JOURNAL OF GEOPHYSICAL RESEARCH,VOL. 103,NO. B2, PAGES2597-2614, FEBRUARY 10, 1998

Fluid venting in the eastern Aleutian subductionzone

Erwin Suess,Gerhard Bohrmann, Roland von Huene, Peter Linke, KlausWallmann, Stephan Lammers, and Heiko Sahling

GEOMAR, ResearchCenter for Marine Geosciences,Kiel, Germany

Gisela Winckler

Institutftir Umweltphysikder Universit•itHeidelberg, Heidelberg, Germany

Richard A. Lutz

Centrefor Deep-SeaEcology and Biotechnology, Institute of Marine andCoastal Sciences RutgersUniversity, New Brunswick,New Jersey

Daniel Orange

MontereyBay AquariumResearch Institute, Moss Landing, California

Abstract.Fluid venting has been observed along 800 km of theAlaska convergent margin. The fluid ventingsites are located near the deformation front, are controlled by subsurface structures,and exhibit the characteristics of coldseeps seen in otherconvergent margins. The moreimportant characteristics include (1) methaneplumes in thelower water column with maximaabove the seafloor which are traceable to theinitial deformation ridges; (2) prolific coloniesof ventbiota aligned and distributed in patchescontrolled by faultscarps, over- steepenedfolds or outcropsof beddingplanes; (3) calciumcarbonate and barite precipitates at thesurface and subsurface of vents;and (4) carbonisotope evidence from tissue and skeletal hardparts of biota,as well asfrom carbonate precipitates, that vents expel either methane- or sulfide-dominatedfluids. A biogeochemicalapproach toward estimating fluid flow ratesfrom individualvents based on oxygenflux measurementsand vent fluid analysisindicates a mean valueof 5.5+ 0.7L m-2 d -1 fortectonics-induced water flow [ Wallmannet al., 1997b].A geophysicalestimate of dewateringfrom the samearea [von Huene et al., 1997]based on sedimentporosity reduction shows a fluid loss of 0.02L m-2 d-1 for a 5.5km wide converged segmentnear the deformationfront. Our video-guidedsurveys have documented vent biota acrossa minimumof 0.1% of the areaof theconvergent segment off KodiakIsland; hence an averagerate of 0.006 L m-2 d -1 isestimated from the biogeochemical approach. The two estimatesfor tectonics-inducedwater flow fromthe accretionary prism are in surprisingly goodagreement.

1. Introduction The circum-Pacificsubduction zones manifest a variety of end-membertectonic settings, studies of which have now and in Fluid venting along the world's subductionzones has been the past contributedtoward an in depthunderstanding of the recognized over the past 10 years as a processof first-order complex processat convergentmargins [Kulm et al., 1986; Le importancefor marinegeosciences and oceansciences [Langseth Pichonet al., 1987; vonHuene and Scholl,1991, 1993; Kastneret andMoore, 1990;Moore and Vrolijk, 1992 ]. Ventingaffects the al., 1991; Carsonet al., 1994; Westbrooket al., 1995; McAdooet budgetsof certainelements in the deepsea [Suessand Whiticar, al., 1996]. Critical regionsfor fluid escapeare ,defor- 1989; Martin et al., 1991, 1996], the material turnover at mationfronts, and initial accretionaryridges. Accreted and sub- specializedvent ecosystems[Suess et al., 1985; Brooks et al., ductedsediments are thoughtto be separatedby interfaceswith 1987; Rio et al., 1992; Childresset al., 1986; Bouldgueet al., low shearstrength and with concentrationsof overpressured pore 1987] as well as the thermalstructure of accretionarycomplexes fluids.This interface decouples the sedimentary sequences during [Le Pichonet al., 1990; Henry et al., 1992, 1996;Hyndman et al., convergenceallowing unconsolidatedsediment to be subducted 1993]. Fluid flow andpressure gradients may in turn influencethe beneaththe margin. Graduallynow, the complexityof these accretionarytectonics such as activity or multiplexing submarinehydrogeologic processes is becomingapparent. So far [Davis et al., 1990; Sammondset al., 1992; Brown et al., 1994]. therehas been evidence reported for outputof freshand super- saline water from accretionaryprisms [Kastner et al., 1991; Wallmann et al., 1997a] and for horizontal and vertical Copyright1998 by theAmerican Geophysical Union. recirculationover considerabledistances through sequences of Papernumber 97JB02131. accretedsediments [Le Pichon et al., 1990; Martin et al., 1995]. 0148-0227/98/97JB-02131 $09.00 An enormous difference of flow rates, however, has been

2597 2598 SUESS ET AL.: FLUID VENTS IN ALEUTIAN ZONE estimatedwith geophysicaland geochemical methods at different Baranoff Fan, is locatedin the southeasternmost part of the Gulf convergencesettings with no clear pictureemerging [Carson et of and began to form in upper Miocene time. The two al., 1990;Linke et al., 1994;Henry et al., 1992, 1996]. older fans have presumably contributed material to the Severalorders of magnitudehave separatedflow estimates accretionaryprisms along the easternAleutian ; however, basedon porosityreduction from those observed directly at vents. most of their sedimentvolume is currently being subducted,the Furthermore,the questionof the relative importanceof focused process which generates the fluids being expelled at the flow as evident at vent fields versus diffuse flow without convergentplate boundary [von Huene and Scholl,1991 ]. conspicuousvent communitiesor chimneysremains unresolved. The westernpart of the Yakutat Block has been subductedas To help clarify this situation, we report here the first shownby magneticanomalies. Assuming it has been coupledto comprehensivedata set as well as hithertounknown evidence for the placespart of the terraneat the baseof the slope, tectonicallycontrolled, large-scale venting phenomena in the deep within the northeastern survey area between 3 and 5 Ma. eastern Aleutian Trench of the Alaska subduction zone. The Subsequently,its point of entry into the subductionzone swept surveyscarded out and samplescollected by R/V Sonnein 1994 northeastward along the trench to its present position in the and 1996 composean 800 km long segmentbetween the Kodiak northern Gulf of Alaska. Thus the tectonic history of the and ShumaginIslands. Here we found manifestationsof fluid accretionary domains in our survey area includes a former ventingin the form of distinctivefaunas, mineral precipitates, collisionalsegment currently in an areareceiving high amountsof methaneanomalies, a temperatureanomaly, in situflow data,and interglacialsediment, a noncollisionalsegment receiving Surveyor contrastingchemistries of porefluids and sediments from on-vent Fan and trench sedimentloaded with glacial debris,and a segment and off-vent settings. where the head of the Eocene Zodiac Fan and the older oceanic The temperatureanomaly is small, yet significant, and crusthave enteredthe subductionzone. In the westernsurvey area documentsan extra heat sourceto the oceanicbottom waters from the thicknessof trench sedimentsis significantly less than in the below the seafloor,thus giving a new meaningto the conceptof easternarea. cold seeps.Overall, the significanceof our discoveryin the Four segmentsof the marginwere investigated during R/V AleutianTrench is seenin thepredictability and documentation of Sonnecruises (Figure 1, SO-96,SO-97, andSO- 110 [ Fliihand vent sites within specificdeformational settings of that vonHuene, 1994;Suess, 1994; Suess and Bohrmann, 1997].All accretionarymargin. Our successin finding theseactive zones of stationsoccupied and surveysconducted during these cruises and fluid escape provides renewed confidence in being able to referred to in this communication are listed in Table 1. Extensive extrapolateand eventuallyto quantifytectonic dewatering within seismicreflection data [von Huene et al., 1987; von Huene, 1989] the entireglobal subduction framework. were mergedwith the high-resolutionswath bathymetry obtained during the R/V Sonnecruises to locate the accretionaryridges which became the focus of our bottom surveys, sampling, and 2. GeologicSetting fluid monitoring. The Edge sector includes an accretionary mass that was The continental margin that borders the eastern Aleutian probablybuilt againstthe erosionalscar formed during collision of Trenchhas an accretionaryterrane that containslithologies as old the Yakutat Block. Here the trenchaxis receivesa large amountof as Late Cretaceous.The Kenai Peninsula, the Shumagin and terrigenousglacial sedimentfrom the adjacentAlaskan mainland Kodiak Islands,and presumablythe shelf betweenthem have an cappedby an interglacialsequence (Deep SeaDrilling ProjectSite upper plate crust consistingof turbiditesand volcanicrocks of 180) [Kulrn and von Huene, 1973]. In the Edge sector the Cretaceousto Paleogeneage. The outershelf and slopeseaward of youngest tectonic structures,forming the deformation front, the islands are characterizedby Eocene to Oligocene accreted consistof two relatively gentlefolds with their asymmetricover- rocksoverlain by Neogenebasins [Moore et al., 1991]. Sediment steepenedflanks facing the trench(Figure 2). Thesestructures are accretedto the continentat the trenchduring the currentepisode is parallelto and situatedjust at the deformationfront in about5000 generallyyounger than 3 Ma. m of water depth. Toward the southwestthey terminateagainst a The oceanic Pacific plate that is subducted near the steepscar, believed to be the lateralstrike-slip fault of a subducted northeasternend of the Aleutian Trench is of Eocene age and seamounttrace. The folds exposetrench fill sediments;their relief increasesin ageto the southwest.The plateconvergence rates are reachesabout 300 m above the trench floor; and their steepened around5.5 cm/yr [DeMets et al., 1990]. The Aleutian Trench is flanks exhibit many slumps visible in the high-resolution generallythe boundaryof the NorthAmerican plate, but below the swathmapping(Figure 2). The escarpmentsexpose gently land- eastern Gulf of Alaska the boundary is complicated by the ward dipping strata.At the baseof theseescarpments, but also presenceof the YakutatBlock, whichis currentlycolliding with higher up at intermediate stepsin the morphology,we found Alaska [Bruns, 1983]. coloniesof vent biota and othermanifestations of fluid escape. Accordingto one plate tectonicreconstruction of the north In the Albatross sector the deformation front consists of a Pacific region, the Yakutat Block resided off Washingtonor growinganticlinal fold on the 5000 m deeptrench axial floor with Oregonduring the Oligoceneand subsequentlymoved northward a relief of up to 400 m. Sedimentin the trenchis probablydistal [Bruns, 1983]. During this transit the adjacent oceanic crust glacialmaterial channeled from southernKodiak Island by a well- received sediment from the North American continent which was definedglacial troughextending from the islandto the upperslope laid downin deep-seafans [Stevenson and Ernbley,1987]. Two of above the surveyedsector. Where it is crossedby seismiclines, thesefans encompass areas larger than or equalto today'sAmazon the structureincludes blind backthrustswith a landward verging Fan. The oldest, the Zodiac Fan in the west of the survey area, fold [von Huene, 1989; von Huene et al., 1997]. The steeperslope receivedhemipelagic sediment during 42 to 24 Ma but is now facing the trenchis also sculptedby slumps,but generally,they coveredby a pelagic sedimentblanket. The secondoldest, the are not as numerousor as extensivelydeveloped as thosein the Surveyor Fan in the east of the survey area, received sediment Edge sector.The failed slopesexpose strata horizontal in the strike duringthe periodfrom about20 Ma to the present.The upperpart directionalong which fluidsescape, apparently rapidly enough to of its clastic material is glacially derived. The youngest,the supportthe vent coloniesobserved here. SUESS ET AL.: FLUID VENTS IN 2599

200' 205' 210' 215' 220' 225' 230' 62' 62'

60' 60'

ßBLOCK ß

EDGE

.. 58' 58' . .. km/Ma

...... PACIFIC PLATE

56' 56' HUMAGIN BAT ROSS

SURVEYOR FAN 54' 54'

o ZODIAKFAN (•(• ø BARANOFFFAN

200' 205' 210' 215' 220' 225' 230'

Figure 1. Gulf of Alaska with major plate boundaries,fan depositsand the areasof investigationduring R/V Sonnecruises SO-96, SO-97 and SO-110. The areasare referredto in the text as the Edge, Albatross,Shumagin, andthe Ugak sectors;white squares show surveys in Edge(Figure 2) andShumagin sectors (Figure 3).

The Shumaginsector, where the trenchaxis is 6000 m deep, difficult.There is a greatdisparity in spacescales between what receives much of its axial fill by lateral transport from the canbe identifiedgeophysically as a deformationfront or the likely northeast.The folds at the deformationfront showlittle evidence projectionto the seafloorof possiblefluid pathwayssuch as faults of slumpingand are discontinuous along strike. Owing to themore and the much smallerdimensions of actual vents.In seismic obliquesubduction in this part of the surveyarea, the lateral sections,faults are of the orderof manyhundreds of meterslong componenton convergencebecomes larger than to the northeast or foldshave hundredsof metersof relief, but activevent sites andhence a patternof shearfaults develops diagonal to the axial coveronly a few squaremeters; perhaps aligned vents are a few trend [Lewis et al, 1988]. Little evidencefor venting was found hundredmeters long. At 5000 m depth,hydrosweep soundings are alongthe folds at thedeformation front; instead, more active vents in a 50 x 500 m grid, and seismictraces are 50 m apart. and extensivecarbonate crusts were observedalong the crestal Bathymetricdepth resolution is 20-30 m at best,seismic resolution regionsof ridgesfarther upslope, where they were cut by canyons 40-60 m at best.The widthof theTV surveyvision is about12 m. (Figure3). The ventbiota colonies were similarto thosein the It mighttake four to five transectsto completelyimage a potential Edgeand Albatross sectors. ventsite in seismicrecords or bathymetry.We havesearched for In the Ugak sector,located about half way betweenEdge and activevent areasby mappinganomalies of dissolvedmethane in Albatrossand within the areareceiving Surveyor Fan sediments, the near-bottomwater columnand by conductingvideo surveys. the seaward flanks of the first three anticlinal ridges were Chemical anomaly mapping detectsvents which emit methane investigated.Vent biotawere found along the seawardface of the assumingthat the backgroundCH4 contentand other nonvent third ridge and the summitof the first ridge.A largeclam field, sourcesof CI-[ areknown [Lammers et al., 1995].Furthermore, it about40 m across,was found at 4880 m nearthe ruggedseaward assumesthat CI-[ andother reduced chemical species, dominantly facingflank of the third ridge.This areawas dominatedby scarps H2S, maintain colonies of characteristicvent biota as well as (2-6 m high) and outcropsof landwarddipping strata. The first createthe environmentconducive to precipitationof crusts,con- ridge,of relativelysmooth morphology, showed a largeclam field cretions,or chimneys. right at the top (5000 m). Evidence for fluid venting was not Duringthe searchfor ventsin the threesectors off Alaska,we discoveredalong the secondridge; however,only one crossing used a combinationof TV surveys,near-bottom temperature with the towed TV system was carried out, and no distinct recordings,and methanemonitoring of the lower water column. structuralfeatures could be associatedwith the active vents. The TV sledEXPLOS (OceanFloor ExplorationSystem)aboard Nevertheless,the discoveriesat Ugak confirmedthat fluids are R/V Sonneis equippedwith a black-and-whiteOsprey video expelledalong the entirelength of the convergencezone surveyed. camera,two Photoseastill cameras,three halogenlamps, and a conductivity-temperature-depth(CTD) System (SIS 6000). All 3. Finding and CharacterizingActive Vent Sites data andimages are continuouslydisplayed in real time aboardthe vessel. The system is navigated with a Super-Short-Baseline Locatingvents at subductionzones without direct observations (SSBL HPR-1507) by Simrad and a respondermounted to the from submersiblesor remotely operated vehicles (ROVs) is sled. The instrumentis towed at approximately0.5 to 1 knot, 2600 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

Table 1. List of StationsSurveyed in the EasternAleutian Subduction Zone DuringR/V SonneCruises SO-96, SO-97, and SO- 110

Station Latitude Longitude Water depth Instrument ID øN øW m

Sonne Cruise 96-2 96-2 57039.55'-57036.16' 147054.20 '- 147ø52.85 ' 3763-4782 EXPLOS 96-3 57ø34.09'-57 ø31.39' 148 ø04.30 '- 148003.26 ' 3847-4691 EXPLOS 96-10 55058.51 '-55ø52.90 ' 152ø38.05 '- 152029.45 ' 3995-5220 EXPLOS

Sonne Cruise 97-1 97-11-1 57040.85 ' 148ø08.47 ' 3286 CTD 97-11-2 57040.76 ' 148ø08.76 ' 3244 CTD 97-19 57032.08 ' 148007.92 ' 4472 CTD 97-20 57026.09 ' 148ø01.14' 4946 CTD 97-21 57029.48 '-57 ø25.89 ' 148 ø03.0 '- 148000.98 ' 4681-4997 EXPLOS 97-22 57ø26.97'-57ø26.97' 148001.21 '- 148ø01.27' 4776-4949 TVG 97-23-1 57025.98' 148001.20' 4969 CTD 97-23-2 57025.99 ' 148001.41' 4978 CTD 97-29-2 57026.55 '- 57025.54 ' 148002.28 '- 148ø00.74 ' 4768-4981 VESP 97-30 57022.99 ' 148ø22.05 ' 4978 CTD 97-35 57029.59 '- 57026.29 ' 148001.04'- 147ø50.39 ' 4527-4979 EXPLOS 97-37 57026.70 ' 148 ø 12.20' 4828 CTD 97-38-2 57027.34'-57026.98 ' 147ø59.81'-147ø59.81' 4924-4974 VESP 97-44 55056.58 ' 152ø00.76 ' 5376 CTD 97-49-1 56005.81' 151044.86' 5379 CTD 97-49-2 56005.92 ' 151 o45.13' 4825 CTD 97-51 56008.60 ' 152ø20.10' 4204 CTD 97-66 57ø27.33'- 57026.79 ' 148ø00.04'- 147059.98 ' 4850-4967 TVG 97-72 57027.68 ' 148000.87 ' 47 44 MUC 97-77 54 o10.81 '-54007.34' 157009.15'- 157006.11' 5562-5872 EXPLOS 97-78 54015.07'-54013.53 ' 157011.45'-157ø09.58 ' 5065-5548 EXPLOS 97-79 54ø 17.73' 157 ø 14.14' 4736 CTD 97-82 54 ø18.61 '-54 o17.57' 157 ø13.82'- 157009.97 ' 4546-4896 EXPLOS 97-88 54 ø16.81' 157 ø14.09' 4776 MUC 97-90 54ø18.86'-54ø17.75 ' 157 ø 11.69'- 157010.46 ' 4715-4939 EXPLOS 97-95 54ø06.30' 157ø22.86 ' 5946 CTD 97-97 45ø18.23'-54ø17.98 ' 157012.75 '- 157011.71' 4619-4861 TVG 97-103-1 54ø 18.10' 157 ø11.38' 4880 CTD 97-103-2 54ø17.71 ' 157ø11.80' 4843 CTD 97-104 54 o18.29'-54 o17.86' 157 ø12.46'- 157 ø11.74' 4662-4892 VESP

Sonne Cruise 110-lb 110-23 57o27.32'-57o26.53 ' 147ø59.73 '- 148ø01.32' 4947-4919 ROPOS 110-27 57ø52.21' 148ø08.01' 4958 CTD 110-35 56ø40.42'-56ø36.71 ' 150o10.58'-50o07.58 ' 4844-5279 EXPLOS 110-39 54 ø18.28' 157ø11.53' 4952 CTD

'•CTD,standard CTD with rosette;EXPLOS, towedvideo-guided survey system; TVG, video-guidedgrab sampler; VESP, video-guidedvent samplerwith benthicchamber; MUC, multicorer;and ROPOS, remotelyoperated vehicle preferablydownslope. The height of 1-6 m abovethe bottom is 4900 m. Of thesesurveys four are referredto here in more detail maintainedby an operator continually adjustingthe length of (stations96-2, 96-3, 97-21, and 97-35). Four surveyswere located winch cable. The sled-mountedCTD records any temperature in the Albatross sector, and data from one of these are used here anomalies.Additionally, hydrocastswith conventionalCTD and (station96-10). Seven surveyswere run in the Shumaginsector, rosettesampler obtained water samplesfor CH4determinations at four of which are dealt with here in more detail (stations 97-77, discreteand closely spaceddepth intervals [Suess, 1994; Suess 97-78, 97-80, 97-82). One survey was run in the Ugak sector and Bohrmann, 1997]. (station 110-35). Extensive venting was shown at Edge and Multibeambathymetry, processed and displayed by the Hydro Albatross at only the first and seconddeformation ridges, at Map System(HMS 300) on boardR/V Sonneprovided updated Shumaginat the third ridge upslope,and at Ugak at the third and backgroundinformation for sitingof the near-bottomhydrocasts as well as the TV sled transect and, where appropriate, for the first ridges. Here we documentand discussthe accumulated deploymentof a TV-guidedgrab and ventfluid sampler(VESP). evidencefor fluid expulsionbased on four criteria:CH 4 anomaly, The depthrange of individualTV surveys,each with 3-4 hoursof potentialtemperature anomaly, benthic colonies,and inorganic bottomtime, coveredup to 1000 m verticallyand of the orderof chemicalprecipitates. Then we considerestimates of flow rates 10 km horizontally, depending on the bottom morphology. and showthat bioirrigationis a major componentof flow emitted Successivetracks were used to providea compositedepth transect. at the seafloor and that true tectonics-induced flow is more than an For the Edge sector,a total of 13 surveyswere conductedwhich orderof magnitudesmaller, though it agreeswell with indepen- coveredthe Neogeneaccretionary prism at depthsfrom 2500 to dentestimates based on a geophysicalapproach. SUESS ET AL.' FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2601

-148 ø 04' -148 ø 02' -148 ø 00' -147 ø 58' 57 ø 30' 57 ø 30'

97-21 97-35 EXPLOS

57" 29' 57 ø 29'

57 ø 28' 57 ø 28'

97-72 MUC

97-66 TVG

57 ø 27' •97-22 TVG 57 ø 27'

97-20C•D'D'D• 57" 26' )7-23-1, -2 CTD 57" 26'

5.5 cm/a 57 ø 25' 57 ø 25' -148 ø 04' -148 ø 02' -148 ø 00' -147" 58'

Figure 2. Locationmap of the vent sitesfrom the Edge sectorwith 20 m depthcontours. The bold segments along the tracksof the surveys(stations 97-21, and 97-35) by the OceanFloor ExplorationSystem (EXPLOS) indicateactive venting.CTD stations97-23-1 and 97-23-2 showmethane anomalies; coring stations97-66 (TV- guidedbox corer,TVG) and97-72 (multicorer,MUC), from whichpore waters were extracted, represent on-vent and off-vent sites,respectively. The outlineof the surveydoes not correspondin detail to the areaof investigation shownin Figure 1.

3.1. Methane launch TV surveys and subsequentsampling of small-scale targets.Frequently, we foundevidence for a plume-shapedCH 4 The methane distribution in the water column shows a near- distributionin the lower water columnwith methaneinjection surfacemaximum with decreasingconcentrations, including some presumedto be lateral at someshort distance away from the cast. fine structure,over the approximatedepth interval of the oxygen The plume patternis best illustratedby data from an array of minimumzone. This patternreflects production of methaneduring hydrocastsin the Edgesector (Figures 4a and4b). A downslope zooplanktongrazing and biogeochemicalcycling during particle profileof four CTD castsshows oceanic CH 4 background to about decomposition[de Angelis and Lee, 1994; Tilbrook and Karl, 3000 m. Between 3000 m and the seafloor,just off the third 1995]. Below about 1000 m the CH 4 contentsstay close to the deformationridge, a well-developedplume was observed (stations detection limit and remain low throughout the deeper water 97-11-1and 97-11-2 ). Fartherdownslope at station97-19, no CH 4 column [ Scranton and Brewer, 1978; Tilbrook and Karl, 1995]. input was detected,but CH 4 contentsincreased dramatically This relativelysimple pattern is well knownand documented in all toward the scar of a subductedseamount (stations 97-30 and 97- castsover the Aleutian Trench (Figures 4a-4d). However, addi- 37). Within the embaymentformed by the collision of the tional CH4input from cold seepsprovides a strong signal for seamountwith the margin, the highestCH 4 contentsanywhere locatingdewatering and degassingsites in subductionzones. CH 4 overthe entireAleutian Trench were detected.Unfortunately, not monitoring of the near-bottom water column at the Edge, enoughtime wasavailable during the R/V Sonnecruises to inspect Albatross,and Shumaginsectors yielded repeatedand consistent this area more closelywith video surveysor to conducthigh- anomaly patterns from which to deduce active venting and to resolutionsampling for methane.From the magnitudeof the 2602 SUESS ET AL.' FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

157 ø 14'W 157 ø 12'W 157 ø 10'W 54 ø 97-90 EXPLOS 19'N

97-82 EXPLOS

110-39 4860 m

97-103-1 97-97 TVG o 54 ø 18'N 97-103-2 CTD

54 ø 30'N

4900 m 54 ø eutian 20'N Terrace 54 ø 17'N

54 ø 10'N Aleutian Trench

170 ø 170 ø 20'W 00'W

Figure 3. Locationmap of the ventsites from the Shumaginsector with 2 m depthcontours. The bold segments alongthe EXPLOS tracks (stations 97-82, and 97-90) indicate active venting. CTD stations97-103-1, 97-103-2, and 110-39 showmethane anomalies and the coringstation 97-97 (TV-guidedbox corer,TVG) yieldedcarbonate precipitates.The surveyis locatedon the Aleutian Terrace (inset); the map outlines do not correspond in detail to the areaof investigationshown in Figure 1. anomaly,we wouldexpect a CH4 sourceof considerablestrength and Leyk, 1995]. The slopewardprojection of the maximumCH 4 somewherevery near by. RepeatedCTD castsover the trench, just value (Figure 4b) coincides with the ridge crest depths,where seawardof the initial deformationridges, also showeda well- chemosyntheticcommunities indicate venting. defined,near-bottom CH4maximum (Figure 4b, stations97-20, In the Albatross sector, three casts, two at the lower slope 97-23-1, and 97-23-2). The sameplume was foundagain 2 years (stations 97-44, 97-49-1, and 97-49-2) and the other on the mid later duringthe recentlycompleted cruise SO-110 (station110- slope(station 97-51), showedpositive CH 4 anomalies (Figure 4c). 27), althoughthe concentration was somewhat diminished. The site of the CH 4 anomaly at the deepestlocation, although Generally,the CH4 plumesare confinedto the lower water rather weak, is populated by prolific bivalve colonies at the column. The anomaliesof between50 and 150 nL/L are not quite trench-facingflank of a growinganticlinal fold. Again,this obser- ashigh as those at mid-oceanridges [Horibe et al., 1986; Lilleyet vationis clearevidence for structurallycontrolled dewatering. The al., 1993];however, they are significantlyhigher than the oceanic CH 4 patternfound at the mid slope station,with the maximum background.A characteristicfeature appearsto be that the concentrationimmediately above the bottom, suggestsventing lowermostsample immediately above the bottom usually contains quite closeby. This potentialvent site was also not investigated lessCH 4 thanthe samples higher up. This probably indicates that further at the time. thesource is notdirectly below the cast but somewhere to theside, In the Shumaginsector three hydrocastscovered the entire or as has been recentlypostulated, it might signify short-term margin horizontallyand vertically from the trenchfloor to the pulsesof verticalCH 4 injectionfollowed by mixing [Radlinski upperslope off the backstopridge (stations97-79, 97-95, and97- SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2603

,1,1,1,1,1,1,1,1,1,

b 1000 1000

2000 2000

[] 97-20 3000 3000 ß 97-23-1 ß 97-23-2

ß 110-27

4000 4000

5000 5000

6000 I i I I I I I I i I I I I I 6000 ' I'1 '1' i'i' I' I'1 '1' 30 60 90 120 150 20 40 60 80 1 O0 Methane [nl/1] Methane [nl/l]

, I , I , I ,_1 , I_l I , I , I , I , . ::::::::::::::::::::::::::::::::;?:i:.i!i:i•.':;.•:•:.ii..:::::'•?.':.::i::..i:'::::7•-.•:.!'.:.:.---.

1000 1000

2000 2000

97-44 ß ' 97-49-1 97-49-2 : ..... ß 97-103-2 3000 97-51 - .•.•j.-•!•

4000 4000 :•--"-"-"-"-"-"•-"-'• :•.:•:..:j:...• - •:....

5000 5000-'

6000 ' i'1'1' !'1'1'1'1' I' 6000 ' •" •' •' •' I' I' •' i' • ' 0 20 40 60 80 IO0 0 20 40 60 80 1 O0 Methane [nl/1] Methane [nl/1]

Figure 4. Methanedistribution in the oceanicwater column over the AleutianTrench. Injectionof CH 4 from the seafloorat the Edgesector; (a) stations97-11-1, 97-11-2, 97-19, 97-30, 97-37, (b) stations97-20, 97-23-1, 97-23- 2, and 110-27, Albatross sector; (c) stations97-44, 97-49-1, 97-49-2, and 97-51, and Shumagin sector; (d) stations97-79, 97-95, 97-103-1, 97-103-2, and 110-39. Bottomdepth is indicatedby horizontallines; note well- developedplumes at differentdepths off the bottomwhich correspond to deformationridges; also note different concentrationscale for Figure4a. The maximaof CH 4 in surfacewaters and decreasing gas contents to about800 m depthare due to biogeochemicalcycling. 2604 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

103). The depthsof the lower water columnthus surveyed ranged Potential temperature [0, øC] from 3000 to 6000 m. Suchcoverage over a distanceof more than 50 km is not detailed enough to locate all possible venting 1.03 1.05 1.07 1.09 1.11 1.1 3 structures,yet the deformationridge at lower slope,between 4300 and 4900 m, showed significant CH 4 anomalies in two casts (Figure 4d, stations97-79, and 97-103). This location, inside a canyon which crosscuts the third accretionary ridge, was 4000.... confirmedto be an activevent siteby EXPLOS-surveysand TVG

sampling. The CH 4 anomaly was located again 2 years later , (station 110-39), generally coinciding with the previous maxi- 4500- . ' ..*...i ...... • ...... : ...... mum. Surprisingly,the water column at the trench axis near the deformationfront showedno CH4 anomaly(station 97-95). Again, a high-resolutionsurvey would most likely ascertainmore widespread and significant CH 4 injection than currently documentedin the Shumaginsector. j 3.2. Temperature No clear evidence for positive temperatureanomalies in the bottom water surroundingsubduction vents had been reported, sooo...... x SO /10 !...... althoughelevated heat flow has generally been associatedwith o SO 96/10b vent colonies[Henry et al., 1992] andhigh sedimenttemperatures, ß S097/35a for example, in excessof 20øC, were recordedon a mud volcano 5500- ]...... :...... •...... •...... •... ß S097/35b.... site off Barbados[Henry et al., 1996]. Here in the Edge sector + S096/391 near the deformationridges as well as at severalother locations o S096/392 acrossthe Alaska margin, the EXPLOS-mounted CTD sensors * S097/77a showed,for the first time, evidence for a slight temperature o S097/77b anomalyin the water column.Figure 5 showscomposite potential [] S097/78a temperatureversus depth recordsobtained from five TV surveys I I while towing the instrumentover the deformationridges (stations ooo ....I,,,,,I .... I ....I ....I .... i .... i .... i .... i .... 96-2, 96-3, 96-10, 97-35, 97-77, and 97-78). During this largely horizontaltrajectory the instrumentcontinually traverses different Figure5. Compositeprofile of potentialtemperature versus depth fromCTD deploymentsalong five EXPLOS tracks.Stations 96-2, in situ temperatureregimes as a functionof adiabaticheating, 96-3, 96-10, 97-77, and 97-78 showambient distribution; station which makes it difficult to identify true anomalies. Potential 97-35 showsa positivetemperature anomaly throughout the temperature,however, compensatesfor the adiabaticheating and surveyas recordedover initial deformationridges with active allows identification of an extra heat source. The composite venting. temperaturerecord from one TV surveyat station97-35 deviates significantlyfrom thoseof all othersurveys which cover the same depthrange but were locatedoff vent (Figure 5). There is a small of theorganisms are closely related to thosefound previously at positivetemperature anomaly (+ 0.010 + 0.002øC) for the entire othersubduction vents and cold seeps. record at station97-35. Furthermorethe crestalregions of each Numerousspecies of vesicomyidclams have previously been ridge show maxima in temperature.Both ridgesshow venting at found at cold seepsoff Oregon,Japan, and Peru and at the theirtrench-facing flanks.It is clear that the reproducibility, basedMonterey andAscension FanValleys offCalifornia [Hashimoto et onrepeated deployment ofthe instrument, andour lack of exact al., 1989' Olu et al., 1996; Barry etal., 1996; Embley etal., 1990; knowledgeabout the regional temperature structure ofthe water Rau et al., 1990; Suess etal., 1985]. Their nutrition isbased on massinthe Aleutian Trench and, particularly, theamplitude of thiotrophic chemoautotrophy, asindicated by light •3C values internalwaves make it difficulttounambiguously provethe [Rauet al., 1990], enzymatic andultrastructural work[Fiala- temperatureanomaly. Nevertheless, weare confident thatfuture Mddioni et al.,1993], and anatomical characteristics [Fiala-

entirehigh-accuracy subductiondata zone.will confirm heat input from vents along the MddioniandLe Pennec, 1989]. Several species ofsolemyids have beenencountered in sewageoutfalls and other highly reduced sedimentsand cold seepsin the deepsea [Felbecket al., 1981; 3.3. Fauna Suesset al., 1985;Embley et al., 1990; Paull et al., 1984].Their nutritionis also basedon thiotrophyof the endosymbionts. The second and by far the most characteristicindicator for Becauseof the largeindividual size of the bivalves,their light activeventing is the occurrenceof seepbiota. These consisted of coloration,their great abundance and characteristic arrangements bacterialmats, pogonophorans,vestimentiferan and omnipresent in clustersand alignments,they are ideally suitedfor visual large coloniesof bivalves.At all four sectors,abundant colonies of detectionof activevent sites.Vestimentiferans and pogono- vesicomyid clams, solemyid protobranchs,perivate pogono- phoransare also chemoautotrophic,although their nutritional phorans, and new, unusual vestimentiferanswere observedand pathwayand sourcesare not fully known.Most speciesare sampled (Plates 1 and 2). During R/V Sonne cruise 110 the thiotrophic;some are clearly menthanotrop'hic[Brooks et al., collectionof specimenswas considerablyenlarged. The work on 1987;Southward et al., 1986,1981; Schmaljohannetal., 1990]. taxonomyof the organismshas not been completedbut will be Our mostextensive survey for ventfauna was along strike of dealtwith in subsequentpapers. Nevertheless, it appears that most the two deformationridges of the Edgesector. Two crossings SUESS ET AL.' FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2605 2606 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

Plate 2. Cold seepfauna; bottom photographs taken by ROPOS-three-chipCCD camerasystem (station 110-23, Edge). For scale;the size of vesicomyidclams range between 135 and 190 mm (a) Alignmentof cold seep communities.(b), (c), and (d) Extensivepopulations of vesicomyidclams (higher magnification of population seenin Plate 2a (e) and (0 Pogonophoransamongst clams and gastropods(Plate 2f hashigher magnification of pogonophoransseen in (e) Plate2e showsgastropod egg cases present on two pogonophorantubes).

(stations 97-21 and 97-35) of these structuresshowed vent fields 110-23 [Suessand Bohrmann, 1997; Orange et al., 1996]). The indicated by bivalve coloniesjust off the ridge crestson the vehicle was equipped with two video cameras, a wide angle trench-facingflank (Figure 2). Examplesof the towedEXPLOS silicon intensified tube (SIT), low light camera, and a three-chip video documentation are shown in Plate 1. The trenchward flank CCD, broadcastquality color camerawith 16X zoom. The video of the first deformation ridge during SO 110 also became the tapes were recordedin Beta 5 P format for the vehicle's color targetof the longestand deepest ROPOS deployment ever (station cameraand SuperVHS for the vehicle's black and white camera. SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2607

Extensivephoto documentationwas obtainedby thesesystems, [Paull et al., 1984], the Skagerrak [ Schmaljohannet al., 1990], and selected views of communities are shown in Plate 2 [Lutz et andthe Louisianaslope [Brooks et al., 1987]. Only Siboglinumsp. al., 1996]. The alignment of cold seeps is along geologic from the Skagerrakappears to incorporateas large a portion of structures,as is their preferredlocation at the baseof stepsor in methane-derivedcarbon as the specimensfrom the Shumagin depressions.Extensive populations of vesicomyidclams, buccinid sectorreported here. The tubes 813C values of around-65%0 PDB gastropods,and pogonophoranswere documented[Lutz et al., are close to the C isotope compositionof a slightly depleted 1996]. Analysis of ROPOS videos showsthat the averageclam biogenicmethane pool from the Cascadiasubduction zone [Suess fieldis -0.7 m2 in area[ Sahling,1997]. and Whiticar, 1989]. This value identifies methanotrophyas the Attemptsat samplingof coloniesin the Edge sectorusing a dominantcarbon metabolizing pathway. Pogonophorans from the TV-guidedgrab (sampling area 1.82 m2) at one station (station 97- Edge sector,tentatively identified as Lamellisabellasp. and Poly- 66) yielded 119 specimensof vesicomyid clams and seven brachiasp., on theother hand, showed 8•3C values from -40 to - specimensof solemyids.At anotherstation (station 97-22), 21 47%0PDB (Table 2), indicatingthiotrophic carbon metabolism. A specimensof vesicomyidclams and 14 specimensof solemyids singlevestimentiferan genus from the Edge sectoranalyzed so far were collectedfrom this samegrab. Although the evaluationof the is evenless depleted in 8•3C(-28.6%0 PDB), thereby supporting observedand sampledbiota is far from complete,our preliminary thiotrophy.The soft tissueparts separated from vesicomyidclam assessmentbased on two surveysat eachsector and the analysisof and solemyid specimensfrom the Edge sector have relatively ROPOS videosshow population densities at vent sitesof between "heavy"•3C valuesof between-33 and-40%0 PDB with slight 159-200 individuals/m2. A total of over 56 m2 of active ventswere enrichmentin the gills (Table 2). Thesevalues indicate thiotrophic documented with several thousand specimens of bivalves carbonmetabolisms, whereby methane,via sulfatereduction, may [Sahling, 1997]. providethe sulfideneeded by the sulfur-oxidizingmicrobes. The nutritionalpathway at the ventecosystems of the Aleutian The carbonate shell material also exhibits evidence for chemo- Trench is also chemoautotrophicas previouslyobserved in other autotrophic-derivedcarbon. Usually, shells of vent bivalves are subductionsettings. The carbonisotopes of the soft tissuesand of only slightlyenriched in •13C becausethey preferentially the chitinoushard partsexhibit •5•3C-enrichedstable isotope incorporatebicarbonate from seawater,but for vesicomyidclams patterns(Table 2). The •5•3Cvalues for all of thenon calcareous from the it was shown that the inner partsanalyzed range from -28.6%0to -64.3%0Pee Dee belemnite complexcross-lamellar layer is moreenriched in •13Cthan the (PDB) and thus show thiotrophicand methanotrophicportions of outermost shell layer [Wagner, 1994]. Presumably, the close carbon incorporated.The values are similar to or exceed those proximity of the shell-secretingmantle to the digestive system, reportedfrom vent coloniesof other subductionzones and are where light metaboliccarbon is generatedand the inner shell layer comparableto thosefrom biota of otherprominent cold seepsnot formed, could retard dilution by seawaterbicarbonate and hence relatedto subduction[e.g., Paull et al., 1992]. The preserve more lighter carbon. For solemyid shells, no such enrichmentof the pogonophorantubes of the Shumaginsector distinction between inner and outer shell layers was possible, (15•3C= -64.3%0 PDB) is thestrongest observed so far. although,in general,the bulk of their carbonatewas isotopically Previousreports on •5•3Cof pogonophorantubes are from the lighter than the carbonate of vesicomyid clam shells from the Cascadia margin [Kulm et al., 1986], the Florida Escarpment Edge sector (Table 2). A value of-7.9%0 PDB indicatesthat

Table 2. StableCarbon Isotope Composition of SelectedTissue and Hard Partsof Vent Fauna

Fauna Station/Sector TissuePart •3C %0 81-•C%0 8•3C%o Maximum Minimum Mean

Vesicomyidclam 97-66/Edge mantel -36.66 -36.64 -36.65 + 0.01 gill -37.31 -37.43 -37.37 + 0.06 foot (inner) -39.87 -39.94 -39.90+0.03 foot (outer) -35.89 -36.28 -36.09 + 0.02 aductor close to -36.30 -36.24 -36.27 + 0.03 foot Solemyidspecimen 97-22/Edge periostracum -30.22 -29.09 -29.60 + 0.5 mantel -34.06 -34.04 -34.05 + 0.01 gill -35.29 -35.40 -35.35 + 0.05 siphon -33.43 -33.29 -33.36 + 0.06 shell -8.09 - 7.80 - 7.90 + 0.10 viscera -34.12 -34.16 -34.14 + 0.02

Pogonophoran tubeincluding -43.25 -40.55 -41.9ñ0.1 Lamellisabella? 97-22/Edge organism

Polybrachia? 97-22/Edge tubeincluding -47.35 -47.50 -47.4 ñ 0.1 organism

Vestimentiferan 97-22/Edge tubeincluding -28.80 -28.43 -28.6 ñ 0.2 organism

Pogonophoran 97-97/Shumagin emptytube, upper -67.37 -61.13 -64 ñ 3 emptytube, lower -55.82 -58.35 -57 ñ 1

•5•3C,%0 relative to PeeDee belemnite (PDB). 2608 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

solemyidsutilize a mixtureof metabolicand bottom water CO2 in composition(•5•80 = +2.1to +3.2%0_+ 0.1%o PDB) is significantly building their shells,probably because of their peculiarhabit of "lighter"than would be expectedfrom equilibriumwith the low living buried in the vent sediment and pumping bottom water bottom-watertemperature and oceanicisotope composition. At through their burrows. Other vent organismswhich showed a present,it is not clear whetherthe "light"oxygen isotope values similar b•'•C-enriched bulk calcareous skeleton have been the indicateformation at elevatedtemperatures or precipitationfrom serpulid worm tubes collected at the cold seeps of the Peru fluidshaving a differentb•80 makeup than present day seawater. subductionzone [ Wagner, 1994], the shellsof Thyasirasp. from The vent carbonatesfrom Edge and Shumaginsectors differ in the Skagerrak (-12%o PDB [Schmaljohannet al., 1990]) and theirb•'•C values by approximatelythe same magnitude asdo the mytilids from the Florida Escarpment(-8%0 PDB [Paull et al., tissuesof the ventorganisms from the two areas,except the range 19851). of valuesfor tube worm from Edge is large. Sinceall specimens fromEdge belong to differenttaxa, isotope fractionation during C 3.4. Precipitates incorporationinto the tubeworms seems likely, althoughdifferent sources of nutritional carbon cannot be excluded. Carbonateand barite crusts,cemented sediment, and chimney- like structures are typical precipitates which form around 4. Fluids and Flow Rates subductionvents and cold seeps[ Ritger et al., 1987; Kulm and Suess, 1990; Paull et al., 1992; Dia et al., 1993]. This Ever sincethe discoveryof subductionvents, the magnitudeof precipitation results from excess carbonategenerated by the flow and the modulation of flow rates over time remain two of the anaerobic microbial oxidation of methane in the case of carbonate most significantopen questions.The flow rate is importantfor and from mixing of Ba-saturatedvent fluid with SO4-richbottom estimatingfluid budgets and mass transportrates. It is a key waterin the caseof barite [Ritgeret al., 1987; Torreset al., 1996]. parameterfor modeling tectonic dewateringduring subduction Aside from their variablecomposition and apparentdisequilibrium accretionand is consideredof major importancein the buildupof mineralassemblages, vent carbonates are always enriched in •51-•C.mechanical stress which eventually may lead to Extremestable isotope values of between-40and-70%,, PDB have [Sammondset al., 1992]. Combiningdirect flow measurementsat beenreported [Ritger et al. , 1987;Roberts and Aharon, 1994].At active vent sites with sequential water sampling and high- both the Edge and the Shumaginsectors, carbonate and sulfate resolutionpore water chemistryis our approachto quantifyflow precipitatesoccur as a result of fluid venting. They range from ratesand chemically diagnose vent fluids. disseminatedmicrocrystalline phases to fragile linings of fluid A TV-guided instrumentfor vent fluid sampling(VESP) has channelsto large, dense impermeable precipitatesand crusts. been available for the in situ measurements of flow rates from Their presencein certain depth horizons is evident from the vents, which is describedin detail elsewhere [Linke et al., 1994; chemicalcomposition of sediments,particularly when coresfrom Carson et al., 1990]. This instrument, not unlike a benthic on-ventlocations are comparedto thosefrom off-vent locations. chamber,is towed close to the seafloornear active dewatering Most prominentare zonesof vent-inducedcalcium carbonateand zonesand placed over vent colonieswhen the track accidentally barite formation at station 97-66 in the Edge sector.At a depth passesdirectly over an active vent. Preciseactive deploymentis between5 and 35 cm below seafloorthe CaCO3 contentincreases not possibleas in the caseof submersibleor ROV operations.On to almost 10 wt % from a level of usually no more than 1 wt % average,six to eight attemptsare neededfor one successfulVESP and the total Ba contentfrom a backgroundof-700 ppm to almost deploymentby towing. Besidesa thermistorflowmeter which is 6000 ppm. The distributionof theseprecipitates downcore is dis- positioned in the channeled flow which emanates from the cussedlater in the context of pore water chemistry from their seafloor,the benthic chambercontains sequentially triggered surroundingsediment, vent fluid composition,and flow rate esti- water samplerswhich enable monitoringof the concentration mates, but first we describe here the appearance of the changes of certain vent tracers with time. Methane, sulfide, precipitates,as diagnosticfeatures for cold vents. oxygen,4He, Ba, andLi havebeen successfully used as tracers Scanning electron microscopic (SEM) images from the [Linkeet al., 1994; Torreset al., 1996].A typicaldeployment time macroscopicallyidentifiable precipitatesof the Mg calcite and at present is between 30 and 60 min. Modifications of the barite linings of openchannels in sedimentsfrom the Edge sector instrumentare underwayto extendon-bottom time considerably (station 97-66), and from the carbonatecrusts from Shumagin by decouplingthe systemfrom the surfacevessel after deployment (station97-97) are shownin Figure 6. The barium sulfatecrystals and utilizing a free-return mode of the instrument when the are composedof a densefilling of fragile barite needles(Figure samplingand measuring cycle is completed. 6d) which show a palisadefabric (Figures6a and 6c) and often a The VESP instrument,when deployed sufficiently long over a concentrically layered structure (Figure 6b) which represents vigorously flowing vent, samplesa mixture of vent fluid and severalgenerations of precipitation.The densecarbonate crusts ambient bottom water. Once vent fluid tracers have been identi- from the Shumagin sector appears under the SEM as a fied, their tlux rate into the bottom chamber of the VESP structurelesscryptocrystalline matrix < 2-3 gm that forms an instrument can be used to derive the water flow. Likewise, the intergranularcement. Euhedral Mg-calcite crystals(Figure 6f) oxygen demand of a vent ecosystem may, under certain were only found in larger pore spacecavities such as in diatom assumptions,also yield vent fluid flow [Wallmann et al., 1997b]. frustulesand radiolarian tests (Figures 6e and6g). The mineralogy Theseestimates may thenbe comparedto directlymeasured rates of bothMg-calcite and barite was confirmed by X ray diffraction. by the VESP flow meter, as previouslyshown by Carsonet al. The carbonisotope signature (•5•3C) of theEdge carbonates [1990] and Linke et al. [1994] and to geophysicallyderived rates rangesfrom -10.7%o to -14.3%oand of theoxygen isotopes (b•80) [vonHuene et al., 1997]. from 1.1%oto +2.9%0PDB (Figure7). The Shumaginprecipitates are also calcitic but consistof cryptocrystallinesubparallel layers 4.1. Direct Flow Measurements which coalesce to form large aggregates.The carbon isotope compositionis between-45.9 and-50.8%0PBD, in agreementwith VESP deploymentsat the easternAleutian subductionzone values of other vent carbonatesworldwide. The oxygen isotope yielded results close to the detectionlimit of the thermistorflow SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2609

d

Figure 6. Scanningelectron photomicrographs (SEM) of bariteprecipitates as fragilelinings in fluid channels from sedimentsof the Edge sector;station 97-66 (Figures6a, 6b, 6c, and 6d) and of carbonatecemented sedimentsfrom station97-79 (Figures6e, 6f, and 6g). Mineralogyby energydispersive (EDS) and X ray diffraction(XRD) analyses.Scale bar for Figures6a and6e is 30 pro;(scale bar for Figures6b, 6c, 6d, 6f, and6g is 10 Bm. 2610 SUESS ET AL.' FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

metaboliteconcentrations in pore fluids from on-vent site 97-66 (Figures8a-81). Authigeniccarbonates (EDGE-sector) -10' 4.3. Fluid Compositionand Mixing The porefluids, especiallythose from the heavilysedimented vents as in the Edge sector,represent a mixture of vent fluid, -20' diageneticpore water and to a considerabledegree ambient bottomwater. The latter is pumpedand recirculated into the vent system by benthic biota [Wallmann et al., 1997b]. The end- membercomposition of the pristinevent fluid and the diagenetic -30' pore water cannot, at present,be exactly differentiated, but the Tube worms •13c (EDGE-sector) contrastbetween the concentration-depthprofiles from on-vent and off-vent locationsidentifies the dominantfluid transport -4O modesand the environmentin whichdiagenetic precipitates form. The mostdramatic differences in porewater chemistryand solid phasecomposition between on-vent and off-vent sitesare shown in Figures8a-81 and briefly summarizedhere for the Edge sector -5O based on data from stations 97-66 and 97-72. A more detailed discussionof thesedata is given by Wallmannet al. [1997b]: (1) Authigeniccarbonates (SHUMAGIN-sector) Methane is virtually absentin the sedimentsfrom the off-vent site -60' but high at the on-vent site (Figure 8c). Owing to the lack of

Tube worms sufficientclosely spacedsamples, the profile is a compositeof (SHUMAGIN-sector) sediments from stations 97-66 and 97-22. The low methane all isotope data in %0 PDB content of two off-vent cores is shown as well (stations97-72 and -70 , [ , [ , ]' , [ , [ , [ , • , 97-88). (2) Calcium carbonateand barite precipitates form distinct -30 -25 -20 -15 -10 -5 0 5 10 layersin the on-ventsediment (Figures 8a and 8b). It appearsthat the CaCO3 layer is located at depth (maximum at 32 cm below seafloor(bsf) with two baritelayers above (maximum at 16 cm bsf Figure 7. Comparisonof stablecarbon versus oxygen isotopes and 8 cm bsf). The process controlling formation of these from vent carbonatesand pogonophorantubes between the Edge precipitatesis not subjectof this paper; (3) however, the pore and Shumagin sectors. Consistently lighter C isotopes at water data on dissolvedcalcium (Figure 81), barium (Figure 8k), Shumaginmight indicatemethylotrophic, and heavierC isotopes and,Y_,,CO2 along with the •5•3Cof ,Y_,,CO2(Figure 8i) clearly at Edge might indicatethiotrophic vent metabolism. document a relationship between diagenetic environment and mineral formation. (4) On the basisof their dissolvedC1 (Figure 8d) the pore watersfrom both sitesare indistinguishable.(5) The meter. In the Shumagin sector (station 97-104), we recorded on-vent site lacks a nitrate-containingsuboxic layer (Figure 8e). directlya wateroutflow of 10+ 8 L m-2d-• (= 240+ 200L m'2 d 'l) (6) Very low sulfatecontents, high sulfideconcentrations (Figure averagedover short deployment periods of up to 40 min. This 8h), and very high ammonia concentrations (Figure 8g) value falls into the lower range of previous measurementsoff characterize the on-vent sediments. These early diagenetic Oregonand Peru where between500-1700 L m'2 d'• were metabolites and methane are oxidized and consumed in or near the recorded, although the margin of error is rather high for the sedimentsurface by the vent megafaunaand the energyreleased Aleutian Trench data. duringthese redox reactionsis usedby the benthicorganisms for growth and metabolism. The shapeof severaldissolved constituents of the on-ventpore 4.2. BiogeochemicalApproach water profiles indicatesthat the topmost25 cm are influencedby the mixing of vent fluid with bottomwater via bioirrigation.The In the Edge area the thermistor measurementsproduced no irregulardistribution and elevated NO 2 contents(Figure 8f) attests reliable results due to very slow flow rates. Nevertheless,the to pumpingactivity. The peculiarlydepleted SiO2 concentrations water samplesrecovered during four successfulVESP deploy- in the on-vent sediment(Figure 8j) is due to the same process. ments allowed the quantification of fluid flow based on a Becausedissolved silicate is not consumedby organismsit may biogeochemicalapproach. A steadydecrease in dissolvedoxygen serve as a tracer for fluid mixing [Wallmann et al., 1997b]. with time was observed at vent sites and used to calculate an Bioirrigation manifestsitself at site 97-66 in lower concentrations averageoxygen consumption rate of 3.3 + 0.4 mmolm -2 h '•. The of silica and elevated NO 2 contents.The benthic megafauna rate is 2 ordersof magnitudehigher than the benthicoxygen flux pumpsbottom water with low SiO2 contentand somefree oxygen at deep-sealocations not influenced by fluid venting[ Wallmannet into the vent system,thereby diluting the diageneticpore water. al., 1997b].It is causedby the oxidationof reducedinorganic Using a transportreaction model, bioirrigationwas identified as compoundstransported to the vent siteby the tectonicallydriven the dominant transport mode at vent sites in the Edge area fluid flow. Using the contentsof methane(0.3 mmol), sulfide(4 [Wallmannet al., 1997b] The bivalvesat vent sitespump water mmol), and ammonia (3 mmol) in vent fluids of core 97-66 from throughtheir bodiesand throughthe surroundingsediments at a the Edge area, the averagefluid dischargeat vent sites was ratethat is severalorders of magnitudehigher than the tectonically calculatedto be0.23 L m-2 d 4 [Wallmannet al., 1997b].The vent driven fluid flow. Taking the abundanceof Solemya sp. and fluid compositionused is that representedby the highest Calytogena sp. at the vent sites,it may be estimatedthat the SUESSET AL.' FLUID VENTS IN ALEUTIAN SUBDUCTIONZONE 2611

CaCO3 [wt-%] Barium[wt-ppm] ON4[wt-ppb] Chloride[mM] 0 2 4 6 8 10 0 2000 4000 6000 0 2000 4000 6000 500 520 540 560 580 600 0i , i , I i I i i , i i I i I i I i i i ! ßß ß i ,• i

IO

20 = 97-88 Depth -22 [cm]

3O

4O

Nitrate[gM] Nitrite[gM] Ammonia[gM] Sulfate[mM] lO 20 30 40 5o o 1 2 3 4 5 6 7 8 o lOOO 2000 3000 o 5 lO 15 20 25 30 i .... , .... i .... , .... i .... i .... !

lO

2o ,

Depth [cm]

3o

4o

•,CO2 [mM] Silicate[gM] Barium[nM] Calcium[mM] 0 5 10 15 20 25 30 250 350 450 550 0 500 1000 1500 2000 7 8 9 10 11 i , i , i , I , i , i o

lO

20

Depth [cm]

30

40 -20 -10 0 (•130-•,002

Figure8. Comparisonof sediment composition andpore water chemistry between on-vent (station 97-66, solid symbols)and off-vent (station 97-72, open symbols) locations atEdge sector (see location map of Figure 2). The on-ventcharacteristics are(a) CaCO 3 and(b) Ba in layers;(c) high CH4; (d) normal oceanic C1; (e) lack of NO3- containingsuboxic layer; (f) unexpectedlyhigh NO2; (g) very high NH4; (h) low SO4 and very high H2S; (i) high •CO2 andvery significantly depleted /5•3C of •CO2 (on-ventlocation only); (j) peculiarlydepleted SiO 2 contents;(k) highdissolved Ba, and(1) low dissolvedCa. 2612 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE biologicalpumping rate is closeto 100 L m'2 d -• whereasthe minimumof 0.1% of the segmentbetween km 7.0 and km 12.5 tectonicallyinduced fluid flow amountsto only0.2 L m'2 d 4. showedventing [Sahling, 1997]. Hence an average rate of 20 L m' 2d-1 per active vent is expected based on the geophysical estimate. 5. Discussion Thisis in surprisinglygood agreement with the biogeochemically The importance of these considerationsfor the deep-sea derivedrate of 5.5+ 0.5L m'2 d -•. The large biological pumping rate at vent sites may bias the thermistorflow measurementsand environmentdepends on the reliability of large-scaleregional and the determinationof fluid flow ratesfrom biogeochemical and eventual global extrapolation to the world's subductionzones. temperature gradients in surface sediments. We feel that the Geophysicalapproaches have in the pastbeen the only attemptat estimating subduction-generatedfluid return fluxes. The dis- previousdiscrepancies in reported flow ratesmay be partially causedby this effect. crepancy among these estimates, based essentially on gross porosityreduction from seismicvelocity analyses,and the directly measuredflow rateshave been pointedout repeatedly.Up to now 6. Conclusion therehas been no compellingidea put forwardon how to reconcile these order-of-magnitude differences. In the context of our Alongthe Aleutian subduction zone, fluid ventingwas directly investigation we can now, for the first time, make a reliable observedalong an 800 km longsegment of its deep-seatrench off comparisonof both approaches.Friihn [1995] and von Huene et Alaska.Geophysical anomalies associated with ventinghere are al. [ 1997] have deriveda detailedporosity reduction for the Edge observed an additional 250 km toward the northeast. The cold sector based on combined reflection and refraction velocity seepsites are concentratedat the initial ridgesassociated with the analysesfor sedimentdeformed during the last 500 kyr (Figure9). deformationfront and are controlledby structure.The character- This showsthat the initial dewateringduring formation of the first istics of cold seeps seen at other convergentmargins are all and secondridges was associatedwith up to 50% of the total net documentedalong the lowermost,seaward facing slopeof the fluid lossoccurring principally across a 5.5 km wide zoneadjacent AleutianTrench. These characteristics include (1) methaneplumes to the deformation front. The highest rates of dewateringand in the lower watercolumn; (2) plumemaxima which are traceable tectonicshortening occur as imbricatestructures develop adjacent to deformationridges and structural settings associated with vents; to the deformation front and sediment becomes consolidated (3) evidencefor a smalltemperature anomaly at the cold seep sedimentaryrock. Thereafterporosity reduction is relatively slow sites;(4) prolific coloniesof ventbiota alignedand distributedin and involvesdeeper sedimentary units. patchescontrolled by fault scarpsoversteepened folds or bedding During our R/V Sonne surveys the areas of the first and exposed by mass wasting; (5) calcium carbonate and barite seconddeformation ridges were imaged by the EXPLOS video precipitatesat the surfaceand subsurfaceof cold vents;and (6) andby ROPOS camerasurveys. The segmentcorresponds to km 7 carbonisotope evidence from tissueand skeletalhard parts of to 12.5 km on the reconstructedfluid lossprofile (Figure9) by von biota, as well as from carbonateprecipitates, that ventsexpel Huene et al. [1997]. Over this distance of 5.5 km, 36.5% of the either methane- or sulfide-dominated fluids. This difference in total fluid lossoccurs which translatesinto a meandewatering rate chemistrymight be relatedto ageand source depth of fluids;they of 0.02L rrf2 d '1 or 8.6L m'2 d '•. Oursurvey clearly revealed that a mayhave migrated from greatdepth along vertical shear faults in

EDGE-Line 302 Third Ridge Second First RidgeRidge AleutianTrench

• igneousbasement -'"-/'-'""-

5%

•[] gradientof fluidloss in %/km profilelength

20 km 10 km 0 km

Figure9. Structureof accretionarycomplex in theEdge sector; line drawing from the SW partof line 302.Fluid lossin percentof total dewateringper kilometerprofile length and 1.8 My of convergenceage; fluid lossis calculatedfrom porosityreduction and convergencereconstruction [modified from Frahn, 1995; von Huene et al., 1997]. SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE 2613 the Shumaginsector. A biogeochemicalapproach toward estimat- Fiala-M•dioni,A., andM. LePennec,Adaptive features of the bivalve ing fluid flow ratesbased on oxygenconsumption and fluid molluscsassociated with fluid venting in thesubduction zones off Japan,Palaeogeogr. Palaeoclimatol. Palaeoecol., 71,161-168, 1989. chemistryatvent sites in the Edge sector converges towards a Fiala-M•dioni, A.,J. Boul•gue, S.Ohta, H.Felbeck, andA. Mariotti, meanvalue of5.5 + 0.5L m'2 d '• for tectonics-induced waterflow. Sourceofenergy sustaining theCalyptogena populations fromthe deeptrenches in subductionzones off Japan,Deep Sea Res.,40, 1241- Acknowledgments.We greatly appreciate the support, cooperation, and 1258,1993. professionalismatsea of masters H.Papenhagen andJ. Wagener andthe 'Fltih, E., and R. von Huene, FSSonne Cruise 96KODIAK SEIS, Kodiak- crew of R/V Sonneduring the cruisesSO-96, SO-97, and SO-110. We Hong Kong, Fahrtbericht210pp., GEOMAR Forschungs-zentrumfar thankB. Domeyer,A. Bleyer, T. Schott,M. Schumann,F. Appel, and A. marine Geowissenschaften der Christian-Albrechts-Universit•it Kiel, Cremer for continuedexcellent technical support;and A. D•ihlmann,J. Germany, 1994. Greinert,N. Maher, E. Zuleger,G. Rehder,N. yon Mirbach,G. Levai, and Frtihn, J., Tektonik und Entwaesserungdes aktiven Kontinentalrandes C. Moyer for theirenthusiasm and indispensable help at seaduring cruise suedoestlichder Kenai-Halbinsel,Alaska, GEOMAR Rep. 39, 93pp., SO-110. A few preliminarydata are includedhere from their work to GEOMAR Forschungszentrumfor marine Geowissenschaftender supportour earlier findings.Special thanks go to Tim Shank(Rutgers Christian-Albrechts-Universit•itKiel, Germany, 1995. University)for preparingand making available the illustrationsof Plate2. Hashimoto, J., S. Ohta, T. Tanaka, H. Hotta, S. Matsuzawa, and H. Sakai, We alsothank G. Fischer(University of Bremen)for the •513C Deep-sea communitiesdominated by the giant clam, Calyptogena measurementson organicmatter and shell material of the ventfauna. The soyoae, along the slope foot of Hatsushima Island, Sagami Bay, commentsby J. Martin,University of Florida,and an anonymousreviewer centralJapan, Palaeogeogr. Palaeoclimatol. Palaeoecol. 71, 179-192, aremost gratefully acknowledged. Financial support for theinvestigations 1989. usingR/V Sonnewas providedby the FederalMinistry of Education, Henry, P., J.-P. Foucher,X. Le Pichon,M. Sibuet,K. Kobayashi,P. Tarits, Researchand Technology,Bonn (grants03G0096A, 03G0097A, and N. Chamot-Rooke, T. Furuta, and P. Schultheiss, Interpretation of 03G0110A/B). The NationalScience Foundation, Washington, made funds temperaturemeasurements from the Kaiko-Nankai cruise:Modeling available through grants OCE-93-02205 and OCE-96-33131 to R. of fluid flow in clam colonies, Earth Planet. Sc. Lett., 109, 355-371, Vrijenhoekand R. Lutz, RutgersUniversity, to help supportthe ROPOS 1992. deploymentduring SO- 110. Henry, P., X. Le Pichon, S. Lallemant, S. Lance, J. Martin, J.-P. Foucher, A. Fiala-M6dioni, F. Rostek, N. Guilhaumou, V. Pranal, and M. Castrec,Fluid flow in and around a mud volcano field seawardof the References Barbados : Results from Manon cruise, J. Barry, J.P., H.G. Green,D.L. Orange,C.H. Baxter,B.H. Robison,R.E. Geophys.Res., 101(B8), 20,297-20,313, 1996. Kochevar,J.W. Nybakken,D.L. Reed,and C.M. McHugh,Biologic Horibe, Y., K.-R. Kim, and H. Craig, Hydrothermalmethane plumes in the and geologic characteristicsof cold seeps in Monterey Bay, Mariana back-arcspreading centre, Nature, 324, 131-133, 1986. California,Deep Sea Res. I, 43 (11-12), 1739-1762,1996. Hyndman,R.D., K. Wang, T. Yuan, and G.D. Spence,Tectonic sediment Boul•gue, J., E.L. Benedetti, D. Dron, A. Mariotti, and R. L6tolle, thickening,fluid expulsion,and the thermalregime of subductionzone Geochemicaland biogeochemicalobservations on the biological accretionaryprisms: The Cascadiamargin off Vancouver Island, J. communitiesassociated with fluid venting in Nankai Trough and Geophys.Res., 98(B12), 21,865-21,876, 1993. JapanTrench subduction zones, Earth Planet.Sc. Lett., 83, 343-355, Kastner,M., H. Elderfield,and J.B. Martin, Fluidsin convergentmargins: 1987. What do we know abouttheir composition,origin, role in diagenesis Brooks,J. M., M. C. Kennicutt II, C. R. Fisher, S. A. Macko, K. Cole, J. J. and importancefor oceanicchemical fluxes, Philos. Trans. R. Soc. Childress,R. R. Bidigare,and R. D. Vetter, Deep-seahydrothermal London A, 335, 243-259, 1991. seep communities:Evidence for energy and nutritional carbon Kulm, L.D. and E. Suess,Relationship between carbonate deposits and sources,Science, 238, 1138-1141, 1987. fluid venting:Oregon accretionaryprism. J. Geophys.Res., 95 (B6), Brown, K.M., B. Bekins, B. Clennell, D. Dewhurst, and G. Westbrook, 8899-8915, 1990. Heterogeneoushydrofracture development and accretionaryfault Kulm, L. D., and R. von Huene, Initial Reportso.f the Deep Sea Drilling dynamics,Geology, 22, 259-262, 1994. Project, vol. 18., U.S. Govt. Print. Off., Washington,D.C., 1973. Bruns, T.R., Model for the origin of the Yakutat Block, an accreting Kulm, L.D., et al., Oregon subduction zone: Venting, fauna, and terranein the northernGulf of Alaska, Geology,11,718-721, 1983. carbonates,Science, 231,561-566, 1986. Carson, B., E. Suess, and J.C. Strasser, Fluid flow and mass flux Lammers,S., E. Suess,and M. Hovland,A largem•thane plume east of determinationsat vent sites on the Cascadia Margin accretionary Bear Island (BarentsSea): Implicationsfor the marine methanecycle, prism,J. Geophys.Res., 95(B6), 8891-8897,1990. Geol. Rundsch.,84, 59-66, 1995. Carson,B., E. Seke, V. Paskevich,and M.L. Holmes,Fluid expulsionsites Langseth,M. G., and J. C. Moore, Introductionto specialsection on the on the Cascadiaaccretionary prism: Mapping diagenetic deposits with role of fluids in sediment accretion, deformation, diagenesis,and processedGLORIA imagery, J. Geophys.Res., 99(B6), 11,959- metamorphismin subductionzones, J. Geophys.Res., 95(B6), 8737- 11,969, 1994. 8741, 1990. ' Childress,J.J., C.R. Fisher,J.M. Brooks,M.C. KennicuttII, R.R. Bidigare, Le Pichon, X., et. al., Nankai Trough and Zenisu Ridge: A deep-sea and A.E. Anderson,A methanotrophicmarine molluscan(Bivalvia, submersiblesurvey. Earth Planet. Sci. Lett., 83, 285-299, 1987. Mytilidae) symbiosis:Mussels fueled by gas, Science,233, 1306- Le Pichon, X., P. Henry, and S. Lallement, Water flow in the Barbados 1308, 1986. accretionarycomplex, J. Geophys.Res., 95(B6), 8945-8967, 1990. Davis, E.E., R.D. Hyndman,and H. Viilinger, Ratesof fluid expulsion Lewis, S.D., J.W. Ladd, and T.R. Bruns, Structuraldevelopment of an acrossthe Northern Cascadia accretionaryprism: Constraintsfrom accretionaryprism by thrustand strike-slip faulting: Shumagin region, new heat flow and multichannelseismic reflection data, J. Geophys. Aleutian Trench., Geol. Soc.Am. Bull., I00, 767-782, 1988. Res., 95(B6), 8869-8889, 1990. Lilley, M.D., D.A. Butterfield, E.J. Olson, J.E. Lupton, S.A. Macko, and de Angelis,M.A., and C. Lee, Methaneproduction during zooplankton R.E. McDuff, Anomalous CH 4 and NH 4 concentrationsat an grazingon marinephytoplankton, Limnol. Oceanogr.39(6), 1298- unsedimentedmid-ocean ridge hydrothermalsystem, Nature, 364, 45- 1308, 1994. 47, 1993. DeMets,C., R.G. Gordon,D.F. Argus,and S. Stein,Current plate motions, Linke, P., E. Suess,M. Torres,V. Martens,W.D. Rugh, W. Ziebis, and Geophys.J. Int., 101,425-478, 1990. L.D. Kulm, In situ measurementof fluid flow from cold seepsat Dia, A.N., L. Aquilina,J. Boul•que,J. Bourgois,E. Suess,and M. Torres, activecontinental margins, Deep SeaRes., 41,721-739, 1994. Originof fluidsand related barite deposits at ventsites along the Peru Lutz, R., et al., Biologicalcommunities and geologicalfeatures associated convergentmargin, Geology, 21, 1099-1102,1993. with methane/sulfideseepage at 4960m in the Aleutian subduction Embley,R.E., et al., Geologicalsetting of chemosyntheticcommunities in zone,EosTrans. AGU 77(46), Fall Meet. Suppl.,F405, 1996. the MontereyFan Valley system,Deep Sea Res., 37(11 ), 1651-1667, Martin, J.B., M. Kastner,and H. Elderfield,Lithium: sources in porefluids 1990. of Peru slope sedimentsand implications for oceanic fluxes, Mar. Felbeck, H., J.J. Childress,and G.N. Somero, Calvin-Bensoncycle and Geol., 102,281-292, 1991. sulphideoxidation enzymes in animalsfrom sulphide-richhabitats, Martin, J.B., M. Kastner,and P. K. Egeberg,Origins of salinefluids at Nature,293, 291-293, 1981. convergentmargins, in Active Margins and Marginal basinsof the 2614 SUESS ET AL.: FLUID VENTS IN ALEUTIAN SUBDUCTION ZONE

westernPacific, Geophys.Monogr. Ser., vol. 88, editedby B. Taylor, and ResourcePotential of the ContinentalMargin of WesternNorth pp. 219-239, AGU Washington,D.C., 1995. Americaand Adjacent Ocean Basins-Beaufort Sea to Baja Cah.'fornia, Martin, J.B., M.M. Kastner, P. Henry, X. Le Pichon, and S. Lallemant, Earth Sci. Set., vol. 6, edited by D.W. Scholl, A. Grantz, and J.G. Chemicaland isotopicevidence for sourcesof fluidsin a mud volcano Vedder, pp. 503-522, Circum-Pac.Counc. for Energy and Miner. field seaward of Barbados accretionarywedge, J. Geophys.Res., Resour., Houston, Tex., 1987. 101(B8) 20,325-20,345, 1996. Suess,E., FS SonneFahrtbericht SO 97 KODIAK-VENT, GEOMARRep. McAdoo, B.G., E.A. Silver, D. Orange, K.D. Mcintosh, L. Kahn, L. 29, p. 220, GEOMAR Forschungszentrumftir marine Geowissen- Abbott, J. Galewsky, and M. Protti, Seafloorstructural observations schaftender Christian-Albrechts-Universit•it Kiel, Germany1994. from Costa Rica accretionaryprism, Geophys.Res. Lett. 23 (8), 883- Suess,E. and G. Bohrmann, FS Sonne Fahrtbericht SO 110 SO-RO, 886, 1996. GEOMAR Rep. 59, 181, GEOMAR Forschungszentrumftir marine Moore, C.J., et al., EDGE deep seismicreflection transect of the eastern Geowissenschaften,Christian-Albrechts-Universit•it Kiel, Germany, -trench layered lower crust reveals underplatingand 1997. continentalgrowth, Geology,19, 420-424, 1991. Suess,E., andM.J. Whiticar,Methane-derived CO 2 in porefluids expelled Moore, J.C., and P. Vrolijk, Fluids in accretionaryprisms, Rev. Geophys., from the Oregon subductionzone, Palaeogeogr.Palaeoclimatol. 30, 113-135, 1992. Palaeoecol., 71,119-136, 1989. Ohta, S., and L. Laubier, Deep biologicalcommunities in the subduction Suess,E., B. Carson,S.D. Ritger, J.C. Moore, M.L. Jones,L.D. Kulm, zone off Japanfrom bottom photographsduring Nautile dives in the and G.R. Cochrane,Biological communities at vent sitesalong the KAIKO project, Earth Planet. Sci. Lett., 83, 329-342, 1987 subductionzone off Oregon,Biol. Soc.Wash. Bull., 6,475-484, 1985. Olu, K., A. Duperret, M. Sibuet, J.-P. Foucher, and A. Fiala-M6dioni, Tilbrook, B.D., and D. M. Karl, Methane sources,distributions and sinks Structure and distribution of cold-seep communities along the fromCalifornia coastal waters to the oligotrophicNorth Pacific gyre, Peruvianactive margin:Relationship to geologicaland fluid patterns, Mar. Chem., 49, 51-64, 1995. Mar. Ecol. ProgressSer., 132,109-125, 1996. Torres, M., G. Bohrmann,E. Suess,J. Boulege, and J. Bourgois, Orange, D., et al., Fluid seepageat the toe of the Aleutian accretionary Authigenicbarites and fluxes of bariumassociated with fluid seepsin complex: Observationsfrom the longest and deepestROPOS dive the Peru subductionzone, Earth Planet. Sci. Lett., 144, 469-481, 1996. yet, Eos Trans.AGU, 77(46), Fall Meet. Suppl.,F 323, 1996. yon Huene, R., Continentalmargins around the Gulf of Alaska, in The Paull, C.K., B. Hecker, B. Commeau, R.P. Freeman-Lyndre, A.C. Eastern Pac!fic Ocean and Hawaii, editedby E.L. Winterer, D.M. Neumann,W.P. Corso, S. Golubic, J. Hook, E. Sikes, and J. Curray, Hussong,and R.W. Decker,pp. 383-401,Geol. Soc.of Am., Boulder, Biologicalcommunities at Florida Escarpmentresemble hydrothermal Colo., 1989. vent communities, Science, 226, 965-967, 1984. yon Huene, R., and D. Scholl, Observationsat convergentmargins Paull, C.K., A.J.T. Jull, L.J. Toolin, and T. Linick, Stableisotope evidence concerningsediment subduction, subduction erosion and the growth of for chemosyntheticin an abyssalseep community, Nature, 317, 709- continentalcrust, Rev. Geophys.,29,279-316, 1991. 711, 1985. yon Huene, R., and D. M. Scholl, The return of sialic material of the Paull, C.K., J.P. Chanton, A.C. Neumann, J.A. Coston, and C.S. Martens, mantleindicated by terrigeneousmaterial subducted at convergent Indicatorsof methane-derivedcarbonates and chemosyntheticorganic margins,Tectonophysics, vol. 219, pp. 163-175,1993. carbondeposits: Examples from the Florida Escarpment,Palaios, 7, yonHuene, R., M.A. Fisher,and T.R. Bruns,Geology and evolution of the 361-375, 1992. Kodiak margin,Gulf of Alaska,in Geologyand Potentialqf the Radlinski,A.P., andZ. Leyk, Formationof light-hydrocarbonanomalies in ContinentalMargin of WesternNorth America and AdjacentOcean oceanicwaters, Geology, 23, 265-268, 1995. Basins-Beau.[brtSea to Baja Cali[brnia, Earth Sci. Set., vol. 6, Rau, G.H., C. M. McHugh, C. Harrold, C. Baxter, B. Hecker, and R.W. Circum-Pac.Counc. for Energyand Miner. Resour.,edited by D.W. Embley,•13C, •15N, and 80 •8 of Calyptogenaphaseohform•s ß ß (bivalve ß Scholl,A. Grantz,and J.G. Vedder, pp. 191-212,Houston, Tex., 1987. mollusc)from the AscensionFan-Valley near Monterey,California, yon Huene, R., D. Klaeschen, M. Gutscher, and J. Frtihn, Mass and fluid Deep SeaRes., 37, 1669-1676, 1990. flux duringaccretion at the Alaskamargin, Geol. Soc.Am. Bull., in Rio, M., M. Roux, M. Renard, and E. Schein,Chemical and isotopic press,1997. featuresof presentday bivalveshells from hydrothermalvents or cold Wagner,C., Cold Seepsan konvergentenPlattenr•indern vor Oregonund seeps,Palaios, 7 351-360, 1992. Peru:Biogechemische Bestandsaufnahme, GEOMAR Rep. 53, 108pp., Ritger, S., B. Carson, and E. Suess, Methane-derived authigenic GEOMAR Forschungszentrumftir marine Geowissenschaftender carbonatesformed by subduction-inducedpore-water expulsion along Christian-Albrechts-Universit/itKiel, Germany,1997. the Oregon/Washingtonmargin, Geol. Soc. Am. Bull., 98, 147-156, Wallmann, K., E. Suess,G.H. Westbrook, G. Winckler, M.B. Cita, and 1987. MEDRIFF Consortium, Salty brines on the Mediterranean floor, Roberts,H.H., andP. Aharon,Hydrocarbon-derived carbonate buildups of Nature, 387, 31-32, 1997a. the northern Gulf of Mexico, Geo Mar. Lett., 14,135-148, 1994. Wallmann, K., et. al., Quantifying fluid flow, solute mixing, and Sahling,H., Untersuchungenan coldseep assoziierten Organismen im NE- biogeochemical turnover at cold vents of the eastern Aleutian Pazifik, Diplomarbeit,96 pp., Univ. of Kiel, Kiel, Germany,1997. subductionzone, Geochim. Cosmochim. Acta, in press1997b. Sammonds,P.R., P.G. Meredith,and I.G. Main, Role of porefluids in the Westbrook, G.K., B. Carson, R.T. Musgrave, and E. Suess (Eds.) generationof seismicprecursors to shearfracture, Nature, 359, 228- Proceedingso.f the OceanDrilling Program,Scient!fic. Results, 146; 230, 1992. OceanDrill. Program,College Station, Tex., 1995. Schmaljohann,R., E. Faber, M.J. Whiticar, and P.R. Dando, Co-existence of methane- and sulfur-basedendosymbioses between bacteria and invertebratesat a sitein the Skagerrak,Mar. Ecol.Prog. Set., 61, 119- G. Bohrmann,S. Lammers,P. Linke, H. Sahling,E. Suess,R. von 124, 1990. Huene, and K. Wallmann, GEOMAR, Research Center for Marine Scranton,M. I., and P. G. Brewer,Consumption of dissolvedmethane in Geosciences,Wischhofstr. 1-3, D-24148 Kiel, Germany. (e-mail: the deepocean, Limnol. Oceanogr. 23(6), 1207-1213,1978. [email protected];[email protected]; [email protected]; Sibuet,M., S.K. Juniper,and G. Pautot,Cold-seep benthic communities in [email protected];[email protected]; [email protected]) the Japan subductionzones: Geological control of community R. A. Lutz, Centrefor Deep-SeaEcology and Biotechnology, Institute development,J. Mar. Res.,46,333-348, 1988. of Marine and CoastalSciences, Rutgers University, New Brunswick, Southward, A.J., E.C. Southward, P.R. Dando, G.H. Ran, H. Felbeck, and 08903NJ. (e-mail:[email protected]) H. Fl0gel,Bacterial symbionts and low •'•C/•2Cratios in tissuesof D. Orange, Monterey Bay Aquarium Research Institute, 7700 Pogonophoraindicate unusual nutrition and metabolism,Nature, 293, SandholdtRd., MossLanding, CA 95039.(e-mail:[email protected]) 616-620, 1981. G. Winckler,Institut ftir Umweltphysikde r Universit•itHeidelberg, Im Southward,A.J., E.C. Southward,P.R. Dando,R. L. Barett,and R. Ling, Neuenheimer Feld 366, D-69120 Heidelberg, Germany. (e-mail: Chemoauthotrophic function of bacterial symbionts in small WI @uphys 1 .uphys.uni-heidelberg.de) pogonophora,J. Mar. Biol Assoc.U.K. 66, 415-437, 1986. Stevenson,A.J., andR. Embley,Deep-sea fan bodies,terrigenous turbidite (ReceivedFebruary 6, 1997; revisedJuly 16, 1997; sedimentation,and petroleumgeology, Gulf of Alaska, in Geology acceptedJuly 23, 1997.)