______-= o~C~E A~N~O~LO~G~I~C~A~A~C~T~A~.~l~9~Bl~.~N~.~S~P_1~~----_

DSDP A summary of Deep Sea TeclOnic TrCllch DSDP Drilling Project Leg 67. Guatemala Subduction Tectonique Shipboard results F.. ~ from the Mid-America trench transect off Guatemala

J . Aubouin . , R. von Huene ", J. Al.éma ", G. Slackinton <. J . A. Carier ' . W. T . Coulbourn d, h l D.S.Cowan ", J.A.Curiale', C.A.Dengo·, R. W, Faas , W.H,\rrÎson', R, Hesse • D, M. Hussong <. J . w. Ladd i, N. MuzyJev k. T . Shiki', P. R. Thompson J, J . Weslbcrg d • • Département de Géologie Stru cturale. Univer ~i té Pierre et Marie Curie. 4, place Jussieu. 75230 Paris. . b US Geologiea! Survcy, Menlo Park, Califorlli« 94025. USA. < Department of Geology and Geophysics, Universit y of Hawaii , Honolulu. Hawaii 96822. USA. d Scripps Institution o f Oceanography, La Jalla, California 92093. USA . • Department of Geological Sciences. Uni versit y of Washington. Seanle. Washington 98195. USA. r School of Oeology and Geophysics. University of Oklahoma. Norm'In. Oklahoma 73019. USA. 1 Center for Teclonophysics. Texas A & M Uni versity. College Stalion. Texas 77843. USA. and ICAITI . Guatemala. . h Department of Geo108Y . Lafayene College. Easton. Pennsylvania 18042. USA. 1 Technische UniversÎtiit . Munchen. fRG. 1 Lamont-Doherty Geological Observatory. Palisades. New York 10964. USA. l Geological InstÎlute o f USSR. Academy of Sciences. Moscow. USSR. , Departmcnt o f Geology and Mineralogy. Kyoto University. Kyoto. Japan.

ABSTRACT The Middle America Trench off Guatemala was transected by 24·channcl seismic-reflection surveys. seismic-refracti on surveys. ;Ind drilling with the G/()mar Chfll/flllger. The drilling was done al three sites on the oceanic and fou r si tes on the . Thcse plaies converge m about 10 cm/yr as indicated by global plate reconstruction. At ail drill sites sediment of upper miocene through Quaternary age is almost entirely hemipelagic mud with interbedded thin volcanic ash. except in the trench where mud and fine sand turbidiles less than 400.000 ye:us old are ponded. However . the undcrlying rocks are very diffcrent. On the oeeanic Cocos plate a bas'II chalk sequence of lower and middJe Miocenc age is overlain by a thin inlerv;,1 of abyssal clay. In contras! is Ihe Crelaceous 10 lower Miocene clayslonc sequence recovered only at a sile 3 km landward of the trench axis where drilJing penetrated the hemipelagic slope deposÎ ls. A large amount of sediment along with ocean crust has been subducted ouring the present (Miocene 10 Quaternary) episode of subduction. and parts o f the continenlal fmmework may have bee n subducted as weIl. No current tcclonic model salisfa ctoril y explains the surprising occurence of Cretaceous to Miocene daYSlone atthe fool of the continenl. Oceaflo{. A cta. 1981. Proceedings 26'· International geotogical Congress. Geology or COnli nental mllrgins symposium . p ~,r i s. July 7-17. 1980. 225-232.

225 J. AUBOUIN et al.

RESUME Bil an des résultats de bord du Lcg 67 (Deep Sen Drilling Projcct : fosse d'Amérique centrale a u large du Guatemala). La fosse d'Amérique centrale a été étudiée selon un profil au large du Guatemala. p;lr sismique réfle:lOion (24 canaux). sismique réfraction. et forage du Glomar Challenger. Les forages ont été effectués il 3 sites sur la plaque de Cocos et 4 sites sur la plaque caraïbe. Le taux de convergence . déduit du mouvement global des plaques. est de 10 cm par an. A tous les site~. les sédiments. du Miocène supérieur au Quaternaire. sont presque entièrement des boues hémipélagiques interstr atîr îée~ de fines couches de cendres \'otcanique~ : sauf dans la fosse où se sont déposées des IUrbidites de boues et sables fin~. d'âge plus récent que 400 000 an~. Cependant les roches sous.jacentes sont très différentes. Sur la plaque de Cocos. une séquence basale de craie d'âge Miocène inférieur et moyen. est recouverte par un fin ni veau d 'argiles abyssales. Au contraire. une séquence de schistes argileux. d'âges allant du Crétacé supérieur au Miocène inférieur. a été forée. à moins de 3 km de la fosse. sous les dépôts hémipélagiques du bas de la pente continentale. Une partie importante des sédiments ont été entraînés avec la croûte océanique pendant le présent épisode de subducti on. du Miocène au Quaternaire. de même. probablement. qu' une partie de la pente continentale elle·même. Aucun modèle tectonique courant n'explique de manière satisfaisante l'existence de la série crétacée-miocène au pied de la marge continentale.

OcelUlOl. Acttl. 1981. Actes 26" Congrès Internati onfll de Géologie, colloque Géologie des marges conti nentales. Paris. 7-17 juil. 1980. 225-232.

INTROD UCTI ON The Acapulco segment of the Middle America Trench lies along the North American continent : the Guatemalan The Mid-America Trench transect off Guatemala is the segment runs along Central America, whîch is basicaJJy a second of two geophysical and drilling transects recommen­ tectonic edifice of old oceanic crust covered by tertiary ded for the Internatiomtl Program o f ocean DrilJina (11'00) volcanics. The Acapulco segment abruptly truncates the across the convergent margin of southern and Sierra Madre structures; the trench axis is ;It 4S ~ to the Central America. The first. a tra nsect off Oaxaca. Mexico. Miocene tectonics axis of the Sierra Madre dei Sur and was drilled on Leg 66 of the Glomar ClrallenRer to in vesti­ there is only a very narrow continental shelf along tho Co.1St. gate a convergent marglO where the leading edge is a The GU>l temala segment does nOI eut the Centnll American complex tecto nically accreted during the Neogene and from structure: the trench lIxis is parnllel 10 the tectonic edifice the mid-slope landward the terrdne is a truncated Precam­ that crops out for instances in the Nicoya peninsul a (Costa brian a nd PaJeozoic continental framework (Moore et al .. Rica) and there is a wide continental shelf. which is 1979). Of( Guatemala our principal objecti ve was to study underlain by a tertiary basin with a thick sedimentary this mit rgin where geophysical data indicated çont i n uou ~ sequence (Seely. 1979). accrelion and imbri cation (Seely et al .. 1974). Seely and his Il has been inferred Ihat the Mesm;oic igneous rock çrop­ colteagues made a case for the imbricate thrust model that is ping out in the Nicoya pc ninsula underlies the Tertiary basin widely accepted and th i~ model is supported by the multi­ o n the continental ~ h el r. A sUong magnetic anomaly, channel seismic refleçtion s ite"Sur vey~ (Ladd el (II. 1978; running alonK the continental ~h~lr. can be projected from Ibrahim el al. 1979). the Nicoy~ peninsula along the edge of the shelf (Ladd This paper is patterned after a more complete report et ai .. 1978). Mesozoit: sequences crop oui. from north to prepared on the prelimin,try results of Leg67. in which the south. in the Santa Elena. Nicoya, Osa. Azuero peninsulas. data from the cores are discussed in greater detail (Aubouin They consist of ; et al.• 1979: von Huene ft al .. 1980). 1) the Nicoya complex. a sequence of oceanic ~ed i ments (graywackes. cheriS, r:.tdiolarites) associated with volcanics (diabasis. pillow htvas). containing rocks of uppermost Jurassic, lower Cretaceous. middle Cretaceous. and upper GEOLOGIC SETIING Cretaceous age: 2) the S'lOta Elena complex. composed of a thick ultramafic The Middle America trench is separated by the Tehuantepec body thrust upon a \'olc~lOo- s ed im enta r y sequence with a ridge in two segmenlS (Auboui n. Tardy, 1980) : the Aca· litho-facies si milar to that o f the Nicoya complex complex pulco segment in the norlh and the Guatemala segment in and consisting of uppermost Crelaceous sedimentary rock the south. A corresponding fealure on land. is the active (upper Campa ni an.Mae~trichlian) which reSh unconforma· Polochic- zone. which separates the mexican bly upon pe rid oti t e~ (Azéma. Tournon. 1980): cordillera (S ie rr

226 • LEG 67 : MID·AMERICA TRENCH (GUATEMALA) sequences of Central America are quite differecl trom the data ind ieate that the oeeanie crust is sub

Figure 1

A . (uppu diagram) PI/lle leetm,ie dial-tr/lm al Ih e Caribbeml aua shawing Ihe ulatian al Ihe Caribbean and Cocos plates alon!; Ihe Middle-America Trench. The plaieS are pre.flnlly eom'erging III 10 cm/yr acroS$ the trench. Key III s)"mbols: 1-4 : Zones wilh cOn/inenllll b/lumenl: 1. andesitic m/canic zones ",ilh grandioriric plU/ons ( ...eslern Sierra Madre. Ctrrfral Culombian cordillera) : 2. non -~o/ca nic zonts, ollen Wilh fI )'$ch ladfs (e /l stern Sierra Madre, sou/hern Siura Madre, taSlern Colombian cordillua. fl ysch nappes 01 Vene<.uela); J. carbo­ nate p/atlorm, lolded {Tt xas platlorm. Sierra Madre dt Chio. pa~} ur non-Iolded (Honduras. Yucalan. Cuba, Bahtlmas) : 4 , mo/assie l ormatiOllS ( ...eS I side ol'he Gull al Mexico. Llullos 01 Vene,uela and Colomhia). 5: Zones with oceanic sequences: chens, radio/arires. assoei(l· ted ... ilh ophiolitle bodies. metamorphosed or nOI in Ihe billes· fZJ • ..z 00 1 I::J. DIIDlJ li O . ....-< J ..,...... • .,-d 1. chisl facies, OI'enhruSled on lhe eOIl/inen/al margins (caribbeml A nappes 01 the greater Anli/les alld VerrHuela. wu/em Colo m. bian cordil/era) or dippillg ullder tht cOllfinenlal margins (Frall­ dscall al Baja CalilOTllia). 6: Plio·Qu/ltemary "o/cullic ;:ones." mainl)' transmexican "o/canle axis and ,'oleanie cordillera 01 Cenlral America. 7-10 ." TtClOn ies: 7, II or"'a/laults. strille slip laults. Ir(JlJslorm laults; 8. Ihrllsling; 9. axial dipping ." A. Parr/l S trans"erse lOlle." B. Guatemala Iran .u·erse lUne: C. Bur'luisimeto Iran S"trse zone; 10. prHent d(l)' .tubdl/ction (Middle-America trell ch, Antilles trench). +• •

B. (lower diagwm) Se/lbeom batlrymelrk mU{J showi" J: delailed bathymllry UNlmd th .. Leg 67 siles.

.. ... _ ...... "" _ '.....0_'" I-...... ~ .. -I

B

227 J. AUBOUIN et s /.

i , consequence the trench appears as a succession of diamond ...... shaped ponds. where the graben cross Ihe trench. separated -:>:..:.... by flat swells representing the partially buried horsts . .. '. '.", }"y':;" The landward slope of the trench is perfectly rectili near ., '-. .. :~I.. :. '. _7<:: along almost 20 km. and perhaps bcyond the seabeam _...... survey ; il shows neither evidence of local or large slum p c. ,,- """"" • "'''''- ...... blocks. nor disturbances related 10 the oceanic plate and .. .. trench Slruçtures. The slope is di vided into rectilinear steps. c. '-~ which seem lîmiled by fault scarps. identified by the faci " """"'" l' the y do nOI respect the detailed lopography . --..._ ".. , • o---'!...'..:..... ,., ln conclusion. in the Leg 67 area. the COt:os plate displays • .. '.- its own typiCOlI Slructure . orie nted N 13()"140". as il enlers the trench, where Ihese structures determine a succession Figure 2 of oblique basins and swells : il di ves under the continental Summaf)' af seismic·uflrction und ufrocllan swdit.l s,If) ... l n ~ ,·e/Qâty strucwu und mu/ur reflrc/ors . Thr Iv,·(l/ion uf sitn dril/(d slope without disturbing il. It seems that here the Cocos by Clomar Chal/engu on L eI: 67 uu i"dicuted. plate essenlially fall s into the subduction zone beneath the American continental margin which is itself extensively faulted parallel 10 the trench. SEABEAM SUR VEY

DurÎng a transît from Panama to Acapulco, a Seabcam survey was made by the S.S. Jean Charcot in the Mid DR ILLING RESULTS Ameriça trench (Renard el al.. 1980). The survey consisted of a continuous corridor of detailed bathymetry wide Cocos Plaie runnÎng along the fool of the continental slope. Two maps including some sites of Leg 66 and Leg 67, of about 50 kml The oceanic Cocos Phlle forms the seaward slope of the and 100 kml respectively were made. The contour interv(11 Midd le America Trench off Guatemala (Fig. 1 A). Igneous was 10 III and mapping W3S donc in real time. using a system ocean basement is overluin by Il thin sediment blanketthat is of the Centre Oceanologique de Bretagne. uniformly 200-300 m thick. Site 495 is on a horst aboui The map of the Leg 67 area (Fig. 1 B) shows 3 system of 22 km seaward of the trench axis and 1.925 m above il. The alternating enlongated depressions and ridges on the Cocos unirormly thick sediment follows topography (excepl where plate. Due to the faCI that the scarps are reclilinear and faulted). a character commonly interpreted in seismic reflection records as that of ··pelagic sediment". cross the regional bathymetry. they can be interpreled as li horSI and graben IcclOnic structure. The corn mo n direction The sediment section penelrated al sile 495 records Ihe of these structures is N 130-N 140. parallel to the magnetic early and middle Mi ocene path of Ih is sile as it passed anomalies of the Cocos plate. This paralleJism suggests thal northward through the equatorial carbonate belt (Fig. 3) into this faulting originated along the grain on the plate nlong the a zone of normal slow pelagie deposition and then. during East Pacific ri se. as observed in the Cyamex and Ri se divÎng the late Miocene. inlo a hemipelagic environmenl influen­ cruises. and that it was reactivated when the plaIe was ced by a terrîgenous source ratheT than the tolally pelagic flexed as it entered the trench. This faulted structure is seen e nvironmenl inferred from se ismic reçords. Supcrimposed in the Irench 10 Ihe land wall toc of the trench. A" a on this record are the effects of subsidence a~ the ocean

I cocos Pl ATE TRENCH TR ENC H SLOPE - - - .' uc,-

1 .-;::Ir- l' 1P •.. _.: :. i ~, C i, ,'.- . Figure J Silllplifi~d hitl' und /i/hu-s/ ru/igruphy vf Siltl drillrd L on LrN 67. ; J" ~, ' 1 ..;;::, 1 - ,- ! i\ ',-<"~ •. - .- - 1 • - i! ;:I]'~! !--i'1 :i 1•-- '1 ..• ' •

228 lEG 67 ; MIO-AMERICA TRENCH (GUATEMALA) cruSI grew older and moved away from the spreading East Cretaceous 10 Quaternary (Fig. 3). The Upper Crelaceous Paciri c Rise. The upper hemipelagic section indicales a rock is about 900 m above the top of igneous oceanic crust surprising volume of sediment transporled across the Irench liS measured in se ismi c records (Fig.2). Il grelll distance seaward. If prese nt raies of convergence At si te 494 upper Miocene 10 Qualernary slope deposits have been conslant during deposition of Ihe hemipelagic cover a sediment seque nce similar 10 that reported by Seely seclion. the lowest beds were deposited when the site was (1979) from the edge of the continental shelL The environ­ aboul 900 km seaward of the trench. Such great dislances of ments represented by Ih is sequence, the nature of hi aluscs. sedimenl transporl arc not seen in Ihe present pattern of and their relation 10 the sequence of the shelf from young 10 hemipelagic sedimentation as recorded al other DSDP sites old are interpreted as follows, in descending order: in the area (van Andel et al.. 1973 : Hays et al .. 1972). 1) a Pliocent and Quaternary deposil consisting principally of sediment transported from the shel! and upper slope to Ihis site: Trench Floor 2) an upper Miocene unconformity coincident with the On Ihe trench floor fla! basinal areas are underJain by widespread upper Miocene unconformit y on the shelf ; horizontal seismic reflectors. These reflectors represe nt 3) a sequence of distal. terrigenous. hemipelagic clay that sediment ponded in graben of the Cocos plate thal are in accumulaled during the earl y Miocene on sea floor near the turn underlain by weak reflectors. The lower rdieciors dip carbonate compensation depth and at rates an order of gently landward and are presumed to be the deep ocean magnitude less Ihan. age-equivalent deposils on the adjll­ sediment seClion riding lnto the Irench axis on the Cocos cent shelf (3 mlMY versus 100 rn/MY) : Plate. 4) an upper Eocene to Oligucenc hi atus thal il! thc age Sites 499 and 500 drilled on Ihe Irench floor include seven equivalent or a widespread unconformity in the adjacent holes. most of which botlomed in basaIt or basait rubble al shelf section but of unknown origin in the section drilled : va rious depths (Fig. 3). Site 499. near Ihe seaward slope, is 5) an Eocene hemipelagic clay deposited below the forami­ located on the southern margin of Il smal1 diamond shaped nireral carbonale compensai ion depth but above the nanno­ basin in the trench. Site 500, at the base of the landward fossil carbonate compensai ion depth at rates at least fiv e slope, is located on the northern fl ank of a horst crossing the times less than age-equivalcnt shel f deposits : trench diagonally as shown by the Seabeam survey (Fig. 1 B). The trenc h fil1. as seen in seismic records. is 6) a hiatus of unknown origin but of an age Ihat is represenled by an upper Quaternary (slig hll y less than represenled by thick. widespread Paleocene sediments 4OO ,OOO-yr-old) sequence of alternating muddy and sandy below the adjacent shetf : turbidites wilh a microfossil assemblage transported (rom 7) an Upper Cretaceous clayslone that accumulated in an the shelf and slope. This turbidile sequence is underlain by a open occan cnvironment above the carbonate: compensation çomplete bul abbreviated oceanic sequence like the one depth CCD at depths about equivalent to, but al rates recovered at sile 495 on Ihe Cocos plate. AI si te 500 Ihis apparenlly less Ihan age-equivalent rocks below the present section is cut bya normal faulilha t probably formed during shel! : the development of the horst and graben structure of the 8) a contact of unknown origin bctween clayslone and trench seaward slope. The sediment benealh the trench igneous rocks; fl oor is compacted normally and shows few signs of 9) igneous rock originally of calkalknline basaltiç composi­ compressional dcformation even llgainst the trench lion llnd atypical of oceanic cruSI : the nature of body landward slope. sampled is unknown. This sequence shows microfracturing The drilling results indicate thal Ihe trench is more complex from tcclonic stress bcginning with rock at the base of the Ihan would be upected from the seismi c records. Basait Plcistocene. but there appear to be no Iluge thrust faults and basait rubble: occur al levels consistent with the cutting the dri11ed secti on. diagonal leg trending basait basement ridge seen in the At site 498 only the slope section (olive-gray mudstone) Seabeam balhymetry bul nol in the seismic records. T he appears tO have been penetrated. We inlerpret Ihe diffe­ drill holes appcar 10 be 2 km from the seismic line. a rcnce in thickncss of thc Miocene seclions of oli ve-gray distance that could llccount for thc llbsence of the ridge in mudstone as rdlecting relief on the upper Miocene disçon· seis mi c records. The trench filt has no pronounced lateral !ormity. dirfercnces in facies transverse to the Irench axis liS mighl be expected from axial turbid it y current channels (Piper el al .. 1973: von Huene, 1974). Middle Siope or Trench

Lower Slupe of Trc: nch Sites on thc trench of Ihe middle slope were positioned 10 penetrale the slope dcposils and 10 sample the underlying Benches o n the lower slopc of the trcnch. such as Ihe more landward dipping refleClors. We planned to test whethcr thc Ihan 20 km lone one on which siles 494 and 498 are localed rdleclOn are part of an imbricated stack resulling from (Fig. 1 A), are commonly intcrpreted as having been formed subduction, as has been inferrcd by Secly t l al.. (1974). by the emergence of a large thruSI fault at the seaflooL Such Ladd el al. (1978). and Ibrahim el al. (1979). Drill ing WllS Ihrusts cannol be seen in seismic refleclion records from the terminated short of the primary objectives for safety Ircnch off Guatemala: below Ihe fl oor o f the bench there is considerations because wc discovered gas hydrate. Siope a weak sequence of nearly hori zonl al renectors thal are deposits. as interpreted from seismic·reflection records. about 1,200m thick. At site 494, only 500m above Ihe uncon!ormably drape the rock that ma kcs up the bulk of the Ircnch floor. wc recovcred II seque nce of rock in normal conli nenlal margin (Fig. 2). The slope deposits thin toward slraligraphic succession thal ranges in age from Late the lowcr slope of the trench.

229 J . AU BOUIN el al.

Sile 497 was drilled in olive-gray mud wi lh varying biogenic indicate linle tectonism other than the horsl and graben of and vitric lu((aceous components (Fig. J). Il is in generala the ocean crus!. uniform seclion of mixed hemipelagic and lerrigenous mud On the landward slope of the trend, a cover of hemipelagic inlerspcrsed wilh a Ihin pcbbly mudSlone tha! represents sedimenl of middle Miocene 10 HoJocene age disconforma­ mass movemenL A similar litholgy was penetrated at bly drapes Ihe undcrlyi ng continental framework, whiçh is site 496 where the upper sequem;:e is biogenic mud of general1y obscured in seismic-rdlection records. Few Qualernary age underJain by biogenic sandy mudstone of DSDP holes have penetrated deeply beneath this disconfor­ Pliocene and Miocene age. Sediment recovered from Ihe mit y so commonly observed in seismic-refleçtÎon records mid-slope area contains volcani c ash beds. rich in lerrige­ across convergent margins. In seismÎc-reflection records. nous delritus_ and yielded a microfossil assemblage displa­ the discontinuity consists of reflections paralJeling Ihe ced rrom upslope. The foraminireral assemblage at sile 496 seafloor overlying landward dipping reflectors (see. for indicates subside nce of the sile in Mi oceoe lime trom shelf instance. Moore Cl al.. 1979). The slope deposils reveal to lower bathyal depths. This subsidence is coincident with morc smal1-scale deformnlion at depths of aboul200 m than li subsidence reporlcd by Seely (1979) from the adj,lcent was seen at site 500 almost al the !andward slope of the shelf edge. trench. Thus. in contrasl to the treneh floor, sedimenl draped over the slope shows sorne effects of tectonÎ sm. At sile 494. only J km landward from Ihe trench noor, we CONCL US IONS pcnetraled the slope deposils and round rock thal b surprisi ngly old and overconsolidated. The sedimenlary Geophysical data from the Middle America trench transect rock here was deposited in w31er shalJower Ihan Ihe o ff Guatemala contain nOlhing unusual when compared with carbonale compensation depth and has distal continenlal geophysical data (rom other convergent margins allhough affinities. From the shipboard studies we are unable 10 Ihe records do nOI resolve much structure al deplh along the ascertain Ihe original envÎronments of deposition except tO landward slope of the trench. as is seen. for instance. along infer thal Ihey are of a lower çontinental slope 10 shal10w the companion transect off Mexico (Moore et al .. 1979). ocean provenance. The Miocene section beneath site 494 The results from drilling off Guatemala also reveal nothing originated seaward of those deposils recovered from the unellpected in the post-middle Miocene sediment. The mid-slope at sites 496 and 497. The seclion at site 494 unupected results QCc ur in Ihc one drill hole that sam pied deposiled close enough 10 land . or entirely al laliludes north roc;k below the cover of slope deposits. rock that is much of the çarbonate zone of high produclivity. nOI to be olde r than that anticipated from previous gcophysical work. inundalcd by cHrbonale ooze. Perhaps anolher unexpeeted aspect is the seant evidence for compressional deformation. However. it is diHicult geophy­ sically to detecl SlruC lure thal is smaller than a few hu ndred TECTONIC INTERPRETATION melers in very deep waler. and structure is not well defined from the study of a n isolated drill core. An interpret3tion of our shi pboard data in the conte xl of The ocean cruSI entering the middle-Ameriça trench off previous geologie studies is difficull because Ihe data. Guatemala has a Mioçene secti on that dearly records particularl y thoses from sile 494. do not fit Ihe commonl y northward passage of the Cocos plate first bencath Ihe applied steady-slHte mode! of tcctonic accretion (Fig. 4) nor equalorial carbonate bel t and then into prollimity of an do any of the less often invoked models seem 10 apply upper Mioçene 10 Holocene terrigenous soun:c. The dis­ wit hout difficulty. Three end-member mode l ~ or explanB­ lançe from la nd where the lerrigcnous sediment fint rea­ ti ons. ail of which may apply in part. include the slumping of ded the area of si te 495 is problematical bUI is ce rtainl y .1 large block. ~ ubduc tion of sediment without accretion. or hundreds of kilometcrs away if plaIe recon~truclion~ for tCC lonic erO lo ion. this area are correct. As Ihe crust was flexed down inlO the trench. the crust was tensionally faulted inlo horst and graben that persist beneath Ihe Irench unlil hidden rrom . ,•• " ....'c. view under the landward slope of Ihe Irench. ,., .. " The diagonal trend of structures in the downflelled ocean crust is interpreted as reactivation of the planes of weaknes~ im p.lrted 10 the crust d uri ng generation along the Ea~t Pacifi e Ri se. Bath yme lric coverage is not sufficient 10 show the begi nning o f horst ilnd gm ben struclure. but beçause the overlying sediment cover does nol seem 10 be draped over fault ~c,l rpi>, the faulling mUS I have been relatively recenl. The Irench floor is d is tinguished from the seaward slope of the tre nch by Ihe ponded turbidiles. The ponded sediment conlains a microfauna leslo than 400.000 years old. attesting to Ihe youlh of the presenl trend floor. The youthful sediment and recenlteClonÎsm. together with active seismi­ cit y and voicanÎsm. are strong evidence for rapid conver­ gence of oceani c and continental crUSls. Nonelheless. the " u- .....-, . .-. fa Hure 10 find any increuse in compressional structure or ' ...... d ' even initial lectonic consolidation of Irench fill and Figure 4 underlying oceanic sediment at thc foot of the land ward Diu/;raru uf s"bs" rfaa Ke%g)' arron Ihe Middle·Amerie/, Tm"h slope is puzzling. ln facto the dril1ing and geophysical data off Gual/'mala haud 011 ,IIeoph)'sics alld Lep' 67 drilli"p'.

230 LEG 67 ; MID-AMERICA TRENCH (G UATEMALA)

From considerations of sedimentary facies. a slump thal ces of Central America. instead of a hypothetic terrane displaced the large bench on which sile 494 was drilJed must rafted on the ocean plate thal was incorporated into the have originated in an environment seaward of the shelf edge continental framework during some ancient collision. (Seely. 1979) and of sit es 496 and 497. A pre-Pliocene age So, if we accept the concept of sediment subduction. we seems to be required by the blanket of Miocene to Quater­ must accept that there is no accretion at the base of Ihe nary slope deposits . which has no slump scars as large as the continental margin of Middle America trench. Some sub­ proposed slump block . and also by the mid-Miocene discon­ ducti on of sediment is required 10 explain the lack of tinuity between slope deposits and the unexpected site 494 accretion in the site 494 area. however sediment subduction section. Although a slump block would a lJ eviate the proble­ was probably not the only process involved. If a block from matieal absence of compressional structure at the foot of the ocean plate was incorporated just prior to the Miocene. the slope and provide a means of transporting an older ail the post Miocenc sediment would have to be subducted accreted section from upslope to the prese nt trench flour. rather than accreted ; if the pre-Neogene section al 494 was the Seal>eam survey shows no scar nor does the perfectly always part of the margin. ail post Cretaceous sediment linellr and more than 20 km long base of the slope have the would have been subducted. Periodic sediment subduction characteristie lumpy morphology of II slump. Slumping does seems reasonable but contÎnuous sediment subduction since not e xplain the absence of accreti on in the Pliocene and the Cretaceous does not. Thus there must be another Quaternary and the slump expia nation requires considerable explanalion for the Jack of accreti on since al r ea .~ 1 Miocene subducti on of sediment. and probably Cretaceous time. The com;ept of sediment subduction since Miocene time Tectoni c erosion is a third concept that might be applied to impl ies disposai of a great amo unt of sediment somewhere explain thc drill data. Because the vector of plate conver­ down the Be nioff zone since the present convergence began gence is normal 10 the trench axis, linle transport in a laieraI (early to mid-Miocene). Perhaps the rather surprising sec­ directi on scems appropriate, a nd tectonic erosion would tion at site 494 is an anomalous accreted block brought in on require subduction of continental fragments, a difficult the ocean plate , but as in the slumping model the block concept. Tec tonic erosion however is an efficient way to would have been accreted in the pre-middle Miocene. truncate the base of the slope despite the conceptionaJ Il is possible that Ihe pre-Miocene sequence at site 494 problems of stuffing roc k of less density beneath that of originated close to the continental margin, of Guatemala. greater density or the abrasion of the conlinenlal framework beginning in the Late Cretaceous. This sequence was on a massive scale. This mechanism has been applied to deposited contemporaneous with the top of the Nicoya e xplain the Neogene Iruncation of the converge nt margin complex that crops out on the continental shelf of Costa oH northern Ho nshu (von Huene el al.. 1980). Ri ca. ln the Santa Elena peninsul a, peridotites arc ovcrlain As problematieal as these explanlltions may seem , they are by a sequence beginning exactly at the same lime (Upper briefly discussed here to provide sorne focus for subsequent Campanian-Maestrichtian) as the sequence above igneous sludy. ft seems paradoxical to Cind more evide nce for a rocks at s it e 494. Because an Upper Cretaceous sequence of passive structural 5eUing than for active tectonism along a sediment is a general feature on the Pacifi c side of Central margin with such high rates of plate convergence. but this America, it seems more real istic to associate the pre-Mio­ paradox might be a maller of the depths of drilling and the cene sequence at site 494 with these late Cretaceous sequcn- resoJulion of geophysical data .

REFERENCES

Auboul n J., Tardy M., 1980. L'Amérique alpi ne; le domaine Symp. Guatemala Februury 4 Earlhquuke and RI!ConSIrUt'fitm Pro­ caraïbe et ses liaisons avec les cordillières nord ct sud-américaines; eus. Gualtmala Ciry. May 1978. in troduction . 16" l'11er. Geo/. Congr .. Paris. C"Uoq. C5. 14 -17. Mlnster J. B., Jordan T. H. , 1978. Pre sen t-day plate motions. J. Aubouln J., von Hoene R., Azéma J., Coullluuro W. T., GeophY5 . Res .• 83, 5331-5B4. Cowan O. S.. Curiale J. A. , l>engo C_ A.. Failli R. W., lIarr!· :;on W., Ife»" H., Ladd J. W. , Muzllev N., Shikl T .. Thomp­ Molnar P., Slkes L. R., 1969. Tec tonics of thc Caribbean and SO n P. R., Weslbe rg J., 1979. Premiers rés ult:IIS des forages pro­ Middle Amenca reglons from focal meehan isms and sd smi eity. fond s dans le Pacifique au niveau de la fosse du Gu aTemala (fosse Geol. Soc. Am. Bull. . 80, 1639-1684 . d'Amérique ce ntrale), C. R. A cad. Sei .. Puris. sir. d. 28\1, 1215-1220. Moore J . C .. Watklru; J .. üllchman S. H., Heghlcl F. W.. Hull A.. Dldyk B. M., Le~ n J .. Lu nd berg N., MeM ilien K. J ., NlIIsu­ Auma J .. Tournon J .• 191:10. La péninsul e de Santa Elena . Costa ma N., Shtpard LE., Shlpley T. H.• Stephan J. F .. Stradner H. , Rica; un ma ss if ultrabasique charrié en marge pa cifique de 1979. Thc Middle America Treneh off Mex ico, GeOlimes. M. 9. l'Amérique centmle. C. R. Acad. Sei., Puris, sir. D, lllO, 9-12. 20-22. Hays J. S., Coo k lU Il. E.. Jenkyns D. G .. Cuuk .-. M., FullerJ. T .. Goll R. M., Mllow E. O. , Orr W, N. , 1972. Initill/ Reports of titI' Piper D. J. W., ~on Huene R. E., Duncan J. R.. 1973. Late Quatcr­ Deep Sea Dri/ling Projeel. Wa5hinHlon. US GD~. Prim. Off., 9, n:lfY sedimentation in the active Eastern Ale utian Trench. Gco/Qg)·. 1 20S p. l , 1. 19-22. Ibrahim A. K. , Latham G. V., l ..dd J ., 1979. Seismi c rdraction Renard V., Auboul n J .. Lonsdale P.. Stephan J. F.. 1980. Premicrs and refieCTion measuremenlS in th e Middl e America Trench résult ats d'une étude de la fosse d'Amérique ce nt rale au so ndeur offshore Guatemala . J. GeOphY5. Res .. 84, BIO . 5643 -5649. m ultifai seea u~ (Seabeam). C. R. Acad. Sâ .. Pli riS. sà. D. 29 t . Jord an T.II .. 1975. The prese nt day motions of the Caribbean 137-142 . Res .• Pla te. J. Geophys. 80. 32, 44 33-4439. Secly D. R.. 1979. Geol'hysicili inves ti gll TiDn ~ of continental ~ I ope s I... dd J. W.• Ibrahim A. K.• McMlllcn K. J .. La tham G. V.. ~o n and ri ses. in : Geological ilnd geol'hys icil l investigation of co ntin en­ Itoene H. E., Watklns J. E., Moore J. C., Worul J. L.. 1978. Te c­ tal margins. Tulsa. Oklahoma. edited by J . S. Wat ki ns and L. Mon­ tonics of the Middle Ameri ca Trench offsllOre Guatemala. I nlu. tadert. Am. AssDe. Perm/. Geai. Mem .. 51, Tulsa.

23. J. AUBOUIN et al.

SHly D. R .• Va il P. R .• Walton G. G .• 1974. Tr~neh slope mode!. af cOlttinelllal margin:;. ediled by C. A. Burk Olnd C. L. Drake. in; Gwlogy of e/mlinl'nwl marllins. editcd by C. A. BlJrk and Springer-Verlag. New York. 1009 p. C. L. Drake . Springcr·Vcrlag. New York. 261-283. "on Huene R .. Auboui n J., An'ma J., Blackln lOn G .. Carter J. '\ .. o'an Andt.! T. H., Heath G. R., Iknnel R. H., lJubry J . O., Chartes· Coulboorn W. T., Co ... an 0 .5., Cur iale J . A.. De ngo C. A.. FallS tlm S., Cronan O. S., OI nkelman M. G. , Kaneps A. G .. KeMn S. R., R. W., Harriso n W. , Hesse R .. Hussong D. 1\1 .. Laad J . W., Muzl. Yeats R. S .• 1973. fnirial Reports of rhe Deep Sea D,ilJjn~ Projre/. lu N .. Shl kl T., Thompson P. R .. Westberg J ., 1980. Leg 67. The Washington. US Go .... Print. Off.. 16, 949 p. Deep Sea Drilling Project Middle America Trench transeet off ,'on Huent R. E .• 1974. Modern Ircneh sediments. in : The Geu/Qgy GUlltemllla. Geu!, Soc. Am. Bull .. III , 421-432.

232

"