to biotic andhundreds ofplantswerealso suddenly wiped crisis. The volcanic theory argues that the pollu- out at this time. The small calcareous plankters tion in theatmosphereandoceans bythemassive that floatatthe ocean surface and theammonites outpourings ofDeccan flood basalt in India had devastating effects on ecology. and rudists from the depths also vanished. The Earthwasdevastated.Lifewasravagedbyoneof KTIne1x9t8in0c,titohne Awlavsarceazusgerdoubpyprthoepoismepdacmtatofthae the worst catastrophes. There were survivors, ofcourse, after the KT g1r0ekatmdmeeatleoorfitien.teTrheists aprmoopnogsalsciheanstigsetsnearnadtedthea dmiasmamstaelrs,(Ficgr.oco1d).ileNse,o—rtnurittlhesi,nelizbairrddss,anpdlascneankteasl mpuibslsiicn.g.BuIftathheugkeeymeptieeocreitoefheavdidienndceeedwcarsasshtieldl all survivedasgroups despitetheextinctionof intotheEarth,wherewasthecrater?Criticssear- some species. From this catastrophe, oppor- ched for alternate explanation. The end of the tunitiesaroseforthesurvivors.TheKTextinction Cretaceous was also a time ofmassive continen- had opened thedoorforthe ageofmammalsand tal flood basalt volcanism, especially the Deccan theriseofbirds andchangedthe courseofevolu- Traps in India. Many palaeontologists believe tionary history. What triggered this catastrophe that such cataclysmic volcanism may have been that led to such an unprecendented ecological the culprit in the KT extinction (McLean, 1985; crisis? Over the years, many theories, some biz- Officer et al., 1987). Over the past 15 years, zare and some plausible, have been offered to exciting new insights have poured in from vir- explain the mystery behind the extinction of tually every branch ofearth and planetary scien- dinosaurs. There is no shortage of murder ces to understand—ing the effects of t—hese suspects. Any explanation ofthecauses ofbiotic catastrophicevents impactandvolcanic on crisesmustfocusonfindingagentsofdestruction earth's ecosphere and the evolution of life. thataffectedenvironments,climates,ecologyand Recently the Chicxulub structure in the northern organisms. coast of Yucatan Peninsula of Mexico has KT EVENTS IN INDIA 49 emerged as a prime candidate for the KT impact lisions is seen in thespectrum ofcratersizes and site (Hildebrand et al., 1991). In this paper w—e morphologies. The recovery of meteorite frag- describe another buried KT impact structure ments and shock effects within or surrounding a the Shiva Crateratthe India-Seychelles rift mar- crater are the most persuasive evidence for an gin, and its relevance to the Chicxulub structure, impact origin, but when large craters are deeply Deccan volcanism and mass extinction (Chatter- eroded or buried, the evidence of impact is jee, 1992; Chatterjee & Rudra, 1993). obscuredorblurred. Suchevidence may beiden- tified indirectly from shock-metamorphiceffects THE IMPACTMODEL on the targetrock andejectacomponents, as well asdistinctivegeochemical signaturesattributable to a particular type of meteoritic projectile Likeotherplanetsinthesolarsystem,theEarth resides in a swarm ofasteroids and comets. It is (Grieve, 1990). These signatures may be now apparent that the Earth has been heavily preservedlocally neartheimpactsite,orglobally bombarded during its history by meteorites of at a particular stratigraphic level containing the varioussources,sizesandcompositions(Clube& ejectafallout.Togetherthey mayprovidecluesto Napier, 1982). The incontrovertible evidence for thenatureofthetargetmaterial and the impactor. large cosmic collisions is the occurrence of cir- Impact craters ofthis obscurenaturearethemost cular craters associated with considerable local controversial and require additional information structural disturbance and shock metamorphism for verification. (French & Short, 1968). Because ofthe dynamic In 1980, theAlvarezgroupadvancedastartling nature of the terrestrial lithosphere where such —theory toexplain the sudden demise ofdinosaurs forcesaserosion,volcanism,deposition,orogeny the most succesful land animals ever to arise andplatetectonicsconstantly restructure thesur- on Earth. They discovered an abnormally high face, impactcraters are often erased orobscured, concentration of iridium (about 30 times more unlikethe morestatic surfacesoftheMoon, Mer- than the surrounding rocks) at the KTboundary cury and Mars. To date over 150 impact craters level ofGubbio,Italy. Soonacomparableiridium have been recognized on the Earth's surface and anomaly was found globally at different KT the list is growing. They range in size from ap- boundary sections (Orth, 1989). Since iridium is proximately 100m to 200km and in age from a very rare element in theearth's crust, butfairly Precambrian to Recent (Grieve, 1987; 1990; abundant in chondritic meteorites, the Alvarez Grieveet al., 1988).The spatial distribution indi- team proposed that the iridium spike at the KT cates concentrations in cratonic areas. Other boundary iscosmic in origin, implying the strike craters may be submerged under oceans and ofa large meteorite.Therewasenoughiridiumin remain inaccessible or undetected. Scientific in- the KT boundary, they calculated, to equal a terest in the role of impact in geological and lOkm-diameterasteroid. They proposed thatthis biological evolution has been enhanced by giant asteroid crashed into the Earth with a severaldevelopments inrecent years. Amongthe velocity of 90,000km/hour to cause the most prominent of these are the hypotheses of worldwidecatastrophicevent.Thisimpact lofted Alvarez et al. (1980) concerning terminal so much debris into Earth's atmosphere as to Cretaceous extinction and lunar and planetary createra 'nuclearwinter'thatcaused muchofthe exploration by manned and unmanned life on Earth to perish. A blackout of sun would spacecraft. As interest in bombardment mounts, kill plants anddestory thefoodchain. Theglobal previously unknown or cryptic impact sites are distributionoftheiridiumlayerwascausedbythe recognized with increasing frequency. As a grim impactandvaporizationofthebolide.Theimpact reminderthat the threatofimpacton ourplanetis theory was reinforced by additional evidence a real possibility, the world's attention was such as shocked quartz(Bohoretal., 1984; 1987; focused during lateJuly, 1994, on the spectacular Owen & Anders, 1988), stishovite (McHone et collision ofcomet Shoemaker-Levy 9on thesur- al., 1989), micro-diamonds (Carlisle & Braman, face ofJupiter, leaving scars the size ofEarth on 1991), impact glasses (Izett et al., 1990), os- the giant planet (Levy et al., 1995; Weissman, mimium isotope ratios (Turekian, 1982), Ni-rich 1955). spinels (Robin et al., 1994), rhodium (Bekov et Hypervelocity impacts can have a large range al., 1988), carbon soots (Wolbach et al., 1988), ofeffects that depend on the strength and density tsunami deposits (Bourgeois et al., 1988), and of the projectile and the nature of the target extraterrestrial amino acids (Carlisle & Braman, material. The most obvious result of larger col- 1993) in the KTboundary layeratdifferent sites. 492 MEMOIRS OFTHEQUEENSLANDMUSEUM Among all thiscumulativeevidence, theshocked Hawaiian and Reunion volcanoes, suggesting quartz is a distinctive signature ofimpact event that the iridium anomaly at the KT boundary as itcan form ata force more than 10gigapascal could have also had a volcanic origin (Olmez et (GPa) that travels through quartz-bearing grains al., 1986). There is no doubt that such a massive ofthe target rock to porduce microscopic shock volcanic outburstoveran extended period would lamellae (Grieve, 1990). Pressures and tempera- have deleterious environmental consequences. tures producedby a large body impact are much Proponents of the volcanic model claim that greater than those generated by other geologic manyofthesupposedimpactsignaturesattheKT processes, such as volcanic activity, mountain boundary layer, such as iridium enrichment, building and earthquakes. shockedquartz,microspherules,claymineralogy However, the strongestevidence in favorofthe and carbon soot, could have volcanic explana- KT impact event would be to locate a crater tions (Officer, atal., 1987; Courtillot, 1990). The marking the point of collision. Such a crater impact proponents disagree. They point out that should be 150km or more in diameter (Grieve, the Deccan volcanism was not of an explosive 1982). The search for an impact site ofthe right typeandcouldnotaccountfortheglobaldistribu- age (~65Ma), and the right size (150km) has tion ofthe iridium anomaly, shocked quartz and continued, includingreassessmentofmany enig- tektites atthe KTboundarylayer(Alvarez, 1986; & maticstructures.Now,afteradecade-longsearch, Alvarez Asaro, 1990). Moreover, the lamellar theChicxulubstructureontheYucatanPeninsula features in quartz grains associated with ex- ofMexico(Hildebrandetal., 1991)andtheShiva plosive volcanism show curvatures contrary to Crater at the India-Seychelles rift margin (Chat- theplanarand parallel lamellae in impact-related terjee, 1992) appear to be two potential can- shocked quartz recoved from the KT boundary didates forthe long-sought KT impact scar. (Izett, 1990). Thegradual extinction pattern seen among organisms may bean artifact ofpreserva- THE VOLCANIC MODEL tion and poorsamplingquality. Others arguethat because Deccan volcanismhadlittleeffectonthe diversity of local Indian biota, its catastrophic Although the impact hypothesis is very com- role in global life is questionable (Prasad et al., pelling, not everybody believes that impacts 1994). killed the dinosaurs and other organisms at the KT boundary. Critics have advanced a volcanic SEARCH FOR THE KT IMPACTSITE alternative. TheendoftheCretaceous was alsoa time of massive continental flood basalt volcanism, especially the Deccan Traps of India. Since Alvarez et al (1980) presented their Recent radiometric dating suggests thatthe main geochemical evidence for an impact event at the pulse of Deccan volcanism may have occured KTboundary, the search forthe proposed impact close tothe KTboundary at 65 million years ago crater has continued. There are a number of (Duncan & Pyle, 1988; Courtillot, 1990). Many candidates forthe KTimpactsite, none ofwhich palaeontologists argue forcefully that such are very compelling at present. The 35km Man- cataclysmic Deccan volcanism may have been son structure in north-central Iowa is such a can- the main contributing factor for the biotic crisis didate (French, 1984; Anderson & Hartung, at the KT boundary (Clemens, 1982; Officer et 1988), but new work suggests that this crater is al., 1987; Hallam, 1987; Keller, 1989; Stanley, older(~74mya)andplayednoroleintheKTmass 1987; Zinsmeisteret al., 1989; Courtillot, 1990). extinction (Izett at al., 1993). Twin impact struc- Othercontemporaryepisodes ofvolcanismatthe tures in the Kara Sea in the former USSR, the KTboundary such as in Cameroon and the Coral Kara (diameter, 60km) and Ust-Kara (diameter, SeahavebeenlinkedtotheKTevent(Sutherland, 25km), have been proposed as possible impact 1994). The proponents of the volcanic model sites (Koeberl et al., 1988), but recent argue that the KT extinction was neither global, geochronologic data suggestthat these structures norinstantaneous, but occurredoveranextended are also older than the KT boundary event period oftime, because different organisms dis- (Koeberl et al., 1990). appeared at different levels at or near the KT Even the general location of the KT impact, boundary. Such step-wise extinction pattern whether continental or oceanic, remains con- could be best explained by prolonged emissions troversial. Trace element and isotopic studies of ofvolcanic pollutants. Large amounts ofiridium the highly altered KT boundary-layer com- have been discovered to be spewing from the ponents tend to support the oceanic impact KTEVENTS ININDIA 493 hypothesis (Gilmore et al, 1984; Hildebrand & The most promising KTimpact site appears to Boynton, 1990). On theotherhand, thepresence be the Chicxulub Crater on the northern margin of shocked quartz at several KT boundary sites of the Yucatan Peninsula, Mexico (Penfield & would indicate a continental site (Bohor at al., Camargo, 1982; Hilderbrand et al., 1991; 1995). 1987). The apparent contradiction can be recon- Itisacircularstructureabout 180kmindiameter, m ciledifthesingleimpactoccurredatacontinental buriedunder 1,100 ofcarbonatestrata,extend- margin involving both oceanic and continental ing outundertheGulfofMexico, and definedby crust, ormultiple impacts at different sites. magnetic and gravity anomalies (Fig. 2A). Itwassoonrealizedthatthesizeandabundance of the ejecta, such as shocked quartz grains and THECHICXULUB CRATER tektites, may givesomecluesastothelocationof the impact crater. Bohor et al. (1987) and Izett (1990) concluded that the largest sizes and The subsurface stratigraphy of the Chicxulub gKrTeatbeosutnadabruyndsaendciemeonftsshooccckuerdinquwaersttzegrrnaiNnosrtihn pstlrourcattuiroenisbkonroewnhoplreismadrriillylefdrobmypePtermoelexu,m ethxe- America, suggesting that the impact occurred on Mexican national petroleum company, in the or near the continent. The discover)' of tsunami 1950s (Lopez-Ramos, 1975; Meyerhoff et al. deposits at the KT boundary sections on the 1994). Unfortunately, most of the critical core BrazosRiver,Texas(Bourgeoisetal., 1988),near samples were destroyed in a warehouse fire. At Braggs, Alabama, (Smit et al., 1994), in the present, samples of the Chicxulub structure are & Caribbean (Hildebrand Boynton, 1988), and limited; as a result, the subsurface stratigraphy is Deep-Sea Drilling Program holes 536 and540in open to various interpretations. In hindsight, it is the southwestern Gulf ofMexico(Alvarezetal., ironic that drilling and exploration were stopped 1992) as well as the identification of tektites at as soon as the andesitic bodies at a depth of Beloc, Haiti (Izettetal., 1991; Maurrasse & Sen, 1500-2000m were encountered; these may have 1991) and Arroyo el Mimbral, northeastern provided the critical evidence for the impact. Mexico (Smit et al., 1992), narrowed the search Penfield & Camargo (1982) suspectedan impact further to the Caribbean region. Impact breccia origin for the Chicxulub Crater on the basis of hasbeenrecoveredfromAlbion IslandofBelize, concentric geophysical anomalies with as- near the Mexican border (Ocampo & Pope, sociated extrusive material such as andesitic 1994). AtleastfourpossibleCaribbeansiteshave bodies. Recently located samples from the old been suggested, including the Colombian Basin, Pemex wells, including brecciated carbonates, western Cuba, Haiti andthe Yucatan Peninsula. andesites and crystalline basement have been Hildebrand & Boynton (1990) placed the KT studied extensively and inferred to support the impactlocation in theColombianBasin between impact scenario for this site. For example, Hil- Colombia and Haiti on the basis ofseismic data debrand et al. (1991) reported shocked quartz andDSDPcore samples,buttheputativecrateris within Chicxulub breccias and documented chemical and isotopic similarities between an- notonly under waterbutburied under 1,000mof desitesandtektitedepositsfromtheKTboundary sediment and is subject to other interpretations, sections of Haiti and Mexico. These findings sneuscshoafsttheectoocneiacnoircigcirnusot.r Baochhoarng&e Sienitthze(t1h9i9c0k)- isnoduirccaetefothratthteheHaCihtiicaxnulaunbdCMraetxeircamnaytebketittehse. speculated that the impact site was near Cuba, Kring & Boynton (1992) interpreted the rocks about 1,350km from the site proposed by Hil- initially thought to be andesites as probable im- debrand and Boynton, on the basis ofa boulder pact melts, whereas Blum et al (1993) found an bed interpreted as ejecta components, but the isotopic match between the Haiti glass and the boulder bed is found to be of local, weathering Chicxulub melt. Subsequently, Sharpton et al. originandtheCubansitehasbeenrejected (Dietz (1992)recognizedthatthebrecciaabovethemelt &McHone, 1990). TheMassifdelaHotteon the rock is suevite breccia, a distinct signature ofan southern peninsula of Haiti, a mountainous impact crater. They recognized that the Chic- region with Cretaceous sediments, has alsobeen xulubmeltrocks showhigh levelsofiridium and proposed as the KT impact site (Maurrasse, theiragecorrespondswell withtheKTboundary. 1990), but closerexamination ofthe area clearly Single crystal °Ar/39Ar dating places the melt indicates that it is not an impact site (Officer et rock at 65Ma (Swisher et ah, 1992). All this al., 1992). combined evidence suggests that the Chicxulub h 1 494 MEMOIRS OFTHEQUEENSLANDMUSEUM UNITED STATES • KT Ejecta Layer • Tsunami Deposits +»-+++ 4-4-+--I-++4-I++ +f-Jt--+4+f+t4t+4-++HH—+- -n-+-[-4++++B4a+s+e|mH-e-tn-t+-++f»-+1+f-if1ftf+ FIG.2. A,LocationoftheChicxulubstructureonthenorthernedgeoftheYucatan Peninsula,Mexico,showing distributionofproximalimpactdeposits.B,Cross-section;estmatedcraterdiameteris 180km(simplifiedfrom Hildebrandetal., 1994). structure may be a prime candidate for the long workers dispute the impact origin of the Chic- sought KTimpact crater(Fig. 2B). xulub structure and interpreted it as a buried It appears from the above discussion that two volcanic complex of Late Cretaceous age (Of- lines ofevidence support the impactscenariofor ficer et al., 1992; Meyerhoffet al., 1994). Their the Chicxulub structure: evidence of shock counterargumentis basedon the original subsur- metamorphism, iridium enrichment, brecciation face correlation proposed by Lopez-Ramos and impact melt within the crater itself, and the (1975) and the unpublished well log for the distributionofproximal ejectacomponentsatthe Yucatan No. 6 well, drilled in 1966 over the KT sections in Haiti, Mexico, Texas, Alabama, Chicxulub structure. One of the authors, Dr Ar- theCaribbeanandadjacentareas(Fig.2A). How- thurA. Meyerhoff, was a consulting geologistto ever, in recent times both interpretations have Pemex atthe time Yucatan No. 6 was drilled and faced strong criticisms. For example, some had first-handinformation onthe biostratigraphy KTEVENTS IN INDIA 495 ofthe site. This well penetrated a superimposed breccia show curvatures typical of volcanic sequence of Pliocene-Miocene, Oligocene, origin. Eocene-Palaeocene and Maastrichtian-Cam- 7) Later thermal events in the Chicxulub vol- panian sediments, and bottomed in andesitic canic sequence might have reset the radiometric rocks and Cretaceous limestone, dolomite and age of the andesite; these authors indicated that anhydrite. The most damaging evidence against out of ten samples analyzed by Sharpton et al. the impact origin is the inverted stratigraphy of (1972), nine gave spurious results; accordingly, theCretaceoushorizon in relationtotheandesite. the correlation between the KTimpacteventand Thesedissenterspointtothepresenceof350mof the presumed andesite melt is tenuous atbest. undisturbed Late Cretaceous sediments with 8)TheChicxulubvolcanicsarenotlocalimpact index fossils (foraminifera) overlying the an- melt, but part of a well-known Late Cretaceous desitic body. If the andesitic rocks were indeed igneous province surrounding the Gulf of impact melt from KTboundary time, the overly- Mexico. ing strata must be Palaeocene oryoungerin age. Not only the impact origin of the Chicxulub Palaeontologic evidence indicates otherwise; structure, but also the interpretation ofproximal these strata are ofCampanian and Maastrichtian deposits ofballistic ejecta and impact-wave dis- agelyingconformablyovertheandesite. Swisher turbances in the Caribbean and GulfCoast is in etal.(1992)explainedthisstratigraphicinversion dispute. Recent studies have shown thatmany of asfallbackbrecciaofCretaceouslimestoneinfill- these socalled impact-generateddeposits may in ing the crater. However these overlying fact represent gravity-flow or turbidite deposits Cretaceous strata are not disturbed, brecciated occurring over an extended period of time, and shocked, and thus posea problem forthe KT whereas supposed impact droplets are altered impact age ofthe andesite body. volcanic particles(Lyons &Officer, 1992; Keller Meyerhoffetal. (1994) indicatedthefollowing et al., 1993; Stinnesbeck et al., 1993; Beeson et additional discrepancies in the impact origin al, 1994; Adatte etal, 1994). hypothesis ofthe Chicxulub structure: The presumed connection between the Haitian 1) The Chicxulub structure is too shallow glasses and Chicxulub has been questioned by (~2000m) for an impact of this dimension; ex- Koeberl (1993) on the basis of geochemical pected excavation depth would be around 10km evidence. He pointed out that, at the time of or greater (Melosh, 1989). impact, the Chicxulub area was covered by 2) Several layers of bentonitic breccia occur evaporitic and carbonate deposits several interbedded with the Cretaceous limestones kilometersthick. Yellow glasses found in the KT withoutanystructuraldisturbanceorobliteration, section of Haiti are not tektities and cannot be so their impact origin is suspect. linkedtoevaporitictargetrocksoftheChicxulub. 3) If the andesite were of impact origin, one Similarly, the interpretation of some breccias would expect highly homogeneous composition within Chicxulub as impact breccias may be with appreciable thickness; in contrast, chemical wrong. However, proponents of the Chicxulub analysis suggests that the Chicxulub andesite is Crater have dismissed most of these criticisms thin, chemically inhomegeneous with a wide (Alvarez etal. 1994; Hildebrand et al., 1994). range of major oxide compositions (Sharpton et Thus,theChicxulubimpactisopentoquestion, al., 1992). and more study is needed before a final assess- 4) In an impact structure, the impact melt rep- ment can be made. The only way to settle this resents the aggregate composition of the target question decisively is by drilling new holes into rock (Engelhardt, 1984); in Chicxulub, on the the Chicxulub structure, as in the case of the other hand, the target country rock is metamor- Manson Crater, to determine more precisely phosed quartzite and rhyolite, whereas the when and how it formed. presumed impact melt is andesite. 5) The anhydrite at the bottom of the Yucatan THE SHIVACRATER No. 6 well would have been completely vaporized at thepoint ofcollision, ifthere wasan Although the Chicxulubstructure hasemerged impact; its presence below the putativemeltrock as the leading candidate for the KT impact scar, is anomalous. anotherpromising KTimpact sitehasbeen iden- 6) Unlike the planar and parallel lamellae in tified at the India-Seychelles rift margin in the shockedquartzassociatedwithanimpactsite,the northwest Indian Ocean, almost antipodal to the lamellar features in quartz grains in Chicxulub Chicxulub structure. Hartnady (1986) suggested 496 MEMOIRS OFTHEQUEENSLAND MUSEUM o 10-s — 20 S 30°S 50°E 607E 70°E FIG. 3. Location ofthe Shiva Crater at the India-Seychelles rift margin during KT boundary; arrow indicates trajectory of meteorite; radial asymmetric distributions of alkaline igneous complexes (impact melt rocks) v downrange ofthe Shiva Crater are shown by closed circles (data from Bose, 1980); areas ofpositivegravity anomaly (datafrom Biswas, 1988)coincidewith theejectameltdistribution; asterisk indicateslocation ofthe Deccan-Reunionhotspot. that the Amirante Basin, south ofthe Seychelles southwest by the structure ofthe Amirante Arc. Island, may be a possible KT impact site. The Sediments from the adjacent Amirante Passage basin has a subcircular shape ofabout 300km in have yielded Late Maastrichtian foraminifera diameter, bounded on the northeast by the Abathomphalus mayorensis (Johnson et al., Seychelles Bank and partially ringed on the 1982), while basalt samples dredged from the KTEVENTS ININDIA 497 Amirante Arc look like Deccan Trap and have a taneoustectonicandvolcanicfeaturesmight—have similarradiometricage(Fisheretal., 1968). Both been triggered by a single physical cause an palaeontologic and radiometric age indicate that enormous impact. the arc was formed nearthe KTboundary. How- Chatterjee (1990; 1992) elaborated upon this ever,itsarcuatestructureisenigmatic.Itdoesnot KT impact scenario at the India-Seychelles rift appearto bea recentorancienttrench, as itlacks margin,andidentifiedtheeasternrimofthecrater volcanic activity, seismicity, and any significant along thePanvel Flexure, nearthe Bombaycoast accretionarysedimentaryprismonits iandward' (Figs. 3-6). The Panvel Flexure is an arcuate side(Johnsonetal., 1982).Thustheinterpretation segment of the crater about 120km long on the of the Amirante Arc as a crater rim is a distinct Deccan Traps, and it is difficult to explain in possibility. The proposed impact also may ex- termsofconventionaltectonics.Itismarkedbya ptlhaeinKtThebopuuznzdlairnygjduurmipngofthteherCifatrilnsgboefrgInRdiidagaenadt slieniesmoifchiotty s(pKraiinglsa,edtikaels.,, d1e9e8p1;cruPsotwalarf,aul1t9s8a1n)d. the Seychelles. Hartnady noticed that the Carlsberg Ridge between the Seychelles and Since the Indian shield is usually aseismic, the Madagascar jumped more than 500km to the seismicityalongtheflexureisunusual,indicating northeastto lie between India andthe Seychelles tectonic instability. The geothermal gradient is andinitiateriftingbetweenthesetwolandmasses. abnormallyhigh along this flexure(36-78°C/knf') He could not find any evidence for plate reor- with evidence ofthinned lithosphere (31-39km), ganization in the Atlantic or Pacific Oceans suggesting melting conditions at shallow depths during this time. He attributed this major plate (Negi etal., 1992). ThePanvel Flexure may rep- tectonic adjustment to the enormous force of a resent the eastern rim ofthe craterin the formof large meteorite. As additional evidence, he a collapsed rim structure. It exercises tectonic pointed to massive tsunami deposits in the KT controlontheattitudeoftheDeccan lavas.Tothe boundary section of Somalia and Kenya, which eastoftheflexure,thebasalticflowsarehorizon- may be linked to this impactevent. tal; to the west ofthe flexure, the basaltic flows dip west to west-southwest at 50° to 60° toward Although Hartnady's model initially had at- the coast. The abrupt change of dip along the tracted wide attention, one majorproblem ofhis flexure axis may indicate the slopeoftheeastern ideaistheenigmaticmorphologyoftheAmirante crater wall, which is now concealed by Deccan Basin.Itissemicircularinoutline,preservinghalf of a supposed crater rim. What happened to the lavas. Seismic data indicate that the basement topography below the Deccan lava west of the otherhalfofthecrater?Althoughcircularityisnot diagnostic of impact origin, there must be flexure has a crater-like depression (Kaila et al., evidence of some structure off the coast of 1981). Completing the oval by combining the Seychelles. Is itpossibletofind the missingrim? Amirante Arc and the Panvel Flexure, the extent ofthe cratercanbe extrapolated. Itis agiantoval Altetal. (1988) remediedthedeficiencyofthe crater,600km longand450kmwide,showingthe TAmhieryanatregBuaesdinthmaotdedluraisngtheKTpoibntouonfdacorlyl,isitohne. morphologyofacompleximpactscar.Chatterjee (1992) named this impact structure the Shiva Seychelleswasadjacenttothewestcoastofindia. Crater, after the Hindu god of destruction and They concurred with Hartanady (1986) that the renewal (Fig. 3). However, the Shiva Crater is westernrimofthecratersurvivesintheAmirante Arc, butthe eastern rim lies along the westcoast dliafrfgiecluyltctooncienatelrepdrebtybetcheauDseecictains sluavbam.arMianneyanodf ofIndia, hidden by the overlying Deccan Traps. They speculated that the impact was forceful theimpactsignaturesarethuserasedorobscured. enough to create not only the enormous crater Moreover, the rifting of the Seychelles from approximately 600km in diameter, but also to India, which occurred along the width of the cause pressure-release melting in the astheno- crater, has obliterated the geomorphology ofthe sphere. Basaltthen filled thecraterbasin toform structure. A series of geodynamic and volcanic an immense lava lake, the terrestrial equivalent eventsthatoccurrednearKTboundarytimemust of a lunar mare. The synchrony of initiation of be untangled and put into proper chronologic spreading along the Carlsberg Ridge, the order to unveil the crater morphology. Recent emplacement of the flood basalts at the Deccan exploratory data from the Bombay High, a giant plateau, SayadeMalhabankandAmranteBasin, offshore oilfield located 160km west ofthe city as well as close spatial association around the ofBombay, hasproduced a wealthofinfomation crater basin, indicate that the array of simul- supporting an impactorigin fortheShiva Crater. 498 MEMOIRS OFTHEQUEENSLANDMUSEUM 90°E 20°N 10°N- 10°S- 20°S- 500 km 30°S- FIG. 4. Presentday location ofthe split Shiva Craterin reference to India and Seychelles on eitherside ofthe CarlsbergRidge.Today,partoftheShivaCraterisattachedtothesouthernpartoftheSeychelles,theother-half tothewesternpartofIndia. Thecraterwasjoined65 millionyearsago when theSeychelles waspartofIndia beforethespreadingoftheCarlsbergRidge. INIDA-SEYCHELLES FIT. Today,theSeychel- of Deccan volcanism and shows matching les microcontinent is separated fromthe western geological provinces. The largely submerged coast ofIndia by 2,800km because of spreading continental block that bears the Seychelles Is- along the Carlsberg Ridge (Fig. 4). This lands contains enormous flood basaltdeposits in midoceanic ridge shows symmetrical magnetic thesubmarineplateauoftheSayadeMalhaBank, anomalies of5, 23, 24, 25, 26, 27, 28 and 29 on which are inferred to be an extension of the eithersideoftheridgeaxisbetweenIndiaandthe Deccan volcanism (Meyerhoff & Kamen-Kaye, Seychelles (Norton & Sclater, 1979; Naini & 1981;Backmanetal., 1988),especiallytheBushe Talwani, 1982). Using both palaeomagnetic and Formation of the Upper Deccan Basalt Group palaeontologic evidence Chatterjee (1992) has (Devey &Stephens, 1992). Thelinkbetween KT restored the India-Seychelles fit for the KT magmatism on the Seychelles and India is em- boundary time (Figs. 3 & 8). The reconstruction phasized by matching the geochemistry and places the western coast of India against the geochronology of alkaline rocks (White & Mc- Seychelles-SayadeMalhaBankataboutthetime Kenzie, 1989;Devey&Stephens, 1992).Similar-