JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 95, NO. B8, PAGES 12,689-12,696, AUGUST 10, 1990

Axial ' An Active Axis on the Central

H. PAUL JOHNSON

Schoolof Oceanography,University of Washington,Seattle

ROBERT W. EMBLEY

NOAA, PacificMarine EnvironmentalLaboratory, Hatfield Marine ScienceCenter Newport,

INTRODUCTION Juande FucaRidge can be dividedinto six 50- to 100-km-long AxialSeamount (some of themanuscripts in this special segments, with the Axial Segment being the third segment sectionrefer to theedifice with the more precise name of northof theBlanco (Figure 2). "AxialVolcano"), a large ridge axis volcano, islocated on the AxialSeamount represents a large mass excess associated centralsegment ofthe Juan de Fuca Ridge approximately 250 with a ""-influencedspreading center. The term nauticalmiles west of the Washington/Oregon/British"hotspot" is usedin quotationshere since geochemical Columbiacoast. Currentlyboth volcanically and hydro- evidence (Desonie and Duncan, this issue; Rhodes etal., this thermallyactive, Axial lies directly at the intersection of the issue)argues that Axial is nota geochemicalhotspot in the Cobb-EickelbergSeamount Chain and the Juan de Fuca Ridge traditional Hawaiian or Icelandicsense. Examination of the (Figure1). Thevolcanic activity associated with the seamount older, more distal portion of the Cobb-Eickelberg (C-E) chain formationstrongly interacts with, and is affectedby, the (Figure1) showsthat it is a relativelythick "band" of normalseafloor spreading processes at the intersection., rather than a singlebathymetric lineation. Becauseof thisunique geologic setting, its proximity to west Riddihough etal. [1983]and Davis and Karsten[1986] coastports and oceanographic institutions, and its shallowobserved that almost all of theseamounts formed at theC-E depth,Axial hasbecome the focusof a largenumber of hotspotand smaller seamounts found in proximityto theJuan scientificinvestigations over the past decade. deFuca Ridge north and south of Axiallie onthe Pacific plate AxialSeamount became an early target for study using the to thewest, rather than on the smaller to newly developedswath-mapping tools of Sea Beam,the east [Riddihough et al., 1983;Davis and Karsten, 1986]. SeaMARC I andII, andGLORIA. During the entire decade Thegeneral setting of inthe eastern Pacific, of the1980s, geologists and geophysicists from a widevariety and the overall morphology of the edifice, can be seenin of institutions,including the U.S. Geological Survey, NOAA, Figures1 and 2, andin thecolor Plates A andB of thisissue. the PacificGeoscience Centre, as well as the Universitiesof To thesouth, structures of theVance Segment overlap with the Washington,Victoria, and British Columbia, and Oregon State South Rift Zone of Axial and interact with the seamount in the University,conducted extensive surveys and detailed sampling SE quadrant. This zone of interaction between the Vance and programsof boththe summit and flanks of thevolcano. Since Axial segmentsis quiteextensive, with at least30 km of 1984,the NOAA VENTS Program has devoted amajor surface overlap. The farthest northward projection of spreading from shipand submersible effort to high-densitystudies of the theVance Segment is a zoneof intensefissuring (observed in summitarea and its hydrothermal systems. The shallow 1500- SeaMARC I andII data)on the outer flank of thevolcano m summitdepth of theseamount has allowed access with [Johnsonand Holmes, 1989; Appelgate, this issue]. Piscessubmersibles (both Canadian and Russian) as well as Interactionofthe Axial Segment with Northern Symmetrical Alvin.This extensive sea-going effort has made Axial one of (alsonamed Cobb) Segment to thenorth is morecomplex. themost intensely studied and best characterized areas of Bothmagnetics and morphology data show that the Axial seafloorin the world. Segmentforms a "bentspreading center" connection zone betweenthe two segments,an unusualintrasegment connection GEOLOGICALSETTING regionthat is apparentlynonoverlapping at the present [Johnsonand Holmes, 1989]. An elevated saddle of basement Axial Seamountis locatedastride the centralsegment of the rock connectsthe west flank of Axial with the easternmost Juande FucaRidge, an intermediatespreading-rate ridge with flank of Brown Bear Seamount(Plate A), the next oldest a total openingvelocity of 6 cm/yr. The Juande Fuca Ridge seamount in the C-E chain. The Son-of-Brown Bear is boundedon the southby theBlanco Fracture Zone and on Seamount,lying directlyeast of Axial Volcano,and three the northby a triple junctionformed by the ridge, the left- smallervolcanos just to thenorth and south (Plate A) arefour lateral Nootka and the SoyancoFracture Zone [Atwater, members of the rare class of that formed near the 1970;Hyndman et al., 1979]. Like the EastPacific Rise, the ridgeaxis and ended up on theJuan de Fucaplate, rather than the Pacific plate.

Copyright 1990 by the AmericanGeophysical Union. STRUCTuRAL COMPONENTS OF THE VOLCANO

Paper number 90JB01037. Structurally,the edifice of Axial Volcanocan be interpreted 0148-0227/90/90JB-01037502.00 to represent a complex interplay of three usually distinct

12,689 12,690 JOHNSONAND EMBLEY.' •NTRODUCTION

t50 ø t40 ø 430 ø •20 ø 60 ø ' I ' I

CANADA /røo/•/r .. PRATT ' •'0, ":"'' •'•__ •,• W,•. '.-..... •1•'e• OH41N .'..'.' • ß • • •:..... •'•,• • . WELKER ,._ • "''.

• WILSON • KNOLLS

VANCOUVER 50 ø T.ZO ' •/c,•4 ß • ISLAND

EICKELBERG COBB

PACIFICPLATE • ß THOMPSON JUAN DE FUCA PLATE

I 0 km 500 BROWN ,• I BEAR U.S.A. at 42øN

a:: RDA

.. 40 ø i ! Fig. 1. Seamountsand spreadingcenters in the northeastPacific, adaptedfrom Riddihoughet at. [1983]. Thick hachured lines are subductionzones. Note that the Cobb-Eickelberg Chain has a finite width and actually consistsof a band of seamounts rather than a linear chain.

geologicalprocesses: , seamount formation, and movement of the water column above Axial; and (4) andHawaiian-style rifting. The elongatestructure of theentire specific experimentsand studiesthat focus on the active edifice (PlateA) appearsto representthe distributionof an hydrothermalsystems. "excess"of magmatic activity along the entire Axial Segment. The elongateNorth and South Rifts show evidenceof "rifting" Axial as a Volcano or, more correctly phrased,longitudinal growth due to along- The large-scale context of Axial is presentedin the papers strike, near-surface magma transport. Finally, the dome- by Desonie and Duncan and by Rhodes et al. Desonie and shapedsummit and argue for seamountlikeformation Duncanexamine the age andchemistry of the Cobb-Eickelberg processes,including eruption from (and subsequentcollapse (C-E) seamounts,using dredgesamples collected mostly from into) localized magma reservoirs. the distal members of the chain, and find a westward age The summit of Axial, with its well-defined, three-sided progressionaway from the spreadingcenter, consistentwith caldera(Figure 3 and PlateB) and associatedactive the general assumption thatAxial is the youngest member of hydrothermalsystems, hasprovided the focus of many of the theprogressively older (to the west) seamount chain. The studiesinthis volume. Floored with very recent Hows, same study also shows aneastward decrease inage difference someof which are young enough toonlap actively venting between theseamounts andthe crust onwhich they are formed, hydrothermal deposits, this caldera is a potential natural arguing that the spreading center and the "mantle source" laboratory. Because of the shallow 1500-m depth of the responsible for the seamountshave been approaching each calderafloor, high-temperature hydrothermal systems can other as the Juan de Fuca migrates toward the fixed C-E undergo"phase separation," a term used by water chemists "hotspot." Desonie andDuncan, however, findthat there isnot when they are discussing"boiling" of the hydrothermalHuid a clear, mantle plume signatureto the geochemistryof the (Butterfieldetal., this issue). The influence ofthis "phase surface rocks of these outer seamounts andcall upon mixing separation" on hydrothermal systems is profound and is a between a weak mantle plume and normal spreading-center themethat runs through several of themanuscripts in this magmas toexplain the observed compositions. specialsection. Rhodesetal., on the other hand, concentrating onsamples from the summit and flanks of Axial volcano itself, find no THISISSUE evidenceat all of any mantleplume influence in the The manuscriptsin this issuecan be divided into four broad geochemistryof the rocks. These authorsargue that although categories: (1) the large view studiesthat look at Axial as an the geochemicaldata (as well as the morphologicalevidence) active ; (2) the more focused summit appearto favor a very activeand extensive magmatic plumbing geologicaland geophysical studies; (3) studiesof thechemistry systemfor Axial Seamountand one that is separateand distinct JOHNSONAND EMBLEY: INTRODUCTION 12,691

t30 ø 126 ø 122 ø

..

...

... ß .. 52 ø 52 ø ß . .ß.

ß

ß

ß ß ß o

ß.Van.couver ß

SFZ

•eattle 48 ø 48 ø ß

ß

ß

"NORTH • NSR CCobb) oo AMERICAN

•" AXIAL PLATE ••. DEJUAN FUCA

PLATE

44 ø 44 ø

PACIFICPLATE ß

o (,•

40 ø 40 ø

130ø 126ø 122ø Fig.2. Schematicdiagram of theridge segments of the Juan de Fuca Ridge, in thenortheast Pacificß Thick lines indicate thespreading center segments; thinner lines indicate fracture zones and the location of thecontinental marginß The followingabbreviations are used: SFZ,Sovanco Fracture Zone; NSR, Northern Symmetrical Ridge; AXIAL, Axial SeamountSegment; VA, VanceSegment (or segmentB in the olderliterature). fromthe rest of theJuan de Fuca Ridge volcanics, the volcano withincrust formed during the normally magnetized Brunhes representsthe surfaceexpression of a thermal,rather that a epic,the edifice does not produce a simplepositive anomaly geochemical,anomaly in themantleß thatcan be simplyrelated to thebathymetry. Both inversion Potentialfield measurementsover Axial Seamount,including of the anomalydata and forwardmodeling coupled with rock seasurface magnetics and both seafloor and sea surface gravity magneticsmeasurements show that the seamount contains large data,have provided some insight into the internal structure of zonesof relativelylow magnetizationassociated with the thevolcanoß Tivey and Johnson, using the closely spaced sea summitregionß While the interpretationis inherently surfacemagnetics data acquired during many years of cruisesnonunique, the anomaliesare bestexplained by a thinned in the area,show that even though the seamount lies entirely (<700 m) sourcelayer underlying the summitof thevolcano. 12,692 JOHNSONAND EMBLEY: INTRODUCTION

04'

02'

46000 '

58'

56'

54'

52'

45050 ' 08' 06' 04' 02' 130000 ' 58' 56' 54' Fig. 3. Detailed bathymetrymap of Axial summitand caldera,from the NOAA Sea Beam coveragein the area. Note the three-sidedcaldera, with an orientationthat is obliqueto all othertectonic trends in the region. Depthsare calculatedusing a watervelocity of 1500m/s. Primarynavigation was by LORAN-C,but wasshifted to GlobalPositioning System base. Contour interval is 10 m.

The necessarythinning could be causedby the presenceof a Rift Zone, describedby Tivey and Johnson,is also consistent shallow, subcalderamagma chamber, residual heat in a with this interpretation. transportpipe from previouseruptions, or a zone of extensive The processof rifting over the SouthRift Zone is discussed subsurfacealteration. at length in the paper by Appelgate. Using high-resolution Hildebrandet al., usingboth sea surface gravity anomalies SeaMARC I data, Appelgate shows that the surficial anddata from a new seafloorgravimeter, report evidence for morphologyof the SouthRift Zone systematicallychanges a shallow,low-density region underlying the summitcaldera, from linear volcanicridges near the summitto smallcratered alsosuggesting a partiallymolten magma chamber at depth conesat the distal end. Furthermore,he showsthat ridge- within the volcano. Away from the summit,the gravitydata parallelfaults along the SouthRift arethe sources for themost of Hildebrandet al. suggestthat the North Rift Zonerepresents recent lava flows in thisregion and that the exposed faults are a tongue of high-density material at shallow depths, an likely to serve as shallow conduits for the horizontal interpretationthat is consistentwith a zone of cooleddike (southward)transport of magma.Finally, using processed side intrusionsassociated with Hawaiian-stylerifting. A narrow, scandata in the SouthernRift Zone,Appelgate identifies a extremelyhigh-amplitude magnetic anomaly over the North verylarge (48 km2 in area;estimated 1.8 km 3 in volume)lava JOHNSONAND EMBLEY: INTRODUCTION 12,693 fieldeast of therift zone,which, he speculates, may have been (Figure3). Althoughthe plumes showed distinct 3He and Mn a major contributorin the formationof the summitcaldera. signatures,there was no detectablethermal signatureover the seamount, an observation attributed by Lupton to the high Summit Geology 3He/heatratios and low salinitiesof the fluidsbeing vented within the caldera. The summit of Axial Seamountrepresents a well-studied Thermal anomalies were, however, found by Baker et al., region of recent eruptive volcanics in the submarine who focus on the results of hydrographic and chemical environment and containsevidence of a variety of geological samplingof the water column directly over the summit. Here, processesthat are currently active. Embley et al., using they found temperature anomalies that were restricted to a Sea Beam and Sea MARC I swathmaps, coupled with towed- 200-m-thick layer above the caldera. The summit water- camera and submersible observations, describe the basic temperatureanomaly over Axial is significanfiy smaller than geology of the summit area and the associatedcaldera; these anomalies associatedwith other hydrothermal fields on the data allow the compilation of maps of the summit area Juande Fuca Ridge, perhapsa surprisingresult considering the comparablein detail to those available for subaerialvolcanos. tremendous amount of magmatic activity that must be Embley et al. show that much of the postcalderavolcanism and associated with the formation of the volcano. associatedhydrothermal venting appearsto be associatedwith Feely et al. examine the chemistryof the particulateswithin South Rift Zone activity. The orientation of the 3 x 8 km the water column over the summit region and use the data to summit caldera, with its roughly rectilinear sidesthat trend in differentiate between distinct regions within the general the N160 ø direction, oblique to all other trendsin the area, has summitplume. These three major regionsinclude (1) a single been a major enigmain interpretationof the morphologyof the buoyant plume, (2) an upper nonbuoyantplume, and (3) a seamount. Embley et al. discuss several of the factors that lower nonbuoyant plume. Sediment trap data reported by may be responsiblefor the caldera and favor a model basedon Feely et al. alsoconfirm that the integratedparticulate fallout a mobile "overlappingspreading center." from the summithydrothermal systems is not as large as some The currenttectonic activity within the calderais graphically of the other vent fields from other segments(i.e., Endearour demonstratedin this volume by Fox, who used a precision Segment)on the Juan de Fuca Ridge. pressuresensor to infer movementof the caldera floor over a period of a year. Data from this recorder showed an abrupt 15-cmdownward movement of thecaldera floor over a period HydrothermalVent Fields of several weeks during this period. Simultaneous hydrographicmeasurements showed the presence of a new Althoughlow-level hydrothermal activity and deposits are water temperatureanomaly and a change in bottom current found in severalareas of the caldera(shown by Embley et al.), velocitiesonthe caldera floor occurring atthe same time as mostof the studies in this issue focus on the ASHES vent theproposed vertical movement. Foxattributes thiscorrelation field, which contains theonly high-temperature ventsin the toa periodof activecaldera deflation and a relatedthermal caldera. Most of these investigations weresupported bythe plumewith associated coldwater entrainment. NOAAVENTS Program. The site is ideal for short- or long- term experiments, being located on the flat, unobstructed caldera floor, away from the calderawall. Hammonddescribes WaterColumn Studies thevent field in detail,drawing from the results of Alvinand A structureon the seafloor the size of Axial Seamountcan Pisces IV dives and deep-tow camera photographs. There is presenta formidablebarrier to thefree circulation of a variety of hydrothermaldeposits in the ASHES field, and in the region. Positionedat the intersectionof the C-E becauseof the intensityof interestin the region,the deposits Seamount Chain, the bathymetric expressionof Axial are well-mappedand sampledfor chemistry,geology, and intensifieswater circulationpatterns in the region. Cannon biology. Hammondproposes that the largestfield, the ASHES and Pashinskidiscuss a seriesof year-longmoored current systemnear the southeastcomer of the caldera,is controlled meter measurementsmade both on the summit of the volcano both by the faulting associatedwith the wall formation and by andon thebathymetric saddle that connects Axial withBrown the permeabilityof the lobatepillow flows that makeup the Bear Seamount(Plate A). The authorsinterpret these data to calderafloor in this region. show that there is substantiallong-term averagedsouthward Butterfield et al. presentthe detailed chemicaldata from the flow alongthe entireeast side of the seamount.West of Axial, ASHES vent field, which is the first well-documentedexample in the saddleseparating it from Brown Bear Seamount,lies a of a phase-separateddeep-sea hydrothermal system. Using the zone of convergence between southward and northward extensivecollection of water samplestaken from the ASHES flowing currents. After converging,these two currentsseem field in 1986, 1987, and 1988, they describethe geochemical to mergeand flow westwardaway from the ridge. In addition effectsof boiling on a hydrothermalsystem. They observea to being importantin their own fight, thesecurrent systems wide rangeof fluid compositionsfrom the field, includingboth also determine the ultimate fate of the hydrothermaleffluent gas-rich and gas-poor compositionsand both high and low that issuesfrom the summit of the active volcano. chlorinity values. Butterfield et al., however,propose that the The hydrothermalvent systemson the summit of Axial entirerange of observedcompositions can be explainedby the Volcano make a detectableheat and chemical signal in the process of boiling and subsequentphase segregationof a overlying water column. Lupron describesthe data from hydrothermalfluid with a single original composition. Fox hydrographicsurveys over andnear the volcano, taken between describesthe physicalseparation of the brine andvapor phases 1980and 1987, which show complex patterns of •He-and Mn- thatresult from a "boiling"vent system and presents a relative rich plumeswithin the water column. Theseplumes represent permeability model within the fluid conduits as a possible regionalconcentrations, as well as local effects,over both the mechanismto produce that separation. Litfie et al. discuss summit and a local topographiclow known as Helium Basin both the theoretical"noise" produced by hydrothermalsystems 12,694 JOHNSONAND EMBLEY: INTRODUCTION in generaland actualsound measurements of noisegenerated Although, as stated earlier, Axial Volcano is already a by vents at the ASHES field. remarkablywell-studied piece of seafloor,the recentvolcanic The natural distributionof hydrothermalfauna, and the eruptions, the concurrent hydrothermal activity, and the impactof geologicalcontrols on this distribution,is described proximityof the seamountto NorthAmerican west coast ports by Arquit. The major impactof submersiblesampling on the virtually guaranteethat Axial Seamountwill continueto be an hydrothermalfauna is discussedby Tunnicliffe,who points out intenselyinteresting natural laboratory for a longtime to come. that vent animalsare very sensitiveto disruption,in spite of living in the extremely harsh environment of a high- REFERENCES temperature vent. Atwater, T., Implication of plate tectonicsfor the Cenozoictectonic evolution of western North America, Geol. Soc. Am. Bull., 81, 3513-3536, 1970. CONCLUSIONS Davis, E. E., and J. L. Karsten, On the causeof the asymmetric Although detailed conclusionsare not appropriatefor an distributionof seamountsabout the Juande FucaRidge: Ridge- introductorymanuscript, it is difficult not to make somefinal crestmigration over a heterogeneousasthenosphere, Earth Planet. Sci. Left., 79, 385-396, 1986. remarks regarding the current statusof our understandingof Hyndman,R. D., R. P. Riddihough,and R. Herzer, The Nootka fault the remarkablegeological feature representedby Axial. zone--A new plate boundaryoff westernCanada, Geophys. J. R. Clearly, Axial Seamountis an active underseavolcano, with Astron. Soc., 58, 667--683, 1979. recentlava flows, active hydrothermal systems, and a calderaJohnson, H. P., and M. L. Holmes,Evolution inPlate Tectonics: the thatmay be presently undergoing inflation/deflation cycles. vol.Juan N,de DecadeFuca Ridge, of NorthinThe AmericanEastern Pacific Geology, Ocean edited and by Hawaii, E. L. Thegravity and magnetics data argue for either a shallow Winterer,D.M. Hussong, andR. W. Decker, pp.73-92, Geological magma chamberunderlying the summitcaldera, or a warm Societyof America,Boulder, Colo., 1989. and/oraltered interior. The vigoroushydrothermal activity Riddihough,R. P., M. W. Beck,R. L. Chase,E. E. Davis,R. D. associatedwith the caldera, while (surprisingly) notof the Hyndman,S.H. Johnson, andG. C. Rogers, inGeodynamics ofthe EasternPacific Region, Caribbean and ScotiaArcs, Geodyn. Ser., samescale as the large fields found on the Endearour Segment vol. 9, edited byR. Cabre, pp.5-23, AGU, Washington, D.C.,1983. and ExplorerRidge to the north,still have a major impacton the regionalwater chemistry, biology, and perhaps hydrology. The detailedstudies of the ASHES vent field have,for the first R.W. Embley,NOAA, PacificMarine Environmental Laboratory, time,documented the chemistryof a phase-separation-Hatfield Marine Science Center, Newport, OR97365. H. P. Johnson,School of Oceanography,University of Washington, dominatedhydrothermal system. Presently only observed on Seatfie,WA 98195. Axial Seamount,this processhas a major impact on the chemistry and geology of the vent field and will no doubt be seenon manyof the shallowsubmarine volcanos of the (ReceivedFebruary 10, 1990; westernPacific in the future. acceptedFebruary 10, 1990.)